March 2020

Center for Disease Control (CDC) Up to date information on the Coronavirus COVID19
Coronavirus Resources in the Work Place: US Department of Labor
World Health Organization (WHO) Coronavirus COVID19

December 2019

Permanent Hair Dye and Straighteners May Increase Breast Cancer Risk

Scientists at the National Institutes of Health found that women who use permanent hair dye and chemical hair straighteners have a higher risk of developing breast cancer than women who don’t use these products. The study published online Dec. 4 in the International Journal of Cancer and suggests that breast cancer risk increased with more frequent use of these chemical hair products.

Using data from 46,709 women in the Sister Study, researchers at the National Institute of Environmental Health Sciences (NIEHS), part of NIH, found that women who regularly used permanent hair dye in the year prior to enrolling in the study were 9% more likely than women who didn’t use hair dye to develop breast cancer. Among African American women, using permanent dyes every five to eight weeks or more was associated with a 60% increased risk of breast cancer as compared with an 8% increased risk for white women. The research team found little to no increase in breast cancer risk for semi-permanent or temporary dye use.

“Researchers have been studying the possible link between hair dye and cancer for a long time, but results have been inconsistent,” said corresponding author Alexandra White, Ph.D., head of the NIEHS Environment and Cancer Epidemiology Group. “In our study, we see a higher breast cancer risk associated with hair dye use, and the effect is stronger in African American women, particularly those who are frequent users.”

An intriguing finding was the association between the use of chemical hair straighteners and breast cancer. Dr. White and colleagues found that women who used hair straighteners at least every five to eight weeks were about 30% more likely to develop breast cancer. While the association between straightener use and breast cancer was similar in African American and white women, straightener use was much more common among African American women.

Co-author Dale Sandler, Ph.D., chief of the NIEHS Epidemiology Branch, cautioned that although there is some prior evidence to support the association with chemical straighteners, these results need to be replicated in other studies.

When asked if women should stop dyeing or straightening their hair, Sandler said, “We are exposed to many things that could potentially contribute to breast cancer, and it is unlikely that any single factor explains a woman’s risk. While it is too early to make a firm recommendation, avoiding these chemicals might be one more thing women can do to reduce their risk of breast cancer.”

Grant Number: Z01ES044005Press Resources:

Dale Sandler, Ph.D.

Dale Sandler, Ph.D.
(Download High Resolution)

Alexandra J. White, Ph.D., M.S.P.H.

Alexandra J. White, Ph.D., M.S.P.H.
(Download High Resolution)


Eberle CE, Sandler DP, Taylor KW, White AJ. 2019. Hair dye and chemical straightener use and breast cancer risk in a large U.S. population of black and white women. Int J Cancer; doi: 10.1002/ijc.32738 [Online 4 December 2019].

Women and Migraine: the Role of Hormones

Current Neurology and Neuroscience Reports

July 2018, 18:42| Cite as

Women and Migraine: the Role of Hormones

  • Candice Todd
  • Ana Marissa Lagman-Bartolome
  • Christine Lay
        1. 1.
        2. 2.

        Headache (R Halker, Section Editor)First Online: 31 May 2018

        Part of the following topical collections:

        1. Topical Collection on Headache


        Purpose of Review

        Migraine is a debilitating disease, that is encountered in countless medical offices every day and since it is highly prevalent in women, it is imperative to have a clear understanding of how to manage migraine. There is a growing body of evidence regarding the patterns we see in women throughout their life cycle and how we approach migraine diagnosis and treatment at those times.

        Recent Findings

        New guidelines regarding safety of medication during pregnancy and lactation are being utilized to help guide management decisions in female migraineurs. There is also new data surrounding the risk of stroke in individuals who suffer from migraine with aura.


        This article seeks to provide an overview of a woman’s migraine throughout her lifetime, the impact of hormones and an approach to management.


        Migraine Estrogen Serotonin Menstrual migraine Oral contraceptive Stroke Pregnancy Perimenopause Menopause 



        Menstrual migraineOCP

        Oral contraceptive pillHRT

        Hormone replacement therapy

        This article is part of the Topical Collection on Headache


        Migraine is an exceptionally common medical condition, recently ranked by the World Health Organization as the third most prevalent and if not managed well, it can be extremely disabling. The incidence of migraine attacks varies considerably between individuals; some women have several attacks a month, while others have less than one a year. Even within the same individual, there may be periods of frequent attacks followed by prolonged migraine-free periods. The effect of migraine on ability to function and the degree of disability it causes also varies, ranging from mild impairment to being unable to work or to enjoy social activities. In 2016, the World Health Organization ranked migraine as the second leading cause of disability [12].

        Migraine is an attack of head pain lasting 4–72 h that is typically disabling and impairing routine activity, and is associated with photophobia and phonophobia, and often with nausea and in some cases vomiting [3].

        Migraine Is More Common in Women

        As migraine is three times more common in women than in men, it behooves clinicians to be well versed in effective diagnosis and treatment. In women, migraine has a lifetime prevalence of 43% and migraine without aura, including menstrual migraine, is the most prevalent subtype [4]. The American Migraine Prevalence and Prevention study notes a higher prevalence of migraine in women as compared to men over the age of 12, which is also consistent across all ethnicities [567]. It is a disorder than impacts women throughout their lifetimes from prepuberty childhood years through to the post-menopausal years. Women report more migraine-related symptoms (i.e., photophobia, phonophobia, nausea, vomiting, and cutaneous allodynia) and more migraine-related disabilities [7]. Menarche is a common time for a teenage girl to experience her first migraine [8].

        It is thought that migraine is more prevalent in women during reproductive years and it is well known that there is a strong relationship between headache and hormone homeostasis, particularly related to ovarian hormones [9]. Estrogen plays a key role in migraine and during times of fluctuation such as puberty, menstruation, pregnancy, and the peri- and post-menopausal years, these fluctuations in estrogen have an impact on migraine. Sex hormones can act as important modulators hence the different effects of migraine in men and women [10•].

        Stroke and Migraine

        While migraine with aura is less common, it is important to understand and recognize it, as it can have implications in hormonal therapies commonly used by women, including the birth control pill and hormone replacement therapy [1112]. Perhaps up to 25% of migraine sufferers also experience an aura of transient neurological symptoms, which typically begin before the headache itself. These symptoms are reversible and often dissipate before or as the headache sets in. Most commonly, aura consists of visual changes with spreading fortification spectra or scotoma, most often in both visual fields. Some patients experience sensory symptoms with numbness or pins and needles on one side of the face and/or body, but often these patients also have visual aura as well [13]. A study recently found that a migraineur who has aura is at an increased risk of stroke, but the risk depends on the frequency of the aura with risk being increased twofold with < 1 aura/month and a fourfold increased if aura is > 1/week [14•].

        While data surrounding the issues of ischemic stroke in a migraineur who is using hormonal contraception is overall poor, emerging evidence and expert opinion suggest that recommendations can be considered. Many previous studies were observational; the type of migraine was not clearly defined and often the specific type of hormone and dose used were not indicated. In a recent study by Champaloux and from the European Headache Federation consensus statement, it is clear that hormonal contraceptive use in women with migraine carries an increased risk of stroke [15••, 16]. While the absolute risk of stroke among women of reproductive age is low, with an incidence of 4.3–8.9 per 100,000 per year, caution should nonetheless be taken when considering hormonal contraception in a migraineur. The consensus statement recommends considering non-hormonal contraception in women with migraine as a first-line method. For a woman with migraine without aura, using hormonal contraception, the risk of stroke is increased fourfold and were she to have aura, the risk is increased sixfold [15••, 1617].

        There are several hypotheses to support the relationship between migraine with aura and ischemic stroke. The concept of migrainous infarction is known; however, in these circumstances, to report the stroke as a migrainous infarction, the symptoms of stroke must exactly match the patient’s known previous aura symptoms. Some believe that women with migraine with aura may have a high prevalence of other vasculopathies that include antiphospholipid syndrome and systemic lupus erythematosus, and/or have vascular risk factors, such as smoking and hypertension, which place them at higher risk of stroke [17]. Migraine with aura has also been linked to patent foramen ovale, which in the right circumstances can result in paradoxic microembolization; however, based upon recent studies, at this time, there is no definitive evidence to recommend PFO closure as a treatment for migraine [18].

        Menstruation and Migraine

        Menstrual migraine (MM) consists of both pure menstrual migraine (PMM) and menstrually related migraine (MRM). MM most frequently occurs in the second decade of life around the onset of menarche and is typically without aura. Prevalence peaks around age 40 and as menopause approaches, prevalence declines. PMM attacks may occur before, during, or after menstruation and these attacks usually occur in at least two out of every three cycles with no migraine at any other time of the month [192021]. It is estimated that only 7 to 35% of women experience PMM. MRM occurs in 60% of women who have migraine attacks perimenstrually as well as at other times of the month [22]. The attacks are often more severe, of longer duration, and less responsive to both acute and prophylactic treatment than migraines occurring at other times of the cycle [2023]. Menstrual migraine typically occurs 2 days prior to or in the first 3 days of the cycle. Premenstrual headache occurs earlier in the cycle, typically 2 to 7 days before the onset of menses and may be part of premenstrual syndrome (PMS) [3]. It is important to note that premenstrual headache typically resolves with the onset of menstruation [20]. Diagnosis can be further verified by having the adolescent or woman keep a calendar or electronic diary of her headache days and menstrual cycles.

        The drop in estrogen during the luteal phase is believed to trigger migraine and this decline may impact blood vessels making them more permeable to pro-inflammatory mediators such as prostaglandins. Prostaglandin levels are elevated threefold in the luteal phase with a further increase during menstruation, and therefore thought to play a role in MRM [924]. MM is often less responsive to acute and abortive therapies [22]. However, prostaglandin inhibitors, nonsteroidal anti-inflammatory drugs (NSAIDs), are effective in treating and preventing MM in some women, thus supporting the role of prostaglandins in MM. Estrogen has been associated with a number of changes in serotonin, including reduced uptake, decreased degradation, and increased production, and thus the fluctuating levels of estrogen throughout a woman’s lifetime likely have an impact on serotonin, leading to physiological consequences, including an impact on migraine. Estrogen increases serotonergic tone and concentration of beta-endorphins hence withdrawal of estrogen can reduce serotonergic tone and change central opioid tonus which results in higher susceptibility to migraine [2425].

        Oral Contraceptives and Migraine

        Despite advances in non-oral contraception methods, oral contraceptive pills (OCP) remain a popular option for women; however, in a migraineur with aura, this may pose a challenge. Migraine is common in the childbearing years, and thus the question of birth control will arise whether in the adolescent girl or in the younger through older-aged premenopausal woman [8]. It may be difficult to predict with certainty what will happen to the migraine pattern when a woman goes on an OCP, as some women may improve, some women may experience more disabling migraine attacks, often during the placebo week, and some women will not experience any change. For the prescribing physician, OCP’s are generally considered safe in women under age 45, who have migraine without aura and who do not have vascular risk factors, such as smoking. If after initiating an OCP a woman experiences a new aura or a change in her typical aura, the OCP should be discontinued since as previously noted, there is a potential increase in the risk of an ischemic event [26].

        When prescribing an OCP, caution must be exercised. It is optimal to choose a pill that is low-dose (20 μg or less) and monophasic, since the “active” pills contain the same amount of ethinylestradiol and progestin. Biphasic and triphasic pills may not be the best choice in migraineurs because of the fluctuating levels of ethinylestradiol and progestin. Of note, in women who suffer from menstrual migraine, skipping the placebo week may reduce the number of menstrual cycles and therefore reduce the number of menstrually related migraine attacks [14•, 27]. However, there remains poor evidence that oral contraceptive pills prevent migraine. In women for whom estrogen-containing contraceptives are contraindicated, a progesterone-only pill can be considered; however, a meta-analysis by Warhurst et al. concluded that while some observational studies found a benefit in migraine day reduction, there were no randomized trials and thus data is not definitive. Adding on estrogen during the placebo week can blunt the impact of the naturally falling estrogen, thus limiting the menstrual migraine [28].

        Treatment of MM can be difficult as often the same effective therapy for a non-MRM attack may not be sufficient for a MM attack. A diary will help the patient identify the relationship between her migraines and her menstrual cycle and non-pharmacologic therapies including sleep routine, hydration, regular meals, and avoidance of known triggers are important. A trigger in a non-MM attack may be different from the trigger for a MM attack. Triptans have shown clear benefit; however, often the addition of NSAIDs or antiemetic therapy will be required. If migraines are occurring 3–4 times per month or the MM attack is not responsive to acute therapy, preventative options, including mini-prophylaxis, must be considered. Both hormonal and non-hormonal prophylactic options are available for mini-prophylaxis, with the latter including the perimenstrual use of standard migraine prophylactic agents and often NSAIDs. Naproxen sodium (550 mg BID) or mefenamic acid (500 mg TID) may be used effectively 2–4 days prior to the MM and continued through day 3 of menstrual flow [2021].

        Triptans including sumatriptan (25 mg TID), naratriptan (1 mg BID), and frovatriptan (2.5 mg BID) are effective for mini-prophylaxis. Patients should be reminded that the triptan is started 2 days prior to the onset of the MM and is continued for a total of 3–5 days. Standard migraine prophylactic medications may be used for 5–7 days prior to the onset of menses and continued through to the end of the vulnerable time period for migraine [31]. For women who are currently taking a preventative agent, the dose can be transiently increased to help prevent the MM attack [24].

        Hormonal prophylaxis to counteract or prevent the luteal phase drop in estrogen may be considered for refractory MM provided there are no contraindications to estrogen therapy, no vascular risk factors, and the lowest effective dose is utilized. While many women prefer oral medications, transdermal or intravaginal may provide more stable estrogen levels. Studies have demonstrated the efficacy of percutaneous estradiol at a dose of 1.5 mg QD, for the 3 days prior to menses and continued for a total of 6 days, as well as transdermal estradiol 100 μg patch placed 3 days prior to menses, then replaced 1 day prior to menses and replaced again on day 2 after menses begins. In refractory cases, possible use of synthetic androgens, tamoxifen, bromocriptine, and GnRH analogs can be considered. No long-term or controlled studies have been undertaken evaluating hysterectomy or oophorectomy in the treatment of MM and in fact, in one study, two thirds of patients undergoing surgical menopause got worse [24293031].

