Pszichológia | Felsőoktatás » Anderl-Chen - Oral Contraceptive Use in Adolescence Predicts Vulnerability to Depression in Adulthood

Oral Contraceptive Use in Adolescence Predicts Vulnerability to Depression in Adulthood Christine Anderla & Frances S. Chena a Department of Psychology, University of British Columbia, Vancouver, Canada Draft version 1.1, 11/06/17 This manuscript has not been peer reviewed Please address correspondence to: Christine Anderl Department of Psychology, University of British Columbia 2136 West Mall, Vancouver BC Canada V6T 1Z4 Email: anderl@psych.ubcca 2 Summary Depression is the leading cause of disability and suicide deaths worldwide1. Beginning in adolescence, women are twice as likely as men to develop depression1,2 and converging evidence suggests that gonadal hormones contribute to these sex differences2,3. Recent findings indicate that current use of oral contraceptives (OC) is associated with increased vulnerability to depression, especially amongst adolescents2. Critically, adolescents’ maturing brains are known to undergo permanent changes due to gonadal

hormones3-5, yet it remains unknown whether effects of adolescent OC use on mental health may persist into adulthood. Here we show that prior OC use in adolescence predicts a higher 1-year prevalence of depression in adulthoodirrespective of current OC use. We found in a nationally representative sample of women in the United States that women who had used OC during adolescence (n = 561) were at least twice as likely (17.1%, 95%CI: 132-209) to meet the criteria for major depressive disorder (MDD) years later compared to women who had never used OC (4.6%, 95%CI: 17-75%; n = 322) or only started using OC in adulthood (87%, 95%CI: 5.0-124%; n = 353) We also show that “third factors” that have previously been proposed to explain the relationship between OC use and depression risk (e.g, timing of menarche and age of first sexual intercourse)6-8 do not account for the pattern of results. Our findings go beyond prior reports of associations between OC use and concurrent depression risk,

in that they suggest that OC use in adolescence may have long-term, possibly irreversible, consequences on mental health. 3 Millions of women worldwide use OC2,3, which expose women to synthetic forms of estrogen and progestogen while suppressing the cyclical endogenous production of these hormones9. The number of adolescent OC users is increasing particularly rapidly; in the United States, over half of the women aged 15–19 years now use OCmore than in any other age group10 and one-third of these adolescent women do so solely for noncontraceptive reasons (reducing cramps or menstrual pain, menstrual regulation, treatment of acne, etc.)11 Adolescence is a time of maturation for the brain and other organs, and adolescents’ developing brains are known to undergo permanent changes due to sex hormones3-5. Thus, the question arises of whether adolescence represents a “critical period” during which the use of OC could affect the course of brain maturation. Building on recent

findings from a large-scale prospective study indicating that current use of OC or other forms of hormonal contraceptives is associated with increased use of antidepressants and a higher likelihood of a first diagnosis of depression, especially amongst adolescents2, here we address the question of whether the effects of adolescent OC use on depression may persist into adulthood, even after a woman stops using OC. To test our hypothesis that OC use in adolescence predicts a higher likelihood for depression in adulthood, we analyzed data from N = 1236 women enrolled in the United States National Health and Nutrition Examination Survey (NHANES)12, a nationally representative sample of people in the United States. Adult women who had used OC during adolescence (n = 561) showed a higher 1-year prevalence rate for MDD (17.1%, 95%CI: 13.2-209%) compared to both adult women who had never used OC (46%, 95%CI: 17-75; P < .001; n = 322) and women who had only started taking OC in adulthood

(87%, 95%CI: 5.0-124%; P = 017; n = 353) As depicted in the Table, these effects remained stable when controlling for important demographic variables and a large number of other factors associated with risk for depression2,6-8,12,13, other mental disorders available in the data set (general anxiety and panic disorder)12,13, and OC use in the past year. Moreover, the effect 4 was robust to using age at first OC use (in years) as a continuous predictor instead of a binary predictor for ever users of OC both across the whole sample (Odds Ratio (OR): 0.89, 95%CI: 0.82-096, P = 007) and within adolescent users of OC (OR: 086, 95%CI: 075-099, P = .04) Our findings indicate that women who had used OC during adolescence showed an at least two-fold higher 1-year prevalence of depression in adulthood compared to both women who had never used OC, and to women who had only started using OC after adolescence. While the correlational nature of epidemiological data precludes causal conclusions,

these effects remained stable when controlling for a large number of other factors that have been associated with risk for depression. Critically, by indicating that women who had used OC during adolescence are at heightened risk for depression in adulthood independent of current use, our findings suggest that taking OC in adolescence may increase women’s likelihood of developing depression across the lifetime. Our results point to a lasting adverse effect of early OC use on mental health. While there is no doubt that empowering women of all ages through access to effective methods of birth control is and should continue to be a major global health priority14, the presented findings reveal an urgent