        Pregnancy and Migraine

        A significant challenge in managing migraine during the childbearing years is the safe treatment of symptoms pre- and post conception and since up to 50% of pregnancies are unplanned, care providers ought to discuss and document risks with patients and choose medications with no known risk of teratogenicity. Hormonal changes, stress, disrupted sleep, nausea, and dehydration of pregnancy can impact migraine. As estrogen levels are rise during the first trimester, transient worsening may occur. Up to 87% improvement in headache symptoms may occur, primarily in the second and third trimester, although migraine can persist beyond the first trimester. Studies suggest that severe maternal migraines may increase the occurrence of adverse delivery outcomes [3233]. One retrospective study found that patients had elevated rates of preeclampsia, preterm birth, and low birthweight but decreased cesarean delivery rates. Migraine presenting for the first time during pregnancy is uncommon and often presents as migraine with aura [34].

        It is imperative to rule out secondary headache due to an underlying pathology, but then simple strategies such as reassurance, rest, ice, acupuncture, biofeedback, and short-term disability leave from work can be helpful. For women who have severe or prolonged migraine, accompanied by nausea, vomiting, and dehydration, medical therapy may be indicated. In the first trimester, small doses of caffeine and acetaminophen are considered safe [35]. Several NSAIDs are generally considered safe in early pregnancy; however, they should be used with caution in the third trimester, as there is an increased risk of premature closure of the ductus arteriosus, impaired renal function, cerebral palsy, and neonatal intra-ventricular hemorrhage [36•]. Some narcotics, including meperidine and morphine, can be used safely, except in the third trimester but their use is cautioned and generally not recommended. Codeine is sometimes used in pregnancy and a study looking at this found no increased risk of fetal malformations in pregnancy outcomes of codeine-exposed compared to non-codeine-exposed pregnancies; however, its use in the third trimester should be cautioned as there was an increased risk of acute C-section delivery and postpartum hemorrhage [36•, 37]. For severe attacks, intravenous fluids and an intravenous antiemetic may be indicated. In suppository form, prochlorperazine, metoclopramide, and promethazine are generally considered safe [36•]. In women who are in status migrainosus and require steroids, prednisone is often preferred to dexamethasone, because dexamethasone can cross the placenta more easily. In general, triptans, ergots, and aspirin should be avoided; however, under careful consideration and monitoring, this may be implicated in specific scenarios. Pregnancy registries have been established to monitor women who have taken triptans while pregnant and yet there remains poor evidence on its specific impact on fetal wellbeing. While a meta-analysis suggested a possible risk of spontaneous abortion in triptan-exposed pregnancies, the evidence is neither conclusive nor causative and requires further investigation [36•, 38]. A recent retrospective study showed no increased risk of behavioral issues in 5-year-old children exposed to triptans in fetal life [39]. When migraine is frequent and disabling, preventative therapy may be required but should be undertaken with the full consent of both the patient and her partner. Beta-blockers (labetalol and propranolol) are common first-line agents for prophylaxis but have been thought to carry a risk of intrauterine growth retardation and preterm birth. However, these studies employed higher doses and the patient population confounded the interpretation of the results. It is suggested to attempt weaning off beta-blockers in the third trimester in order to mitigate these risk factors. Occasionally, amitriptyline or fluoxetine are employed when beta-blockers are contraindicated, but again the risk of fetal malformation has been reported and this should be discussed with patients [3236•].

        Breastfeeding and Migraine

        Migraine often recurs in the postpartum period, presumably related to the rapid fall in estrogen. Although breastfeeding women have a lower recurrence rate of migraine than bottle feeding women, more than half of breastfeeding women still experience migraine recurrence within 1 month from delivery. Therefore, in some women, it may be necessary to treat during lactation [40]. When choosing a medication, consider the drug as well as the protein binding capabilities of the drug, as these factors influence maternal plasma levels and, in the case of a highly protein bound drug, entry into breast milk may be impeded. Oral bioavailability can impact the amount of drug that enters the plasma compartment. A review of the literature and safety profile of acute migraine drugs during lactation found that low-dose aspirin, ibuprofen, acetaminophen, caffeine, metoclopramide, eletriptan and sumatriptan can be used safely [4142]. For many lactating mothers, pumping and dumping breast milk for 4–6 h after taking a medication provides some additional reassurance. Preventative medications that are contraindicated during lactation include among many, high-dose aspirin atenolol, nadolol, cinnarizine, flunarizine, ergotamine, methysergide, and pizotifen [4041].

        Perimenopause, Menopause, and Migraine

        Migraine tends to improve with age, especially in women who have MM and while women expect that following menopause, a reduction in headache will occur, but this transition could take many years. On average, women enter perimenopause in their mid-forties, or earlier, and tend to reach menopause by age 55 [43]. Unfortunately, the perimenopausal transition can give rise to a temporary worsening of headaches due to significant fluctuations in hormone levels, mainly estrogen levels. With this increased headache frequency, patients should be cautioned regarding medication overuse. Migraineurs who “grew out” of their migraines may see a return of their headache symptoms and in women with MRM, there may be an increase in frequency and severity of their headaches due to the loss of the predictable menstrually related pattern. Infrequent and unpredictable menses poses a challenge to the patient as well as the treating physician. Perimenopause can be associated with fatigue, insomnia, irritability, night sweats, hot flashes, forgetfulness, drop in libido, and difficulty concentrating, which may contribute to migraine, especially when hormone replacement therapy (HRT) is deemed necessary [4344••].

        HRT is commonly provided to women during perimenopause and depending on the duration, type, dose, and route of administration of estrogen, migraine can be exacerbated, unchanged, or improved. Generally, HRT is not prescribed for migraine alone, especially in a migraineur who has aura or other vascular risk factors, but rather when it is deemed necessary to control the symptoms of peri-or post-menopause. If HRT worsens migraine, clinicians will have to elicit new strategies for treatment, such as dose reduction, non-cycling estrogen use, switching from conjugated estrogens to pure estradiol or from synthetic to bioidentical estrogen, with the goal being to attempt to provide stable physiologic levels. If adjustments in the estrogenic component of the HRT are unsatisfactory and migraine continues to be poorly controlled, the decision should be addressed as to whether to cease HRT completely versus adjustment of the progesterone component. Other non-hormonal options include gabapentin, venlafaxine, and a myriad of natural supplements [274546].


        Migraine is a lifelong disorder in women, often beginning just before or around puberty and continuing through the post-menopausal years. Fortunately, through standard migraine lifestyle recommendations, abortive and preventative therapies including hormonal options, the majority of women can experience improvement in their migraines and associated symptoms. Clinicians need to be well versed in this extremely common and disabling condition, including the implications of hormonal therapies on migraine, as well as the concerns regarding hormonal use in a woman who has migraine with aura or other vascular risk factors. Clearly, more definitive research on the use of current low-dose hormonal therapies in a migraineur is needed to help guide treatment.


        Compliance with Ethical Standards

        Conflict of Interest

        Candice Todd, Ana Marissa Lagman-Bartolome, and Christine Lay each declare no potential conflicts of interest.

        Human and Animal Rights and Informed Consent

        This article does not contain any studies with human or animal subjects performed by any of the authors.


        Papers of particular interest, published recently, have been highlighted as: • Of importance •• Of major importance

        1. 1.Steiner T, Stovner L, Vos T. GBD 2015: migraine is the third cause of disability in under 50s. The Journal of Headache and Pain. 2016;17:104. Scholar
        2. 2.Global, regional, and national incidence, prevalence, and years lived with disability for 328 diseases and injuries for 195 countries, 1990–2016: a systematic analysis for the Global Burden of Disease Study 2016. The Lancet Neurology. 2017;390:1211–859.Google Scholar
        3. 3.(2018) Headache Classification Committee of the International Headache Society (IHS) The international classification of headache disorders, 3rd edition. Cephalalgia 38:1–211.Google Scholar
        4. 4.Sacco S, Ricci S, Degan D, Carolei A. Migraine in women: the role of hormones and their impact on vascular diseases. J Headache Pain. 2012;13:177–89.CrossRefPubMedPubMedCentralGoogle Scholar
        5. 5.Stewart W. Prevalence of migraine headache in the United States. JAMA. 1992;267:64–9.CrossRefPubMedGoogle Scholar
        6. 6.Stovner L, Hagen K, Jensen R, Katsarava Z, Lipton R, Scher A, et al. The global burden of headache: a documentation of headache prevalence and disability worldwide. Cephalalgia. 2007;27:193–210.Google Scholar
        7. 7.Buse D, Loder E, Gorman J, Stewart W, Reed M, Fanning K, et al. Sex differences in the prevalence, symptoms, and associated features of migraine, probable migraine and other severe headache: results of the American Migraine Prevalence and Prevention (AMPP) study. Headache: J Head Face Pain. 2013;53:1278–99.Google Scholar
        8. 8.Crawford M, Lehman L, Slater S, Kabbouche M, LeCates S, Segers A, et al. Menstrual migraine in adolescents. Headache: J Head Face Pain. 2009;49:341–7.Google Scholar
        9. 9.Martin V, Behbehani M. Ovarian hormones and migraine headache: understanding mechanisms and pathogenesis-part I. Headache: J Head Face Pain. 2006;46:3–23.CrossRefGoogle Scholar
        10. 10.•• Vetvik K, MacGregor E. Sex differences in the epidemiology, clinical features, and pathophysiology of migraine. Lancet Neurol. 2017;16:76–87. The authors review sex differences regarding migraine characteristics and treatment. This paper discusses the possible exogenous and endogenous hormonal influences on genes that lead to women experiencing more migraine than men.CrossRefPubMedGoogle Scholar
        11. 11.Sacco S, Merki-Feld G, Ægidius K, et al. Hormonal contraceptives and risk of ischemic stroke in women with migraine: a consensus statement from the European Headache Federation (EHF) and the European Society of Contraception and Reproductive Health (ESC). J Headache Pain. 2017;18:108. Scholar
        12. 12.ACOG practice bulletin no. 73: use of hormonal contraception in women with coexisting medical conditions. Obstet Gynecol. 2006;107:1453.Google Scholar
        13. 13.Cutrer F, Huerter K. Migraine aura. Neurologist. 2007;13:118–25.CrossRefPubMedGoogle Scholar
        14. 14.• Calhoun A, Batur P. Combined hormonal contraceptives and migraine: an update on the evidence. Cleveland Clin J Med. 2017;84:631–8. This article highlights the issues around oral contraceptive use in women migraineurs with and without aura. It discusses the algorithm physicians should use in order to safely prescribe estrogen-containing contraceptives.CrossRefGoogle Scholar
        15. 15.•• Champaloux S, Tepper N, Monsour M, Curtis K, Whiteman M, Marchbanks P, et al. Use of combined hormonal contraceptives among women with migraines and risk of ischemic stroke. Obstet Gynecol Surv. 2017;72:543–5. This article finds a 6-fold increased risk of stroke in women of childbearing age who use combine oral contraceptive medications and suffer from migraine with aura.Google Scholar
        16. 16.Tepper N, Whiteman M, Zapata L, Marchbanks P, Curtis K. Safety of hormonal contraceptives among women with migraine: a systematic review. Contraception. 2016;94:630–40.CrossRefPubMedGoogle Scholar
        17. 17.Sacco S, Ricci S, Degan D, Carolei A. Migraine in women: the role of hormones and their impact on vascular diseases. J Headache Pain. 2012;13:177–89.CrossRefPubMedPubMedCentralGoogle Scholar
        18. 18.Tariq N, Tepper S, Kriegler J. Patent foramen ovale and migraine: closing the debate-a review. Headache: J Head Face Pain. 2016;56:462–78.CrossRefGoogle Scholar
        19. 19.Vetvik K, Benth J, MacGregor E, Lundqvist C, Russell M. Menstrual versus non-menstrual attacks of migraine without aura in women with and without menstrual migraine. Cephalalgia. 2015;35:1261–8.CrossRefPubMedGoogle Scholar
        20. 20.Sullivan E, Bushnell C. Management of menstrual migraine: a review of current abortive and prophylactic therapies. Curr Pain Headache Rep. 2010;14:376–84.CrossRefPubMedPubMedCentralGoogle Scholar
        21. 21.MacGregor E. Review: menstrual migraine: therapeutic approaches. Ther Adv Neurol Disord. 2009;2:327–36.CrossRefPubMedPubMedCentralGoogle Scholar
        22. 22.Calhoun A. Understanding menstrual migraine. Headache: J Head Face Pain. 2018;58:626–30. Scholar
        23. 23.Güven B, Güven H, Çomoğlu S. Clinical characteristics of menstrually related and non-menstrual migraine. Acta Neurologica Belgica. 2017;117:671–6.CrossRefPubMedGoogle Scholar
        24. 24.MacGregor E. Migraine headache in perimenopausal and menopausal women. Curr Pain Headache Rep. 2009;13:399–403.CrossRefPubMedGoogle Scholar
        25. 25.Sacco S, Ricci S, Degan D, Carolei A. Migraine in women: the role of hormones and their impact on vascular diseases. J Headache Pain. 2012;13:177–89.CrossRefPubMedPubMedCentralGoogle Scholar
        26. 26.Allais G, Chiarle G, Sinigaglia S, Airola G, Schiapparelli P, Bergandi F, et al. Treating migraine with contraceptives. Neurol Sci. 2017;38:85–9.Google Scholar
        27. 27.Lay C, Broner S. Adolescent issues in migraine: a focus on menstrual migraine. Curr Pain Headache Rep. 2008;12:384–7.CrossRefPubMedGoogle Scholar
        28. 28.Warhurst S, Rofe C, Brew B, Bateson D, McGeechan K, Merki-Feld G, et al. Effectiveness of the proestin-only pill for migraine treatment in women: a systematic review and meta-analysis. Cephalalgia. 2017;38:754–64.Google Scholar
        29. 29.Neri I, Granella F, Nappi R, Manzoni G, Facchinetti F, Genazzani A. Characteristics of headache at menopause: a clinico-epidemiologic study. Maturitas. 1993;17:31–7.CrossRefPubMedGoogle Scholar
        30. 30.Calhoun A. Hormonal contraceptives and migraine with aura-is there still a risk? Headache: J Head Face Pain. 2016;57:184–93.CrossRefGoogle Scholar
        31. 31.Lay C, Broner S. Migraine in women. Neurol Clin. 2009;27:503–11.CrossRefPubMedGoogle Scholar
        32. 32.Jarvis S, Dassan P, Piercy C. Managing migraine in pregnancy. BMJ. 2018:k80.Google Scholar
        33. 33.Pakalnis A. Migraine and hormones. Semin Pediatr Neurol. 2016;23:92–4.CrossRefPubMedGoogle Scholar
        34. 34.Grossman T, Robbins M, Govindappagari S, Dayal A. Delivery outcomes of patients with acute migraine in pregnancy: a retrospective study. Headache: J Head Face Pain. 2017;57:605–11.CrossRefGoogle Scholar
        35. 35.Wells R, Turner D, Lee M, Bishop L, Strauss L. Managing migraine during pregnancy and lactation. Curr Neurol Neurosci Rep. 2016;16:40. Scholar
        36. 36.• Negro A, Delaruelle Z, Ivanova T, Khan S, Ornello R, Raffaelli B, et al. Headache and pregnancy: a systematic review. J Headache Pain. 2017; article reviews the literature regarding the safety of pharmacological treatment of migraine during both the peri-partum and post-partum period. The authors update readers on the change in safety classifications.
        37. 37.Nezvalová-Henriksen K, Spigset O, Nordeng H. The effects of codeine on pregnancy outcome: results from a large population-based cohort study. Reprod Toxicol. 2011;31:263–3.Google Scholar
        38. 38.Marchenko A, Etwel F, Olutunfese O, Nickel C, Koren G, Nulman I. Pregnancy outcome following prenatal exposure to triptan medications: a meta-analysis. Headache: J Head Face Pain. 2015;55:490–501.CrossRefGoogle Scholar
        39. 39.Harris G, Wood M, Ystrom E, Nordeng H. Prenatal triptan exposure and neurodevelopmental outcomes in 5-year-old children: follow-up from the Norwegian mother and child cohort study. Paediatr Perinat Epidemiol. 2018;
        40. 40.Hoshiyama E, Tatsumoto M, Iwanami H, Saisu A, Watanabe H, Inaba N, et al. Postpartum migraines: a long-term prospective study. Intern Med. 2012;51:3119–23.Google Scholar
        41. 41.Davanzo R, Bua J, Paloni G, Facchina G. Breastfeeding and migraine drugs. Eur J Clin Pharmacol. 2014;70:1313–24.CrossRefPubMedGoogle Scholar
        42. 42.Hutchinson S, Marmura M, Calhoun A, Lucas S, Silberstein S, Peterlin B. Use of common migraine treatments in breast-feeding women: a summary of recommendations. Headache: J Head Face Pain. 2013;53:614–27.CrossRefGoogle Scholar
        43. 43.Ibrahimi K, Couturier E, MaassenVanDenBrink A. Migraine and perimenopause. Maturitas. 2014;78:277–80.CrossRefPubMedGoogle Scholar
        44. 44.•• Martin V, Pavlovic J, Fanning K, Buse D, Reed M, Lipton R. Perimenopause and menopause are associated with high frequency headache in women with migraine: results of the American Migraine Prevalence and Prevention Study. Headache. J Head Face Pain. 2016;56:292–305. This is a cross-sectional study that reviews the frequency of headache during the various stages of menopause, particularly the perimenopausal period.CrossRefGoogle Scholar
        45. 45.Lauritsen C, Chua A, Nahas S. Current treatment options: headache related to menopause—diagnosis and management. Curr Treat Options Neurol. 2018;20
        46. 46.Marino J, McNamara H, Hickey M. Managing menopausal symptoms after cancer: an evidence-based approach for primary care. Med J Aust. 2018;208:127–32.CrossRefPubMedGoogle Scholar

        Copyright information

        © Springer Science+Business Media, LLC, part of Springer Nature 2018

        About this article

        CrossMarkCite this article as:Todd, C., Lagman-Bartolome, A.M. & Lay, C. Curr Neurol Neurosci Rep (2018) 18: 42.

        • First Online31 May 2018
        • DOI
        • Publisher NameSpringer US
        • Print ISSN1528-4042
        • Online ISSN1534-6293
        Overweight and obesity are associated with clustering of metabolic risk factors in early pregnancy and the risk of GDM

        Overweight and obesity are associated with clustering of metabolic risk factors in early pregnancy and the risk of GDM

        Overweight and obesity are associated with clustering of metabolic risk factors in early pregnancy and the risk of GDM

        • I-Weng Yen, 
        • Chien-Nan Lee, 
        • Ming-Wei Lin, 
        • Kang-Chih Fan, 
        • Jung-Nan Wei, 
        • Kuan-Yu Chen, 
        • Szu-Chi Chen, 
        • Yi-Yun Tai, 
        • Chun-Heng Kuo,  …




        Overweight and obesity are important risk factors of gestational diabetes mellitus (GDM). Clustering of metabolic risk factors in early pregnancy may be a potential pathogenesis between the link of overweight/obesity and GDM. Since it remains unexplored, we investigated if overweight and obesity are associated with clustering of metabolic risk factors in early pregnancy and the risk of GDM in this cohort study.


        Total 527 women who visited National Taiwan University Hospital for prenatal care in between November 2013 to April 2018 were enrolled. Risk factors of GDM in the first prenatal visit (FPV) were recorded. Overweight/obesity was defined if body mass index ≥24 kg/m2. GDM was diagnosed from the result of a 75g oral glucose tolerance test in 24–28 gestational weeks.


        Overweight/obesity was associated with clustering of metabolic risk factors of GDM, including high fasting plasma glucose, high HbA1c, insulin resistance, high plasma triglyceride and elevated blood pressure in FPV (p<0.05). There was a positive relationship between the number of metabolic risk factors and the incidence of GDM (p <0.05). The odds ratios of HbA1c and diastolic blood pressure were higher in overweight/obese women, compared with those in normal-weight women.


        Overweight/obesity is associated with clustering of metabolic risk factors in early pregnancy, which is correlated with higher risk of GDM. Our findings suggest that metabolic risk factors during early pregnancy should be evaluated in overweight/obese women.


        Fig 2
        Table 1
        Table 2
        Fig 1
        Fig 2
        Table 1
        Table 2
        Fig 1

        Citation: Yen I-W, Lee C-N, Lin M-W, Fan K-C, Wei J-N, Chen K-Y, et al. (2019) Overweight and obesity are associated with clustering of metabolic risk factors in early pregnancy and the risk of GDM. PLoS ONE 14(12): e0225978.

        Editor: Frank T. Spradley, University of Mississippi Medical Center, UNITED STATES

        Received: August 28, 2019; Accepted: November 15, 2019; Published: December 3, 2019

        Copyright: © 2019 Yen et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

        Data Availability: All relevant data are within the paper and its Supporting Information files.

        Funding: This work is supported in part by grants (MOST 103-2314-B-002-157-MY2 and MOST 106-2314-B-002-197) from the Ministry of Science and Technology, Taiwan, and a grant (NTUH.107-S3744) from National Taiwan University Hospital, Taiwan to to H-YL and C-NL.

        Competing interests: The authors have declared that no competing interests exist.


        Gestational diabetes mellitus (GDM) is defined when carbohydrate intolerance is developed or recognized during pregnancy for the first time [1]. GDM is prevalent among pregnant women. According to the report of International Diabetes Federation, the prevalence of GDM is about 14% worldwide in 2017 [2]. GDM results in increased risk of adverse pregnancy outcomes, including macrosomia, premature birth, hypoglycemia at birth, neonatal jaundice and congenital anomalies [3]. Furthermore, it is associated with a higher incidence of type 2 diabetes after delivery [4].

        Obesity is an important risk factor of GDM [5]. Women with pre-pregnancy BMI over 30 have been shown to have a 3-fold increased risk of GDM, compared to women with normal weight before pregnancy [6]. Several mechanisms have been proposed, including elevated pro-inflammatory cytokines in maternal and fetal circulations and inflammation at placenta [78]. In addition, clustering of metabolic abnormalities in obese women at early pregnancy may be another pathophysiology for the link between obesity and GDM. In non-pregnant status, metabolic abnormalities such as hypertension, central obesity, insulin resistance and atherogenic dyslipidemia tend to cluster. Clustering of these metabolic abnormalities is associated with the development of type 2 diabetes mellitus in the future [9]. During early pregnancy, clustering of these metabolic risk factors has been reported to correlate with increased risk of GDM [10]. Nonetheless, there is no report investigating the relationship among obesity, clustering of metabolic risk factors and GDM in early pregnancy.

        In this cohort study, we enrolled 527 pregnant women and recorded their metabolic risk factors in early pregnancy. The relationship between overweight/obesity and clustering of metabolic risk factors were investigated. Moreover, the effect of clustering on the risk of GDM was explored.

        Materials and methods


        We conducted a prospective cohort study, which recruited all the pregnant women having visited the obstetric clinic for prenatal care at National Taiwan University hospital obstetric clinic from November 2013 to April 2018. The pregnant women with overt diabetes, defined as diabetes diagnosed before pregnancy or at the first prenatal visit, and those with twin or multiple pregnancies, were excluded. The medical history, findings from physical examination, and results of laboratory tests of the participants were recorded at the first prenatal visit. All participants underwent a 75g oral glucose tolerance test (OGTT) at 24th–28th gestational weeks to diagnose gestational diabetes. Body mass index (BMI) was calculated by body weight in kilograms divided by the square of body height in meters. Plasma glucose, hemoglobin A1c (HbA1c), total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), plasma triglyceride (TG) and C-peptide were measured with an autonomic analyzer (Toshiba TBA 120 FR, Toshiba Medical Systems Co., Ltd., Tokyo, Japan). Written informed consent was obtained from each patient before enrollment in the cohort. This study was reviewed and approved by the Institutional Review Board of National Taiwan University Hospital.


        Family history of diabetes was defined if one of the parents has diabetes. Overweight/obesity was defined as BMI ≥24 kg/m2. Updated computer models for homeostasis model assessment were used for calculation of HOMA2-IR. Insulin resistance was defined if HOMA2-IR of the subject was ≥ 25 percentile in this cohort. GDM was diagnosed according to the American diabetes association (ADA) criteria. Specifically, the diagnosis was made when any of the following criteria were met during an OGTT: 1). FPG ≥92 mg/dL (5.1 mmol/L); 2). 1-hour plasma glucose during OGTT (1hPG) ≥180 mg/dL (10.0 mmol/L); 3). 2-hour plasma glucose during OGTT (2hPG) ≥153 mg/dL (8.5 mmol/L).

        Statistical analysis

        Data were presented as means and standard deviations for continuous variables, and as number and percentages for categorical variables. Student’s t-tests, Chi-squared tests and Fisher’s exact tests were used to identify the differences in clinical characteristics between the GDM and non-GDM groups. Linear regression were used to compare the difference in the clustering of risk factors between subjects with overweight/obesity and subjects with normal BMI. A logistic regression analysis was applied for the relationship between the number of risk factors and GDM, using GDM as the dependent variable. Logistic regression analyses were performed to identify important risk factors and to estimate their odds ratios of GDM. Interactions between risk factors and overweight/obesity on the risk of GDM were calculated. Statistical analyses were performed using Stata ⁄ SE 14.0 for Windows (StataCorp, College Station, TX, USA). The level of significance for all tests was p<0.05.


        A total of 527 women were included in this study. The median gestational age of the first prenatal visit was 9.7 weeks (inter-quartile range 8.7–11 weeks). Their clinical characteristics in early pregnancy are shown in Table 1. The incidence rate of gestational diabetes was 12.6% in the normal weight group and 20.4% in the overweight/obesity group. Those developing gestational diabetes had older age, higher percentage with history of GDM and family history of diabetes, higher fasting plasma glucose, higher HbA1c, higher HOMA-IR (in the normal weight group), higher plasma triglyceride (in the normal weight group) and higher blood pressure (in the overweight/obesity group).



        Table 1. Clinical characteristics of the study subjects in early pregnancy by gestational diabetes mellitus (GDM) and body mass index (BMI).

        As shown in Table 2, age, fasting plasma glucose, HbA1c, HOMA-IR, plasma triglyceride concentration and blood pressure at the first prenatal visit were significantly associated with the incidence of GDM (all p<0.05). In the overweight/obesity group, age, fasting plasma glucose, HbA1c, systolic blood pressure and diastolic blood pressure at the first prenatal visit were significant risk factors of GDM. Among all the risk factors, there were significant interactions for HbA1c, diastolic blood pressure, BMI group and GDM (both p<0.05). In other words, the odds ratios of GDM for HbA1c and diastolic blood pressure were significantly higher in overweight/obesity group than those in the normal weight group. Besides, plasma TG at FPV, but not other risk factors, was significantly associated with the gestational age at FPV (p<0.05). Therefore, we adjusted gestational age at FPV to explore the interaction between plasma TG at FPV, overweight/obesity and GDM. The result showed that the interaction was not statistically significant (p = 0.244). We performed linear regression for HbA1c and HOMA2-IR of these two groups, which showed: 1. Every 1% elevation of HbA1c increased HOMA2-IR 1.14 in normal weight group and 1.57 in overweight/obesity group. The p value of interaction between BMI and HbA1c to the influence of HOMA2-IR was 0.025, statistically significant. 2. Every 1 mmHg elevation of diastolic pressure increased HOMA2-IR 1.005 in normal weight group and 1.013 in overweight/obesity group. The p value of interaction between BMI and diastolic pressure to the influence of HOMA2-IR was 0.022, also statistically significant.



        Table 2. Interaction of overweight and other factors at the first prenatal visit (FPV) on the risk of gestational diabetes (GDM).

        Odds ratios of every 1 standard deviation increase in risk factors for GDM were shown.

        Fig 1 shows the distribution of different numbers of risk factors in pregnant women with normal weight and overweight/obesity. Women in the overweight/obesity group tented to have more risk factors than those in the normal weight group (p<0.05), which means that risk factors are clustered in women with overweight/obesity. Further adjustment for gestational age resulted in similar finding (p<0.05)



        Fig 1. The proportion of different numbers of risk factors in early pregnancy in pregnant women with normal weight and overweight/obesity.

        Risk factors were measured in the first prenatal visit and included impaired fasting glucose (≥92mg/dl), HbA1c ≥5.7%, insulin resistance (HOMA2-IR ≥75 percentile), plasma triglyceride ≥150mg/dl and blood pressure ≥130/85mmHg. * p<0.05 vs. BMI<24.

        Fig 2 shows the relationship between the numbers of risk factors in early pregnancy and the incidence of GDM in both normal weight and overweight/obesity group. In normal weight group, the incidence of GDM was 8.7% in women without any risk factor, 16.5% in women withone risk factor (p = 0.041 vs. women without risk factors), 28% in women with two risk factors (p<0.01 vs. women without risk factors) and 66.7% in women with more than three risk factors (p<0.01 vs. women without risk factors). In overweight/obesity group, the incidence of GDM was 10.5% in women without any risk factor, 11.5% in women who with risk factor (p = 0.899 vs. women without risk factors), 33.3% in women with two risk factors (p<0.05 vs. women without risk factors) and 43.8% in women with more than three risk factors (p<0.01 vs. women without risk factors). The incidence of GDM increased by the numbers of risk factors in the early pregnancy in both groups (normal weight group, p for trend<0.01, r2 = 0.9115; overweight/obesity group, p for trend<0.01, r2 = 0.8498) Further adjustment for gestational age showed similar results (p<0.01)



        Fig 2. Number of risk factors in early pregnancy and the incidence of GDM.

        (A) normal weight group and (B) overweight/obesity group. * p<0.05 vs. 0 risk factor. P for trend <0.01.


        The major findings of this study are that metabolic risk factors in early pregnancy tend to cluster in pregnant woman with overweight/obesity, and clustering of these risk factors is associated with a higher risk of GDM. Our results suggest that clustering of metabolic abnormalities in overweight/obesity women in early pregnancy may be another pathophysiology for the link between overweight/obesity and GDM. The present study is the first report investigating the relationship among overweight/obesity, clustering of metabolic risk factors in early pregnancy and GDM. However, there are two articles including obesity as one of the metabolic risk factors and studying the relationship between clustering of these factors and GDM. In 2009, Chatzi et al. have described that metabolic syndrome can also be found during pregnancy [10]. They defined metabolic syndrome using the same criteria in non-pregnant people, except replacing waist circumference by BMI. Their findings supported the present study that obesity, higher blood pressure, fasting plasma glucose, HOMA2-IR and plasma triglyceride concentrations, and clustering of these factors in the first trimester were associated with higher risk of GDM. In another study, clustering of metabolic factors including obesity was associated with GDM as well as other adverse pregnancy outcomes [11].

        In the present study, we found that obesity was associated with clustering of risk factors including insulin resistance, hypertriglyceridemia and elevated blood pressure during pregnancy. Obesity results in elevated plasma free fatty acid (FFA) levels, which leads to increased intracellular lipid accumulation in non-adipose cells, such as hepatocytes and skeletal muscle cells [1213]. Ectopic lipid accumulation in these cells can result in insulin resistance through the activation of protein kinase C and diacylglycerol pathways [14]. On the other hand, insulin can regulate plasma triglyceride concentrations by downregulation of microsomal triglyceride transfer protein (MTP) and activation of lipoprotein lipase (LDL) [1516]. In insulin resistant status, failure to inhibit MPL and activate LPL would lead to hypertriglyceridemia [17]. In addition, obesity can also activate renin-angiotensin-aldosterone system (RAAS), which is an important cause of hypertension[18]. Taken together, our findings are supported since obesity is a common cause of insulin resistance, hypertriglyceridemia and elevated blood pressure.

        In this study, we found significant interactions between HbA1c, diastolic blood pressure and overweight/obesity on the risk of GDM. In other words, the odds ratio of HbA1c or diastolic blood pressure on GDM was significantly higher in pregnant women with overweight/obesity than those with normal weight. In the literature, HbA1c has been used to define high-risk group of GDM. One report in 2014 suggested the cutoff value to be set at 5.7%, and there was 30% of pregnant woman in their study group having BMI ≥30 kg/m2[19]. Our findings suggest that different cutoff values for HbA1c and blood pressure are needed to define high-risk group of GDM in women with overweight/obesity. However, the underlying pathophysiology for these interactions remains unclear. In our analyses, every unit increased in A1c and diastolic blood pressure was associated with a greater increase in HOMA2-IR in overweight/obese women, although the difference was borderline statistical significant due to the limited sample size. These findings suggest that with similar HbA1c or blood pressure, overweight/obese women were more insulin resistant than women with normal weight, which is a promising mechanism to be explored in the future.

        This study is the first report which explored the relationship of overweight/obesity, clustering of metabolic risk factors and GDM with an adequate sample size. Our findings provide supporting evidence for clustering of metabolic risk factors as a pathogenic link between overweight/obesity and GDM. Besides, these metabolic risk factors were recorded at the first prenatal visit. Therefore, the temporal relationship between clustering of risk factors and GDM is clear. However, this study limited in the detail molecular mechanisms among overweight/obesity, clustering of metabolic risk factors and GDM were not explored, which should be studied in future studies.

        In conclusion, the present study found that metabolic risk factors in early pregnancy tend to cluster in pregnant woman with overweight or obesity, and clustering of these risk factors is associated with a higher risk of GDM. Our findings suggest that clinicians should evaluate metabolic risk factors in early pregnancy, especially in obese or overweight pregnant women.


        The authors would like to thank Mr. Chin-Mao Huang,Ms. Kuei-Chen Chih of National Taiwan University Hospital, and the staff of the eighth Core Lab, Department of Medical Research, National Taiwan University Hospital, Taipei, Taiwan, for technical and computing assistance.


        1. 1.Practice Bulletin No. 137: Gestational diabetes mellitus. Obstetrics and gynecology. 2013;122(2 Pt 1):406–16.
        2. 2.IDF Diabetes Atlas, 8th edn. Brussels, Belgium. International Diabetes Federation. 2017.
        3. 3.Osmundson SS, Zhao BS, Kunz L, Wang E, Popat R, Nimbal VC, et al. First Trimester Hemoglobin A1c Prediction of Gestational Diabetes. American journal of perinatology. 2016;33(10):977–82. pmid:27120479
        4. 4.13. Management of Diabetes in Pregnancy: Standards of Medical Care in Diabetes-2018. Diabetes care. 2018;41(Suppl 1):S137–S43. pmid:29222384
        5. 5.Doherty DA, Magann EF, Francis J, Morrison JC, Newnham JP. Pre-pregnancy body mass index and pregnancy outcomes. International journal of gynaecology and obstetrics: the official organ of the International Federation of Gynaecology and Obstetrics. 2006;95(3):242–7.
        6. 6.Torloni MR, Betran AP, Horta BL, Nakamura MU, Atallah AN, Moron AF, et al. Prepregnancy BMI and the risk of gestational diabetes: a systematic review of the literature with meta-analysis. Obesity reviews: an official journal of the International Association for the Study of Obesity. 2009;10(2):194–203.
        7. 7.Aye IL, Lager S, Ramirez VI, Gaccioli F, Dudley DJ, Jansson T, et al. Increasing maternal body mass index is associated with systemic inflammation in the mother and the activation of distinct placental inflammatory pathways. Biology of reproduction. 2014;90(6):129. pmid:24759787
        8. 8.Pantham P, Aye IL, Powell TL. Inflammation in maternal obesity and gestational diabetes mellitus. Placenta. 2015;36(7):709–15. pmid:25972077
        9. 9.Shin JA, Lee JH, Lim SY, Ha HS, Kwon HS, Park YM, et al. Metabolic syndrome as a predictor of type 2 diabetes, and its clinical interpretations and usefulness. Journal of diabetes investigation. 2013;4(4):334–43. pmid:24843675
        10. 10.Chatzi L, Plana E, Pappas A, Alegkakis D, Karakosta P, Daraki V, et al. The metabolic syndrome in early pregnancy and risk of gestational diabetes mellitus. Diabetes & metabolism. 2009;35(6):490–4.
        11. 11.Lei Q, Niu J, Lv L, Duan D, Wen J, Lin X, et al. Clustering of metabolic risk factors and adverse pregnancy outcomes: a prospective cohort study. Diabetes/metabolism research and reviews. 2016;32(8):835–42. pmid:27037671
        12. 12.Garg A. Lipodystrophies. The American journal of medicine. 2000;108(2):143–52. pmid:11126308
        13. 13.Boden G. Obesity and free fatty acids. Endocrinology and metabolism clinics of North America. 2008;37(3):635–46, viii-ix. pmid:18775356
        14. 14.Samuel VT, Petersen KF, Shulman GI. Lipid-induced insulin resistance: unravelling the mechanism. Lancet (London, England). 2010;375(9733):2267–77.
        15. 15.Sadur CN, Eckel RH. Insulin stimulation of adipose tissue lipoprotein lipase. Use of the euglycemic clamp technique. The Journal of clinical investigation. 1982;69(5):1119–25. pmid:7040473
        16. 16.Au WS, Kung HF, Lin MC. Regulation of microsomal triglyceride transfer protein gene by insulin in HepG2 cells: roles of MAPKerk and MAPKp38. Diabetes. 2003;52(5):1073–80. pmid:12716735
        17. 17.Brunzell JD, Hokanson JE. Dyslipidemia of central obesity and insulin resistance. Diabetes care. 1999;22 Suppl 3:C10–3.
        18. 18.Hall JE, do Carmo JM, da Silva AA, Wang Z, Hall ME. Obesity-induced hypertension: interaction of neurohumoral and renal mechanisms. Circulation research. 2015;116(6):991–1006. pmid:25767285
        19. 19.Fong A, Serra AE, Gabby L, Wing DA, Berkowitz KM. Use of hemoglobin A1c as an early predictor of gestational diabetes mellitus. American journal of obstetrics and gynecology. 2014;211(6):641 e1–7.
        Anxiety reduction through art therapy in women. Exploring stress regulation and executive functioning as underlying neurocognitive mechanisms

        Anxiety reduction through art therapy in women. Exploring stress regulation and executive functioning as underlying neurocognitive mechanisms

        Anxiety reduction through art therapy in women. Exploring stress regulation and executive functioning as underlying neurocognitive mechanisms

        • Annemarie Abbing, 
        • Leo de Sonneville, 
        • Erik Baars, 
        • Daniëlle Bourne, 
        • Hanna Swaab




        To explore possible working mechanisms of anxiety reduction in women with anxiety disorders, treated with art therapy (AT).


        A RCT comparing AT versus waiting list (WL) condition on aspects of self-regulation. Stress regulation (heart rate and heart rate variability) and executive functioning (daily behavioural and cognitive performance aspects of executive functioning (EF)) were evaluated in a pre-post design. Participants were women, aged 18–65 years with moderate to severe anxiety symptoms.


        Effectiveness of AT compared to WL was demonstrated in a higher resting HRV post treatment, improvements in aspects of self-reported daily EF (emotion control, working memory, plan/organize and task monitor), but not in cognitive performance of EF, stress responsiveness and down regulation of stress. The decrease in anxiety level was associated with improvements in self-reported daily EF.


        AT improves resting HRV and aspects of EF, the latter was associated with art therapy-related anxiety reduction.


        Table 4
        Fig 1
        Table 1
        Fig 2
        Table 2
        Fig 3
        Table 3
        Table 4
        Fig 1
        Table 1
        Fig 2

        Citation: Abbing A, de Sonneville L, Baars E, Bourne D, Swaab H (2019) Anxiety reduction through art therapy in women. Exploring stress regulation and executive functioning as underlying neurocognitive mechanisms. PLoS ONE 14(12): e0225200.

        Editor: Jim van Os, Maastricht Universitair Medisch Centrum+, NETHERLANDS

        Received: July 3, 2019; Accepted: October 26, 2019; Published: December 3, 2019

        Copyright: © 2019 Abbing et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

        Data Availability: Abbing, Annemarie (2019), “Dataset RCT: art therapy for anxiety in women”, Mendeley Data, v2.

        Funding: This study was co-funded by the Iona Foundation (, Stichting AG Phoenix and the Dutch association of anthroposophic art therapy (NVKToag). These organisations had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

        Competing interests: The authors have declared that no competing interests exist.


        Every person experiences fear and anxiety in life to some degree, but for people with an anxiety disorder, the anxiety increases over time, is disproportionate to the actual danger or threat and becomes permanent [1].

        The Diagnostic and Statistical Manual of Mental Disorders (DSM-5) [1] distinguishes between different types of anxiety disorders. The most common anxiety disorders are phobias, followed by social anxiety disorder (SAD), generalized anxiety disorder (GAD) and panic disorder (PD) [2]. Although the anxiety disorders may have different triggers, they share underlying features [34]. An important feature that applies to all anxiety disorders is the exaggerated cognitive appraisal that is associated with the threatening situation: hyper-alert cognitive schemes lead to pathological anxiety [5].

        Anxiety consists of physiological, emotional and cognitive aspects, and arises from specific personal characteristics combined with genetic, neurobiological and social factors [6], including hypersensitivity to stress and the tendency to experience strong negative emotions (nervousness, sadness, anger). According to Clark and Watson [7], anxiety is characterized by negative affect and high physiological hyperarousal. Anxiety disorders are associated with dysfunctions in self-regulation [89], such as problems concerning emotion regulation, stress regulation (hyperarousal) and difficulties with cognitive regulation.

        Treatment of anxiety is often aiming at changing maladaptive beliefs through cognitive behavioural therapy [10,11] and/or reducing anxiety symptoms through medication. Art therapy (AT) is also often provided in anxiety disorders, although little is known about its effectiveness. AT is a non-verbal, so-called ‘experience-oriented’ intervention that uses the visual arts (e.g. painting, drawing, sculpting, clay modelling) and is provided as standalone therapy or in multidisciplinary treatment programs for anxiety disorders. Outcomes of a RCT comparing three months AT with three months waiting list condition in women with anxiety disorders showed that anxiety symptoms can be reduced by AT and that there are indications that improved perceived emotion regulation plays a role in this reduction [12]. However, other aspects of self-regulation may also play a role in the reduction of anxiety symptom severity. As discussed above, individuals suffering from anxiety have also problems concerning stress regulation and difficulties with cognitive regulation, expressed as executive functioning.

        Anxiety disorders and stress regulation

        Stress regulation concerns dealing with stressors. People with anxiety problems usually have stress responses that are typically accompanied by sweating, shaking, dizziness and increased heart rate [13]. These physical reactions are driven by the autonomic nervous system in the presence of a stressor. An indicator for the functioning of the autonomic nervous system is heart rate variability (HRV), which refers to the fluctuations in heart rate, also known as the variation in time between heartbeats (InterBeat Interval) [1415]. Parasympathetic influences from the brainstem alter the heart rate, and HRV is thus an index of cardiac flexibility [16]. HRV decreases with age and cardiac comorbidities and increases with physical activity [17]. During stress, HR increases and HRV decreases due to more sympathetic and less parasympathetic activation.

        High HRV, expressing more variation in time between heart beats, is seen as an indicator for a well-functioning autonomic nervous system, which can respond to varying demands in different situations [1518]. HRV at rest (resting HRV) reflects self-regulation ability [19].

        A review of 36 studies into the relationship between the presence of anxiety symptoms and HRV, in people with anxiety disorders compared to control groups without anxiety symptoms [20], reported that the presence of an anxiety disorder is significantly associated with lower HRV. This was the case for people with generalized anxiety disorder (GAD), social anxiety disorder (SAD) and panic disorder (PD): lower resting HRV is associated with PD, SAD [21] and GAD [921]. Lower HRV indicates autonomic inflexibility and deficits in anxiety related inhibitory processes [21]. Resting HRV is lower in individuals that worry more [22] and have high (trait) anxiety [23]. Decrease of HRV is shown in conditions of stress (e.g. time pressure), emotional strain and increased anxiety [24]. Based on these findings, heart rate and heart rate variability are considered physiological indicators of stress and anxiety [25]. An increase in heart rate may be an indication of an increased emotional state [26].

        Anxiety disorders and executive functioning

        A risk factor in the development and persistence of anxiety disorders is a limitation in executive skills [27]. Executive functions (EFs) are cognitive processes that are necessary for efficient and goal-oriented behaviour. Important EF components are inhibition (the ability to stop and / or slow down behaviour (actions and thoughts)), working memory (the collection of cognitive processes that keep information temporarily accessible in order to perform mental tasks), cognitive flexibility (changing and adjusting behaviour) and planning (being able to think ahead and subdivide the process into intermediate steps towards a goal) [282930].

        Although the body of knowledge on the relationship between anxiety and EF is small, it is assumed that anxiety disorders may be the result of suboptimal cognitive regulation processes involving executive functioning. Difficulties with inhibition are found to be associated with higher levels of anxiety symptoms. Problems with EF are a risk factor for developing social anxiety [31] and adults with a generalized anxiety disorder are found to have more difficulties with inhibition than healthy controls [32].

        Present study

        There is already some evidence for anxiety symptom reduction through AT and there are indications that improved perceived emotion regulation plays a role in this reduction [12]. To gain further understanding of other regulating processes that might play a role in the reduction of anxiety symptoms through AT, the question is addressed whether stress regulation and executive functioning improve as a result of AT and whether these mechanisms are related to the reduction of anxiety symptoms and may provide (further) mechanistic evidence and explanations of the AT treatment effect.

        Some studies suggest a possible stress regulating effect of AT, since AT is thought to promote relaxation [3334]. AT is also believed to improve several aspects of executive functioning, like inhibition, because it is supposed to contribute to decrease of (a.o.) impulsivity [3536]. Based on these studies and expert opinions we hypothesized that AT treatment contributes to better stress regulation, because AT is thought to induce relaxation, presumably comparable to mindfulness [3738] and may thus have a dampening effect on the arousal. This could become visible in a lower stress response, improved down regulation after facing a stressor and as a generally lower stress level which is reflected in a higher HRV at rest.

        AT may also improve cognitive regulation, which can be reflected in improvement of several aspects of executive functioning, like inhibition, sustained attention, flexibility, working memory and task monitor, because these competences are needed to perform artistic exercises and are thought to be practiced and trained during the therapy process.

        Materials and methods

        In a RCT design, comparing three months AT with three months waiting list condition in women with anxiety disorders, psychophysiological outcomes (stress responsivity) as well as daily behavioural and cognitive performance aspects of executive functioning were measured. Data were collected as part of a single-blind RCT on the effectiveness of art therapy in women with anxiety disorders. Detailed information about the study is reported elsewhere [12]. The study was approved by the Medical Ethics Committee of the Leiden University Medical Centre, the Netherlands (NL36861.018.11) and the trial was registered in the Dutch Trial Registration (

        Sample randomisation and intervention

        Included were adult women, aged 18–65, with a primary diagnosis of generalized anxiety disorder, social phobia and/ or panic disorder (with or without agoraphobia) and with moderate to severe anxiety symptoms (scoring >7 for anxiety and/or >10 for distress on the Four Dimension Symptoms Questionnaire (4SDQ) [39]. Patients were excluded if they were aged less than 18 years or older than 65 years, had psychosis or hallucinations, alcohol or drug addiction, suicidal risk, brain pathology. All participants were recruited through posters/flyers in the practices of family doctors and by social media.

        Sample size

        Based on a pre-post measurement difference in the primary outcome of 15% (considered to be a clinically relevant LWASQ total score reduction), an alpha of 0.05, a power of 0.80 and a dropout rate of 15%, the sample size was calculated: 60 participants in total (30 participants per group) (

        Randomisation method and allocation concealment

        A pre-stratification procedure was executed with four strata: use of psychotropic drugs (yes/ no), moderate or severe depression symptoms (4SDQ: depression >6) (yes/ no). After pre-stratification, through block randomization (blocks of 2), participants were at random assigned to either the treatment group (AT) or the control group (WL), according to a computer-generated list ( Blinding of art therapists and participants was not possible.

        The study took place at 25 private art therapy practices spread throughout the Netherlands, in the period between January 2017 and March 2018.

        A total of 59 women was included between January and July 2017. After pre-stratification on comorbid depression symptom level and psychopharmaceutical use, randomization resulted in an experimental group of 30 participants and a control group of 29 participants.


        The experimental group (AT group) received 10–12 sessions AT of one hour each, during a period of three months. The specific intervention type was anthroposophic art therapy. The control group (WL) was wait listed for three months. In order to assure that the intervention tested in the study was representative for the general approach of this type of AT, only qualified and registered Dutch anthroposophic art therapists, with more than five years’ experience in working with adults with anxiety, treated the participants.

        Study population

        During the study, 12 patients (20%) dropped out and 47 patients (80%) completed the trial. Dropouts concerned six from AT group and six from WL group, due to lack of time (n = 3), not willing to wait for the intervention (n = 3), hospitalization or physical illness (n = 3), non-response (n = 2) and migration (n = 1). There were no significant differences between dropouts and completers on baseline parameters (T0), so per-protocol analysis was justified.

        The participants in the two groups did not differ on key variables, including age, diagnosis, use of medication, occupation, education and familiarity with anthroposophic healthcare and outcome variables at baseline. The analysed sample of 47 patients had a mean age of 44.4 years (SD = 14,0), moderate to severe anxiety symptoms: 11.2 (SD = 4.6) and a mean duration of anxiety symptoms of 17.6 years (SD = 18.9) (range: three months—64 years (lifetime)). Medication for anxiety was used by 15 participants. Multiple diagnoses applied to all participants: 25 participants met the criteria for the diagnosis GAD, 21 for social phobia and 28 for panic disorder. Ten participants suffered from (comorbid) PTSD, five participants had current comorbid depression and 16 patients experienced one or more depressive episodes prior to this study.


        The study contained two measurement waves: pre- and post-treatment, three months apart, which consisted of online questionnaires and home-visits with physiological and neuropsychological measurements. The online questionnaires were completed prior to the home-visit. The protocol during the two home-visits at T0 and T1 included the measurement of stress regulation and of performance-based executive functioning. The measures at the home-visits were taken by trained research assistants who were unaware of allocation. Outcome assessors who judged and analysed the results were blinded as well.

        Measures and instruments

        Measures of stress regulation

        Stress regulation was measured as stress responsiveness (response and recovery), with physiological responses using a Biopac MP150 Acquisition System (Biopac Systems Inc., Santa Barbara, CA) during a stress-evoking task (Fig 1), based on the Trier Social Stress Test (TSST) [40]. The TSST is developed to measure regulation of the autonomic nervous system (ANS) during stress.



        Fig 1. Content of the stress-evoking task.

        Heart rate (HR), and heart rate variability (HRV) were recorded responses. Three ECG electrodes were attached to the chest, one located near left mid-clavicular line directly below the clavicle, one near the right mid-clavicular line and one between 6th and 7th intercostal space on left mid-clavicular line. To stabilize the ECQ signal, a 2 Hz high pass filter and a 50 Hz notch filter were applied in AcqKnowledge software (version 0.3.0, Biopac System Inc). R-peaks and IBIs were visually inspected and manually corrected by two researchers (AA and DB). The corrected recordings were analysed using the PhysioData Toolbox, a MATLAB-based application [41]. For each phase (resting, stress and cool down) we calculated mean HR: mean of the continuous HR, as interpolated for the accepted Inter-Beat-Interval (IBI) data point, and HRVRMSSD: the square root of the mean squared differences between successive IBIs (nonadjacent IBIs disregarded). HRV is the variability in the distance between R peaks, RR-interval or Inter-Beat-Interval (IBI) and refers to beat-to-beat alterations in heart rate (HR). It is a measure of both sympathetic and parasympathetic influences on the heart (Levine et al., 2016) and is related to emotional arousal [24].

        The Root Mean Squared Successive Differences (RMSSD) of the HRV was used as this is the recommended measure for calculating high frequency HRV from recordings of several minutes since this measure is indicative for parasympathetic nervous system and is most commonly used and preferred to pNN50, as it has better statistical properties [42]. Normal RMSSD mean value in a healthy population is 42ms (range 19–74) [17].

        Because of the vulnerability of the study population, it was decided that the experiment would be stopped immediately if the subject indicated that she wanted to stop. Also, when research assistants observed or suspected a too anxious, confused or emotional state, the subject was asked if she wanted to stop the experiment, and if so, it was stopped.

        Measures of executive functioning

        We used two measures of executive functioning: behavioural EF and cognitive EF. The daily behavioural EF was measured with a self-report questionnaire (BRIEF-A) and cognitive aspects of EF were measured with performance-based measures (subtests of the Amsterdam Neuropsychological Tasks (ANT)).

        The Dutch version of the Behaviour Rating Inventory of Executive Function for Adults (BRIEF-A) was used to measure various aspects of daily executive functioning [43]. The BRIEF-A is a questionnaire developed for adults, and it consists of 75 items with nine clinical scales that measure various aspects of EF: four behavioural regulation scales and five metacognition scales. The behaviour regulation scales are: inhibit: ability to control impulses (inhibitory control) and to stop engaging in a behaviour; shift: cognitive flexibility, ability to move freely from one activity or situation to another; to tolerate change; to switch or alternate attention; emotional control: ability to regulate emotional responses appropriately; and self-monitor: ability to keep track of the effect of one’s own behaviour on other people. The metacognition scales are: initiate: ability to begin an activity and to independently generate ideas or problem-solving strategies; working memory: ability to hold information when completing a task, when encoding information, or when generating goals/plans in a sequential manner; plan/organize: ability to anticipate future events; to set goals; to develop steps; to grasp main ideas; to organize and understand the main points in written or verbal presentations; organization of materials: ability to put order in work, play and storage spaces (e.g. desks, lockers, backpacks, and bedrooms); and task monitor: ability to check work and to assess one’s own performance. T-scores were calculated from the raw scores. The ranges for the clinical scales are: <60 normal; 60–65 subclinical; >65 clinical [43].

        For measuring cognitive performance-based aspects of executive functioning, the Amsterdam Neuropsychological Tasks (ANT) a computer-aided assessment, was used. The ANT allows for the systematic evaluation of neuropsychological performance [44]. It has been proven to be a sensitive and valid tool in research on executive functions. Test–retest reliability and validity of the ANT-tasks are satisfactory and have been extensively described elsewhere (e.g. [45,46]).

        A test battery of three tasks was chosen for this study: Baseline Speed (BS), Shifting Attention Set Visual (SSV) and Sustained Attention Dots Patterns (SAD). These tasks cover the following neuropsychological domains: alertness (intensity of attention) (BS), inhibition / mental flexibility (SSV) and sustained attention (continuous performance) (SAD). These tasks are shortly described, for detailed information including illustrations, see [47].

        The BS task is a simple reaction time task, measuring of ‘intensity’ aspects of alertness and attention, as described by Konrad, Günther, Hanisch & Herpertz-Dahlmann [48]. On the screen a (fixation) cross is continuously projected. This cross changes unexpectedly into a square requiring the participant to press a mouse key as fast as possible, after which the square turns into a cross again, and this is repeated in 32 trials. Main outcome parameters are reaction time (RT), reflecting alertness, and the response speed stability (SD of RT), reflecting fluctuation in alertness.

        The SSV task aims at measuring inhibition and attentional flexibility. The signal consists of a horizontal bar that is permanently present on which a square jumps randomly to the left or the right. In part 1, participants have to copy the movement of a green-coloured square (press left/right button on a left/right move). In part 2, participant have to do the opposite, i.e. ‘mirror’ the movement of a red-coloured square, and in part 3, the square randomly changes colour, requiring participants to either copy or mirror the movement of the square. The contrast between performances in part 1 and part 2 reflects inhibitory control; the contrast between performances in part 1 and part 3 (compatible responses) reflects cognitive flexibility, with larger values indicating poorer performance. Main outcome parameters of the SSV task are reaction time and accuracy (percentage of errors).

        The SAD task measures sustained attention, i.e. the ability to keep performance at a certain level during a longer period of time. In this task, 600 dots patterns with 3, 4, or 5 dots appear on a computer screen in 50 series of 12 trials, each consisting of three 3-dots, 4-dots, and 5-dots patterns, presented in a pseudo random order. Participants are required to respond to 4-dots patterns by pressing a mouse key with their preferred hand (‘yes’-response) and to press the other mouse key with the non-preferred hand (‘no’-response) whenever 3- or 5-dots patterns are shown. Inaccurate responses, misses (‘no’-responses to 4-dots) and false alarms (‘yes’-responses to 3 or 5 dots) are directly followed by a beep signal. Task duration is approximately 15–20 min. Main outcome parameters are tempo (mean series completion time across 50 series), accuracy, and fluctuation in tempo. Fluctuation in tempo, the WS subject SD of 50 completion times, is taken as the primary index of sustained attention. As participants were informed about errors by a beep signal, and correct responses following an error are separately registered, post-error slowing (sensitivity to feedback) can be estimated.

        Measure of anxiety symptoms

        The Dutch version of the Lehrer Woolfolk Anxiety Symptom Questionnaire (LWASQ) [49] was used to measure the anxiety level. The LWASQ is a self-report, generic anxiety instrument with 36 questions which assesses the cognitive (worry and rumination), behavioural (avoidance) and somatic (physical symptoms) aspects of anxiety. In the present study, the difference between pre- and post-measurement was used for further analysis. The reliability of the LWASQ is sufficient (α = .83 tot .92) and the questionnaire is suitable for the measurement of treatment effects [50].

        Statistical analysis

        Statistical analyses were conducted using SPSS statistics (version 23.0) [51]. All data was checked for normal distribution using the Shapiro Wilk test, Q-Q plot and histogram.

        Reasons for missing values were reported. Dropouts were compared to completers using pre-test measures on age and anxiety score, by use of independent students t-tests. No significant differences were found, so missing cases were listwise deleted and per protocol (PP) analyses were performed for all outcomes.

        The following hypotheses were tested:

        1. AT results in higher resting HRV
        2. AT results in a lower stress response: lower increase in HR and lower decrease in HRV from resting to stress phase, respectively
        3. AT results in faster recovery (HR) during cooling down
        4. AT leads to improvements on self-reported daily EF
        5. AT leads to improvements on performance-based EF
        6. The reduction of anxiety symptoms is associated with improvements in stress recovery
        7. The reduction of anxiety symptoms is associated with improvements of EF

        Evaluation of treatment effects–stress regulation

        To determine if the stress paradigm did work, we evaluated the changes in HR and HRV from resting to stress and cool down phase, with all subjects of waitlist and treatment groups included. Expected were an increase in stress level at the start of the stress induction phase (shown as increase of HR and decrease of HRV) and a decrease during the cool down (shown as decrease of HR and increase of HRV). This was tested with a general linear model repeated measures analysis for variance (RM-ANOVA), with Test phase (resting, stress induction, cooling down) as within-subject (WS) factor and the pre-test HR and HRVRMSSD as dependent variable respectively, using a repeated contrast (resting vs. stress induction, stress induction vs. cooling down).

        To examine hypothesis 1, we tested whether the therapy had influenced resting HRV (resting phase) by using a RM- ANOVA with Test moment (pre- vs. post-test) as WS factor, Group (AT vs. WL) as BS factor and resting HRVRMSSD as dependent variable.

        Hypothesis 2 was tested using a RM-ANOVA with Test moment (pre- vs. post-test) as WS factor, and Group (AT vs. WL) as BS factor, with stress response (stress induction HR minus resting HR) as dependent variable. The same procedure was followed for HRVRMSSD (stress induction HRV minus resting HRV).

        To examine hypothesis 3, to test whether the experimental group improved on downregulation (recovery speed), we divided the Cooling down phase in nine slices of 30 seconds each and analysed changes in HR during using a RM-ANOVA with the Cooling down phases (slice 1–9) and Test moment (pre- vs. post-test) as WS factors, Group (AT vs. WL) as BS factor, and HR as dependent variable.

        Evaluation of treatment effects–executive functioning

        To examine hypothesis 4, a RM-ANOVA was performed, using Test moment (pre- vs. post-test) as WS factor and Group (AT vs. WL) as BS factor, with BRIEF-A subscale (T) scores as dependent variables, respectively.

        To establish whether the study population deviated from the norm on performance-based EF measured with the ANT, we performed a MANOVA, with the z-scores of alertness (speed, fluctuation in speed), inhibition and flexibility (reaction time, error percentage) and sustained attention (tempo, fluctuation in tempo, error percentage) as dependent variables, using the intercept test for deviations from zero.

        Subsequently, the treatment effects were evaluated (hypothesis 5) by means of RM-ANOVAs with Group (AT vs. WL) as BS factor and Test moment (pre- vs. post-test) as WS factor, and the outcomes on alertness (reaction time, fluctuation in reaction time) and sustained attention (tempo, fluctuation in tempo, error percentage) as dependent variables respectively. To analyse treatment-related changes in inhibition and flexibility, differences scores were computed for Inhibitory control and Cognitive flexibility, for both reaction time and error percentages. These dependent variables were analysed with RM-ANOVAs, with Group (AT vs. WL) as BS factor and Test moment (pre- vs. post-test) as WS factor.

        For all analyses, a p-value of 0.05 was considered statistically significant. The effect size Partial Eta Squared (ηp2) was calculated to assess the relevance of the effect. An effect size of 0.01–0.06 is considered a small effect, 0.06–0.14 a medium effect, and >0.14 a large effect in RM analysis [52].

        Exploration of associating factors

        The sample size was calculated based on our primary aim to study the effectiveness of art therapy. The secondary aim was to explore factors influencing anxiety reduction (testing hypotheses 6 and 7). For this purpose, correlations were computed between the reduction in anxiety score and the pre-post treatment differences in EF, and aspects of stress regulation (HR and HRV). Only those aspects of EF and stress regulation that were demonstrated to improve significantly after treatment within the experimental group were entered in the correlation analysis. Only variables that significantly correlated with anxiety reduction were subsequently entered in regression analysis. Hierarchical regression analyses were planned to examine whether improvements of EF and stress recovery contributed to anxiety symptom reduction. To examine baseline predictors of anxiety symptom reduction, regression analyses were planned with the primary outcome variable (pre-post treatment difference in anxiety symptom severity) compared to baseline variables (pre-treatment values).


        Anxiety symptom severity

        In a previous paper [12] that reported on this RCT, was shown that anxiety symptom severity was significantly reduced in the AT group but not in the WL group [F(1,45) = 11.49, p = 0.001, ηp2 = 0.20].

        The Within-Group outcomes are presented in Table 1.



        Table 1. Anxiety symptom severity: Within-Group outcomes [12].

        Mean, standard deviation, 95% CIs and p-values from pre- to post-treatment (paired t-tests).

        Stress regulation

        A number of measurements of participants (n = 11) could not be used due to a very distorted signal (n = 6) or uncompleted tests due to refusal or impossibility of participants to complete the paradigm (n = 5). To investigate whether there were outliers, the 1.5 x interquartile distance rule (IKA) was used. Analyses are carried out both with and without outliers. Because this did not yield different results, the outcomes of the analyses with outliers are reported.

        Evaluation of stress paradigm

        The first WS contrast (resting vs. stress induction) revealed a significant effect for HR [F(1,51) = 158.72, p<0.0001, ηp2 = .757] and HRV [F(1,51) = 5.666, p = 0.021, ηp2 = .100], respectively, with HR increasing from [mean (SD)] 69,59 (9,58) to 90,28 (15,96), and HRV decreasing from [mean(SD)] 40,02 (25,29) to 29,25 (25,25). The second contrast (stress induction vs. cool down) showed a significant effect for HR [F(1,51) = 166.47, p<0.0001, ηp2 = .765] and HRV [F(1,51) = 6.342, p = 0.015, ηp2 = .111], with HR decreasing from [mean(SD)] 90,28 (15,96) to 70,14 (9,83) and HRV increasing from 29,25 (25,25) to 42,27 (29,49). This confirms that the stress paradigm worked: an increase of HR is shown in the stress induction phase and a decrease is shown in the cool down phase as expected. The HRV decreases during stress induction and recovers during cool down.

        Treatment effects on HRV

        For HRVRMSSD, the interaction effect Test moment*Group was trend significant [F(1,35) = 3.96, p = 0.054, ηp2 = .102], indicating that the AT group improved more than the WL group. This was further explored within the three phases. During resting phase, the interaction Test moment*Group was significant [F(1,35) = 4.54, p = 0.04, ηp2 = .115], reflecting that the AT group had a higher HRV during the resting phase at post treatment and the WL group had a lower HRV at T1, indicating improved HRV in AT group only (Table 2 and Fig 2).



        Fig 2. Mean HRV.



        Table 2. Outcomes HRV and HR (stress regulation).

        Mean, standard deviation at pre- and post-treatment (RM-ANOVA).

        For the stress induction phase, no significant Test moment*Group interaction was shown (p = 0.81) and in the cool down phase, the Test moment * Group interaction on HRV was trend significant (p = 0.068): the WL group appeared to have a lower HRV at T1, compared to T0, and the AT showed an increase of HRV at T1, indicating an improvement (Table 2 and Fig 3).



        Fig 3. Mean HR.

        Treatment effects on stress responsivity (HR)

        The RM-ANOVAs testing the treatment effects of AT on HR showed a significant main effect for Test moment, indicating an increase in HR from pre- to post treatment [F(1,35) = 11.46, p = 0.002, ηp2 = 0.247], but no significant Test moment * Group interaction (p = .649) (Fig 3).

        Stress response was calculated as the difference in mean HR between stress induction phase and resting phase. No significant differences in stress response were found between groups (p = 0.444), indicating that there were no improvements in stress responsivity.

        Treatment effect on stress recovery (HR)

        Stress recovery was calculated as the difference in mean HR between stress induction phase and cool down phase. No significant differences in stress recovery were found between groups (p = 0.374), indicating that there were no improvements in stress recovery.

        To test stress recovery speed, the cool down phase was analysed in slices of 30 seconds. No differences between the groups were observed on HR, indicating that both groups did not differ in stress recovery speed.

        Exploration of factors contributing to anxiety symptom reduction

        The only aspect of stress regulation that improved (resting HRV) did not significantly correlate to anxiety symptom reduction, so mediators and predictors were not analysed.

        Executive functioning–behavioural aspects

        On the behavioural aspects of EF (BRIEF-A total score), the interaction effect Test moment*Group was significant: F(1,44) = 827, p = 0.006, with a large effect size (ηp2 = .16), showing that the total EF improved in the AT group but not in the WL group. Four of the nine subscales of the BRIEF showed significant interactions on Test moment*Group: emotion control [F(1,44) = 4.26, p = 0.045, ηp2 = .09]; working memory [F(1,44) = 5.49, p = 0.024, ηp2 = .11]; plan/organize [F(1,44) = 5.87, p = 0.020, ηp2 = .12] and task monitor [F(1,44) = 10.79, p = 0.002, ηp2 = .20], indicating that AT was effective in these domains. AT was not effective in the domains inhibit (p = 0.13), shift (p = 0.24), self-monitor (p = 0.94), initiate (p = 0.66) and organization of materials (p = 0.56) (Table 3).



        Table 3. Outcomes BRIEF-A (executive functioning).

        Mean, standard deviation, p-values and effect sizes from pre- to post-treatment (RM-ANOVA).

        Exploration of contributing factors

        Only the variables that improved significantly in the AT group (analysed with a MANOVA intercept test) were added in a regression analysis with the LWASQ difference score (pre-post treatment). These variables were shift (cognitive flexibility), emotion controlplan/organizeworking memory and task monitor. A backward regression analysis with these variables resulted in a significant model [F(3,22) = 13,09, p<0.0001, R2 = .674], consisting of three subscales of the BRIEF, indicating that improvements in emotion control (β = .364, t = 2.54, p = 0.020), plan/organize (β = .406, t = 2.76, p = 0.012), and task monitor, (β = .319, t = 2.21, p = 0.039), explained 67,4% of the variance in anxiety symptom reduction.


        To explore possible predictors of therapy success, baseline scores of the subscales were used in a backward regression analysis in relation to anxiety symptom reduction. This resulted in a significant model [F(2,20) = 4,15, p = 0.031, R2 = .293] consisting of two subscales of the BRIEF, showing that higher baseline scores of shift (β = .381, t = 2.02, p = 0.057) and organization of materials (β = .347, t = 1.84, p = 0.081) led to larger reduction of anxiety symptoms, suggesting that subjects who experience many problems with these EF aspects are more likely to benefit from AT.

        Executive functioning–cognitive aspects

        Intercept tests on baseline z-scores showed that some of the EF variables deviated significantly from the mean norm score, but were within normal range (-1 to 1): fluctuation in tempo: [mean(SD)] 0.37 (1,16); F(1,44) = 4,61, p = 0.037, ηp2 = .095; accuracy (mistakes) in SSV1: [mean(SD)] 0.63 (1.34); F(1,44) = 10,35, p = 0.002, ηp2 = .19; accuracy of inhibition: [mean(SD)] 0,80 (2,26); F(1,44) = 5,68, p = 0.022, ηp2 = .114; accuracy of cognitive flexibility: [mean(SD)] 0,68 (2,12); F(1,44) = 4,64, p = 0.037, ηp2 = .095.

        The variable impulsivity falls within clinical range: [mean(SD)] 1,65 (1,65); F(1,44) = 57,07, p<0.0001, ηp2 = .52. This indicates a clinical problem (poor inhibition) in this study population.

        Treatment effects

        Due to procedural errors, three cases had to be excluded from analysis, two from the AT group and one from the WL group. The RM-ANOVAs testing the treatment effects of AT on inhibition, cognitive flexibility and sustained attention showed no significant differences between experimental group and control group (0.15<p<0.91). Some on the tasks showed significant outcomes of test moment only, indicating a learning effect. This applied to inhibition (speed), flexibility (speed) and sustained attention (speed and stability), but not for number of errors on the tasks and stability in speed in the inhibition and flexibility tasks. Outcomes of the tasks (mean(SD)) are presented in Table 4.



        Table 4. Outcomes of ANT tasks BS, SSV, SAD (mean(SD)).

        Mean, standard deviation, p-values and effect sizes from pre- to post-treatment (RM-ANOVA).

        Exploration of contributing factors

        Since no significant treatment effects on cognitive performance EF were observed, associations between performance EF and anxiety symptom reduction were not analysed.


        Performance-based inhibition scores at baseline did correlate to anxiety symptom reduction (r = -.416; p = 0.043), indicating that that subjects with poorer inhibition showed a larger reduction of anxiety symptoms, suggesting these subjects are more likely to benefit from AT.


        In this explorative study, the effects of AT on stress responsivity and executive functioning were assessed in order to further study the effectiveness of AT in adult women with anxiety and to explore possible working mechanisms of this treatment. Data were collected as part of a single-blind RCT on the effectiveness of AT in women with anxiety disorders, comparing an experimental AT treatment group and a waitlist control group [12].

        Our first hypothesis that AT would contribute to better stress regulation, is partially supported. Subjects in the intervention group showed higher resting HRV after treatment, indicating a lower stress level and/or reduction of anxiety, meeting our expectation. The stress response measured after treatment was however as strong as before the treatment and no improvements in stress recovery were observed, contrary to our expectations.

        Our second hypothesis that AT would result in executive functioning, was also partially supported. The results of the self-reported EF show that there were significant improvements in emotion controlworking memoryplan/organize and task monitor, but the changes in AT group in the domains inhibitshiftself-monitorinitiate and organization of materials were not significant compared to WL group. Regarding performance-based cognitive EF, there were no significant post treatment differences between the experimental group and the control group on Inhibition, Cognitive flexibility and Sustained attention.

        The third hypothesis, that improvements in stress regulation and EF were associated with anxiety symptom reduction, was only partly substantiated. Improvements in the self-reported EF domains emotion controlplan/organize and task monitor were associated with anxiety symptom reduction, with an explained variance of 67,4%. Analysis of predicting factors demonstrated that lower Inhibition scores on performance EF at T0 were associated with larger reduction of anxiety symptoms, and lower self-reported cognitive flexibility and organization of materials were associated with a larger anxiety reduction.

        Interpretation and comparison to literature

        The finding that improvement of resting HRV was shown in the experimental group, indicates an improved autonomic regulating ability [1853] and, according to the Neurovisceral Integration Model, an improved ANS regulation [54]. The higher resting HRV in the experimental group may be indicative for a lower overall stress level and can be considered an index for improved self-regulatory ability [55]. Because HRV is strongly associated with the presence of an anxiety disorder [921], and HRV is positively correlated with adaptive emotion regulation, according to the Polyvagal Theory [16], the outcomes of this study substantiate our earlier finding: anxiety symptom reduction and improvement of emotion regulation [12]. Furthermore, there is neurophysiological evidence for associations between resting HRV and executive brain regions [54]. Resting HRV does not only represent overall health, but is also an index for the degree of brain flexibility concerning self-regulation processes, such as executive functions and cognitive control [5657].

        Subjects in the experimental group showed the same stress response as before treatment and did not improve on stress recovery (down regulation). The fact that the stress response after treatment did not differ from before treatment, can have several explanations. Firstly, it is possible that subjects in the intervention group were just as sensitive/susceptible to stress induction as before treatment, leading to the preliminary conclusion that AT does not affect the direct stress response (stress responsiveness). In other studies it was shown that the stress response did not differ between healthy populations, people with intense worry and patients with GAD [58]. This implies that the stress response itself cannot easily be influenced. Secondly, the treatment period (three months) might have been to short and the number of sessions (10–12) too little to realize significant changes in stress responsiveness. Thirdly, the Trier Social Stress task is originally developed to induce stress in healthy populations. A worry task [9] may also be suitable for this study population and may lead to other outcomes.

        Another important outcome is that the treatment group experienced improvements in daily behavioural executive functioning in the domains emotion control, working memory, plan/organize and task monitor, but did not show pre-post treatment differences regarding performance-based executive functioning (Alertness, Inhibition, Cognitive flexibility and Sustained attention) compared to the control group. It is known that self-report measures are prone to a higher risk of bias / overestimation, due to positive expectations of the treated participants and to placebo effects, which are thought to account for 15% of treatment effects [59]. Positive expectations generally lead to a more positive self-evaluation of mental health [60].

        A possible explanation for not finding improvements in performance EF is that the study population was not in the clinical range on this aspect (except for Inhibition), meaning that there were no major problems with EF, thus making occurrence of improvement less likely. On the other hand, the small sample size may also have compromised the outcomes. It is not unlikely that significant improvements in accuracy (error percentages) of Inhibition and Cognitive flexibility can be found in a larger study population, because these variables improved in AT group and not in WL group.

        The study population showed poorer Inhibition skills compared to a healthy study population, and subjects with larger Inhibition problems showed a larger anxiety reduction. This is consistent with several studies that showed that poor behavioural inhibition is associated with anxiety and high physiological arousal [61].

        Hypotheses on working mechanisms of AT

        Although there is still much unclear about the exact working mechanisms of AT, the forgoing results allow for the hypothesis that AT is effective in the treatment of anxiety symptoms due to the improvement of specific aspects of self-regulation. In our study, AT led to improvement in overall stress reduction (higher resting HRV), and the treated subjects reported improvements in several aspects of executive functioning; emotion controlcognitive flexibility (shift)working memoryplan/organize and task monitor.

        This hypothesis can be substantiated by the body of knowledge on anxiety reduction, which states that both higher resting HRV and improvements in EF contribute to lower anxiety levels. These improvements may be caused by the therapy. Firstly, creating visual art is linked to improved psychological resilience (i.e. stress resistance) on a neural level, due to improvements in functional connectivity of the medial parietal cortex and the praecuneus [62]. Secondly, specific skills are practiced during the artistic exercises, which are carefully chosen by the therapist. These exercises provide experiences within a safe environment and are not only intended to gain insight in emotions and responses, but are also intended to practice skills [63]. These skills may be related to aspects of executive functioning: e.g. following instructions (working memory), working autonomously on an assignment (plan/organize), tracking and evaluating own actions during the art work (task monitor), learn to interact with and adjust to the qualities of different art materials (shift), and learn to explore and regulate their emotions. This hypothetical working mechanism is substantiated by the finding that improvements of the aspects emotion control, plan/organize and task monitor contributed for 64,7% of the anxiety symptom reduction through AT.

        Strengths, limitations and generalizability

        Strengths of this study are that this is the first RCT on AT for anxiety studying clinical outcomes and working mechanisms, with both self-reported and objective measurements and analyses of outcomes and influencing factors. This study is important for the scientific underpinning of AT in general and for the AT treatment of anxiety disorders specifically. Based on this study, specific hypotheses on working mechanisms can be formulated that can be tested in further research.

        Although this study provides important contributions to the sparse AT literature, it does have some limitations. A first limitation is the lack of an active control group. It is therefore not possible to conclude with certainty that the observed effects are caused by therapy-specific factors. Second, the relatively small sample size may have compromised the outcomes and may have led to non-detection of significant outcomes of stress responsivity and performance EF, or to non-detection of associations between improvement of resting HRV and anxiety reduction. Also, the outcomes of the regression analyses should be handled with caution since the small sample (n = 22) that was used for these analyses compromises the generalizability of these results.

        Third, cofounding and effect modifying factors such as cardiac comorbidities were not measured, which could have influenced the HRV outcomes. Fourth, the study population consisted of a specific subgroup: higher educated women, with a long duration of anxiety symptoms and probably with a specific motivation for the treatment, because the participants applied for the study. This implies that the results are not generalizable to all women with anxiety, nor to men [12].

        Fifth, the study population was heterogeneous in nature, due to dimensional inclusion on anxiety symptom severity. Participants did not belong to one specific anxiety disorder classification, so conclusions on the effectiveness of art therapy for specific anxiety disorders cannot be drawn. Sixth, this study did not result in a clear substantiation of the working mechanism of art therapy, but led to preliminary hypotheses that need to be tested in further studies. Because the present study was explorative in nature and had a small sample size, straightforward analyses have been performed in order to detect promising areas for further study. Some insight into possible working mechanisms has been gained, but still many factors need to be considered before concluding on the exact working mechanism(s) of art therapy.

        Future perspectives

        Outcomes of this study show that AT is a promising intervention for anxiety disorders, but studies with active controls are needed to prove efficacy and cost-effectiveness of AT. Recommendations are the testing of specific hypotheses in larger samples, testing with other objective measures and/or a different psychophysiological protocol.

        Hypotheses on working mechanisms generated from this study, regarding reduction of anxiety symptom severity through improvements in self-reported daily executive functioning, should be evaluated in future studies with larger samples, and should also include mediation and moderation analyses, and should control for confounding and effect modification. Associations between outcomes on emotion regulation and executive functioning, and emotion regulation and resting HRV, need to be explored as well in order to obtain a better understanding of the route along which art therapy reduces anxiety symptom severity.

        Because this explorative study contained a limited set of outcome measures, other hypotheses on working mechanisms are possible and should be studied as well as well. For instance, social regulation can be taken into account in future studies, as well as studying the effects of AT on for example work and use of pharmacotherapy for anxiety.

        Other hypotheses need to be formulated more firmly and narrower, before new RCTs are designed. Hypotheses can arise in general from expert experience, practice knowledge and scientific literature. Since scientific literature on AT is in its infancy, expert experiences and practice knowledge should be used to provide input for specific clinically relevant hypotheses. Case studies are suitable for this aim [64], because they can provide basic information on therapeutic processes, are aimed at explicating expert experience and give insight in the approach that is used in this intervention ( Case studies on AT may provide insight in the route along the improvements of executive functioning and stress level are accomplished, which may further substantiate the preliminary working mechanisms found in this study. In addition, case studies can be used to describe how positively tested interventions in RCTs can be tailored to the needs of the individual patient.


        The authors thank all participants and therapists who took part in this study.


        1. 1.American Psychiatric Association (2013). Diagnostic and Statistical Manual of Mental Disorders (Fifth ed.). Arlington, VA: American Psychiatric Publishing. P. 189. ISBN 978-0-89042-555-8.
        2. 2.Anxiety and Depression Association of America (ADAA, 2018). About AADA. Facts and Statistics. Accessed September 16, 2018 from
        3. 3.Cisler J., Olatunji B., Feldner M., & Forsyth J. (2010). Emotion regulation and the anxiety disorders: An integrative review. Journal of Psychopathology and Behavioural Assessment, 32(1), 68–82. pmid:20622981
        4. 4.Rosellini A., Boettcher H., Brown T., & Barlow D. (2015). A Transdiagnostic Temperament-Phenotype Profile Approach to Emotional Disorder Classification: An Update. Psychopathology Review, A2(1), 110–128.
        5. 5.Beck A., & Haigh E. (2014). Advances in Cognitive Theory and Therapy: The Generic Cognitive Model. Annual Review of Clinical Psychology, 10(1), 1–24.
        6. 6.Hassink-Franke L.J.A., Terluin B., Van Heest F.B., Hekman J., Van Marwijk H.W.J., & Van Avendonk M.J.P. (2012). NHG-Standaard Angst (tweede herziening). Huisarts Wet, 55(2), 68–77.
        7. 7.Clark L.A & Watson D. (1991). Tripartite model of anxiety and depression: psychometric evidence and taxonomic implications. J Abnorm Psychol. 1991;100:316–336. pmid:1918611
        8. 8.Mennin D., Heimberg R., Fresco D., & Ritter M. (2008). Is generalized anxiety disorder an anxiety or mood disorder? Considering multiple factors as we ponder the fate of GAD. Depression and Anxiety, 25(4), 289–299. pmid:18412056
        9. 9.Levine J. C., Fleming R., Piedmont J. I., Cain S. M., & Chen W. J. (2016). Heart rate variability and generalized anxiety disorder during laboratory-induced worry and aversive imagery. Journal of affective disorders, 205, 207–215. pmid:27449553
        10. 10.Hofmann S., & Smits J. (2008). Cognitive-behavioural therapy for adult anxiety disorders: A meta-analysis of randomized placebo-controlled trials. The Journal of Clinical Psychiatry, 69(4), 621–32 pmid:18363421
        11. 11.Smits J. A., Julian K., Rosenfield D., & Powers M. B. (2012). Threat reappraisal as a mediator of symptom change in cognitive-behavioural treatment of anxiety disorders: A systematic review. Journal of Consulting and Clinical Psychology, 80(4), 624. pmid:22686124
        12. 12.Abbing A.C., Baars E.W., De Sonneville L., Ponstein A.S., & Swaab H. (2019). Effectiveness of art therapy for anxiety in adult women: a randomized controlled trial. Frontiers in Psychology.
        13. 13.Kazdin A. (2000). Psychotherapy for children and adolescents. Directions for research and practice (Second ed.). New York: Oxford University Press.
        14. 14.Akselrod S., Gordon D., Ubel A., Shannon D., Barger A., & Cohen R. (1981). Power spectrum analysis of heart rate fluctuation: A quantitative probe of beat-to-beat cardiovascular control. Science, 213, 220. pmid:6166045
        15. 15.Cohen H., Benjamin J., Geva A. B., Matar M. A., Kaplan Z., & Kotler M. (2000). Autonomic dysregulation in panic disorder and in post-traumatic stress disorder: application of power spectrum analysis of heart rate variability at rest and in response to recollection of trauma or panic attacks. Psychiatry research, 96(1), 1–13 pmid:10980322
        16. 16.Porges S. W. (2007). The Polyvagal perspective. Biological Psychology, 74, 116–143. pmid:17049418
        17. 17.Nunan D., Sandercock G., & Brodie D. (2010). A Quantitative Systematic Review of Normal Values for Short-Term Heart Rate Variability in Healthy Adults. Pace-Pacing And Clinical Electrophysiology, 33(11), 1407–1417.
        18. 18.Appelhans B., Luecken L., & Leucken B. (2006). Attentional processes, anxiety, and the regulation of cortisol reactivity (vol 19, pg 81, 2006). Anxiety Stress And Coping, 19(2), 93.
        19. 19.Segerstrom S. C., Boggero I. A., & Evans D. R. (2016). Pause and plan; the physiology of self-regulation. Handbook of self-regulation: Research, theory, and applications, 131–145.
        20. 20.Chalmers J. A., Quintana D. S., Abbott M. J., & Kemp A. H. (2014). Anxiety disorders are associated with reduced heart rate variability: a meta-analysis. Frontiers in psychiatry, 5, 80. pmid:25071612
        21. 21.Pittig A., Arch J., Lam C., & Craske M. (2013). Heart rate and heart rate variability in panic, social anxiety, obsessive–compulsive, and generalized anxiety disorders at baseline and in response to relaxation and hyperventilation. International Journal of Psychophysiology, 87(1), 19–27. pmid:23107994
        22. 22.Brosschot J.F., Van Dijk E. & Thayer J.F. (2007). Daily worry is related to low heart rate variability during waking and the subsequent nocturnal sleep period. International Journal of Psychophysiology. 63 (1): 39–47. pmid:17020787
        23. 23.Miu A. C., Heilman R. M., & Miclea M. (2009). Reduced heart rate variability and vagal tone in anxiety: trait versus state, and the effects of autogenic training. Autonomic Neuroscience, 145(1–2), 99–103. pmid:19059813
        24. 24.Nickel P., & Nachreiner F. (2003). Sensitivity and Diagnostics of the 0.1-Hz Component of Heart Rate Variability as an Indicator of Mental Workload. Human Factors. 45 (4): 575–590. pmid:15055455
        25. 25.Huijbregts S., Van Berkel S., Swaab-Barneveld H., & Van Goozen S. (2011). Neurobiological and behavioural stress reactivity in children prenatally exposed to tobacco. Psychoneuroendocrinology, 36(6), 913–918. pmid:21255938
        26. 26.Cacioppo J., Tassinary L., & Berntson G. (2007). Handbook of psychophysiology (3rd ed.). Cambridge [etc.]: Cambridge University Press.
        27. 27.Sharp P.B., Miller G.A. & Heller W. (2015). Transdiagnostic dimensions of anxiety: neural mechanisms, executive functions and new directions. Int. J. Psychophysiology 98: 365–377.
        28. 28.Miyake A., & Friedman N. (2012). The Nature and Organization of Individual Differences in Executive Functions: Four General Conclusions. Current Directions in Psychological Science, 21(1), 8–14 pmid:22773897
        29. 29.Lezak M. (2012). Neuropsychological assessment (5th ed.). New York [etc.]: Oxford University Press.
        30. 30.Huizinga M., Burack J. A., & Van der Molen M. W. (2010). Age-related change in shifting attention between global and local levels of hierarchical stimuli. Journal of Cognition and Development, 11(4), 408–436.
        31. 31.Gladstone G. L., Parker G. B., Mitchell P. B., Wilhelm K. A., & Malhi G. S. (2005). Relationship between self-reported childhood behavioural inhibition and lifetime anxiety disorders in a clinical sample. Depression and anxiety, 22, 103–113. pmid:16149043
        32. 32.Hallion L. S., Tolin D. F., Assaf M., Goethe J., & Diefenbach G. J. (2017). Cognitive control in generalized anxiety disorder: relation of inhibition impairments to worry and anxiety severity. Cognitive Therapy and Research, 41(4), 610–618.
        33. 33.Kaimal G., Ray K., & Muniz J. (2016). Reduction of cortisol levels and participants’ responses following art making. Art therapy, 33(2), 74–80. pmid:27695158
        34. 34.Sandmire D. A., Gorham S. R., Rankin N. E., & Grimm D. R. (2012). The influence of art making on anxiety: A pilot study. Art Therapy, 29(2), 68–73.
        35. 35.Haeyen S., van Hooren S., van der Veld W., & Hutschemaekers G. (2018a). Efficacy of art therapy in individuals with personality disorders cluster B/C: A randomized controlled trial. Journal of personality disorders, 32(4), 527–542.
        36. 36.Haeyen S., van Hooren S., van der Veld W. M., & Hutschemaekers G. (2018b). Promoting mental health versus reducing mental illness in art therapy with patients with personality disorders: a quantitative study. The Arts in Psychotherapy, 58, 11–16.
        37. 37.Brown KW, Ryan RM. The benefits of being present: Mindfulness and its role in psychological well-being. Journal of Personality and Social Psychology. 2003;84(4):822–48. pmid:12703651
        38. 38.Egger N, Konnopka A, Beutel ME, Herpertz S, Hiller W, Hoyer J, et al. Short-term cost-effectiveness of psychodynamic therapy and cognitive-behavioral therapy in social anxiety disorder: Results from the SOPHO-NET trial. Journal of Affective Disorders. 2015;180:21 pmid:25879721
        39. 39.Terluin B. (1996). De vierdimensionale klachtenlijst (4DKL). Een vragenlijst voor het meten van distress, depressie, angst en somatisatie [The Four-Dimensional Symptom Questionnaire (4DSQ). A questionnaire to measure distress, depression, anxiety, and somatization] Huisarts Wet, 39, 538–547.
        40. 40.Kirschbaum C., Pirke K-M., & Hellhammer D.H. (1993). The “Trier Social Stress Test”–A Tool for Investigating Psychobiological Stress Responses in a Laboratory Setting. Neuropsychobiology. 1993;28(1–2):76–81. pmid:8255414
        41. 41.Sjak-Shie E. (2017). PhysioData Toolbox. Version 0.2.Leiden University, Faculty of Social and Behavioural Sciences, The Netherlands.
        42. 42.Camm A. J., Malik M., Bigger J. T., Breithardt G., Cerutti S., Cohen R. J. et al. (1996). Heart rate variability: standards of measurement, physiological interpretation and clinical use. Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Circulation, 93(5), 1043–1065 pmid:8598068
        43. 43.Roth R. M., Isquith P. K., & Gioia G. A. (2005). BRIEF-A: Behavior rating inventory of executive function—adult version: Professional manual. Psychological Assessment Resources.
        44. 44.De Sonneville L. M. J. (1999). Amsterdam Neuropsychological Tasks: A computer-aided assessment. Cognitive ergonomics, clinical assessment and computer-assisted learning, 6, 187.
        45. 45.De Sonneville L. M. J. (2014). Handbook Amsterdam Neuropsychological Tasks. Boom Testuitgevers, Amsterdam, The Netherlands.
        46. 46.Rowbotham I., Pit-ten Cate I. M., Sonuga-Barke E. J. S., & Huijbregts S. C. J. (2009). Cognitive control in adolescents with neurofibromatosis type 1. Neuropsychology, 23(1), 50. pmid:19210032
        47. 47.De Sonneville L. M. J., Boringa J. B., Reuling I. E. W., Lazeron R. H. C., Ader H. J., & Polman C. H. (2002). Information processing characteristics in subtypes of multiple sclerosis. Neuropsychologia, 40(11), 1751–1765. pmid:12062887
        48. 48.Konrad , Günther , Hanisch , & Herpertz-Dahlmann . (2004). Differential Effects of Methylphenidate on Attentional Functions in Children With Attention-Deficit/Hyperactivity Disorder. Journal of the American Academy of Child & Adolescent Psychiatry, 43(2), 191–198.
        49. 49.Lehrer P. M., & Woolfolk R. L. (1982). Self-report assessment of anxiety: Somatic, cognitive and behavioural modalities. Behavioural Assessment, 4(2), 167–177.
        50. 50.Scholing A., & Emmelkamp P. M. G. (1992). Self report of anxiety: a cross validation of the Lehrer Woolfolk Anxiety symptom Questionnaire in three populations. Behav. Res. Ther. 30, 521–531. pmid:1520238
        51. 51.IBM Corp. Released 2015. IBM SPSS Statistics for Windows, Version 23.0. Armonk, NY: IBM Corp.
        52. 52.Borenstein M. (2009) Effect sizes for continuous data. In Cooper H., Hedges L. V., & Valentine J. C. (Eds.), The handbook of research synthesis and meta-analysis (2nd ed., pp. 221–235). Sage Foundation: New York, NY.
        53. 53.Friedman B. H., & Thayer J. F. (1998). Anxiety and autonomic flexibility: a cardiovascular approach. Biological psychology, 47(3), 243–263. pmid:9564452
        54. 54.Thayer J. F., Åhs F., Fredrikson M., Sollers J. J., & Wager T. D. (2012). A meta-analysis of heart rate variability and neuroimaging studies: Implications for heart rate variability as a marker of stress and health. Neuroscience & Biobehavioral Reviews, 36(2), 747–756.
        55. 55.Segerstrom S. C., & Nes L. S. (2007). Heart Rate Variability Reflects Self-Regulatory Strength, Effort, and Fatigue. Psychological Science, 18(3), 275–281. pmid:17444926
        56. 56.Williams D.P., Cash C., Rankin C., Bernardi A., Koenig J., & Thayer J.F. (2015). Resting heart rate variability predicts self-reported difficulties in emotion regulation: a focus on different facets of emotion regulation. Front. Psychol. 6:261. pmid:25806017
        57. 57.Williams D., Feeling N., Hill L., Spangler D., Koenig J., & Thayer J. (2017). Resting Heart Rate Variability, Facets of Rumination and Trait Anxiety: Implications for the Perseverative Cognition Hypothesis. Frontiers In Human Neuroscience, 11, Frontiers In Human Neuroscience, 2017 Oct 31, Vol.11
        58. 58.Fisher A., & Newman M. (2013). Heart rate and autonomic response to stress after experimental induction of worry versus relaxation in healthy, high-worry, and generalized anxiety disorder individuals. Biological Psychology, 93(1), 65–74. pmid:23384513
        59. 59.Asay T.P., & Lambert M.J. (1999). The empirical case for the common factors in therapy: Quantitative findings. In Hubble M. A., Duncan B. L., & Miller S. D. (Eds.), The heart and soul of change: What works in therapy (pp. 23–55). Washington, DC: American Psychological Association.
        60. 60.Taylor SE. & Brown JD (1988). Positive illusions and well-being revisited: separating fact from fiction. Psychological Bulletin 1994, Vol. 116, No.1. 21–27.
        61. 61.Suveg C., Morelen D., Brewer G. A., & Thomassin K. (2010). The emotion dysregulation model of anxiety: A preliminary path analytic examination. Journal of Anxiety Disorders, 24(8), 924–930. pmid:20634040
        62. 62.Bolwerk A., Mack-Andrick J., Lang F. R., Dörfler A., & Maihöfner C. (2014). How art changes your brain: Differential effects of visual art production and cognitive art evaluation on functional brain connectivity. PloS one, 9(7), e101035. pmid:24983951
        63. 63.Van Hooren, S. (2018). Vaktherapie. Doen wat werkt. Rede. Open Universiteit [Art therapy. Do what works. Inaugural speech. Open University, The Netherlands].
        64. 64.Driggers R., Ho C., Korhonen E., Kuivanen S., Jääskeläinen A., Smura T. et al. (2016). Zika Virus Infection with Prolonged Maternal Viremia and Fetal Brain Abnormalities. The New England Journal of Medicine, 374(22), 2142–2151. pmid:27028667

        Download PDF




        Subject Areas?

        The FDA estimates that Essure has been used by more than 750,000 women worldwide since it was first approved by global regulatory authorities. It involves insertion of coils into the Fallopian tubes to block passage of an egg from the ovary. 

        Implantation of the device has been associated with serious risks including persistent pain, uterine and Fallopian tube perforation, and coil migration into the pelvis or abdomen. Recent reports also have focused on issues related to surgery for removal of Essure. To address the reports, FDA ordered its manufacturer, Bayer, to conduct a post-market study and to add a boxed warning and a Patient Decision Checklist to the labeling.   

        In a press release, Bayer said that their decision “is based on a decline in U.S. sales of Essure in recent years and the conclusion that the Essure business if no longer sustainable.” A statement from FDA Commissioner Scott Gottlieb, MD, indicates that since the change to the labeling, sales of the device in the United States have fallen by approximately 70%. 

        Bayer will continue to enroll new participants in the post-marketing study, each of whom will be followed for 3 years. The company also will continue to submit reports to the FDA on the study’s progress and results. 

        The FDA said that women who use Essure can continue to do so and should consult with their doctors if they have symptoms related to the device. In its response to the impending removal of the device from the market, the American College of Obstetricians and Gynecologists encouraged ob/gyns to consult its Sterilization: Resource Overview for publications and resources on permanent contraception. 

        Survival after Minimally Invasive Radical Hysterectomy for Early-Stage Cervical Cancer
        • Alexander Melamed, M.D., M.P.H., 
        • Daniel J. Margul, M.D., Ph.D., 
        • Ling Chen, M.D, M.P.H., 
        • Nancy L. Keating, M.D., M.P.H., 
        • Marcela G. del Carmen, M.D., M.P.H., 
        • Junhua Yang, M.S., 
        • Brandon-Luke L. Seagle, M.D., 
        • Amy Alexander, M.D., 
        • Emma L. Barber, M.D., 
        • Laurel W. Rice, M.D., 
        • Jason D. Wright, M.D., 
        • Masha Kocherginsky, Ph.D., 
        • et al.



        Minimally invasive surgery was adopted as an alternative to laparotomy (open surgery) for radical hysterectomy in patients with early-stage cervical cancer before high-quality evidence regarding its effect on survival was available. We sought to determine the effect of minimally invasive surgery on all-cause mortality among women undergoing radical hysterectomy for cervical cancer.


        We performed a cohort study involving women who underwent radical hysterectomy for stage IA2 or IB1 cervical cancer during the 2010–2013 period at Commission on Cancer–accredited hospitals in the United States. The study used inverse probability of treatment propensity-score weighting. We also conducted an interrupted time-series analysis involving women who underwent radical hysterectomy for cervical cancer during the 2000–2010 period, using the Surveillance, Epidemiology, and End Results program database.


        In the primary analysis, 1225 of 2461 women (49.8%) underwent minimally invasive surgery. Women treated with minimally invasive surgery were more often white, privately insured, and from ZIP Codes with higher socioeconomic status, had smaller, lower-grade tumors, and were more likely to have received a diagnosis later in the study period than women who underwent open surgery. Over a median follow-up of 45 months, the 4-year mortality was 9.1% among women who underwent minimally invasive surgery and 5.3% among those who underwent open surgery (hazard ratio, 1.65; 95% confidence interval [CI], 1.22 to 2.22; P=0.002 by the log-rank test). Before the adoption of minimally invasive radical hysterectomy (i.e., in the 2000–2006 period), the 4-year relative survival rate among women who underwent radical hysterectomy for cervical cancer remained stable (annual percentage change, 0.3%; 95% CI, −0.1 to 0.6). The adoption of minimally invasive surgery coincided with a decline in the 4-year relative survival rate of 0.8% (95% CI, 0.3 to 1.4) per year after 2006 (P=0.01 for change of trend).


        In an epidemiologic study, minimally invasive radical hysterectomy was associated with shorter overall survival than open surgery among women with stage IA2 or IB1 cervical carcinoma. (Funded by the National Cancer Institute and others.)

        Leave a Comment

        This site uses Akismet to reduce spam. Learn how your comment data is processed.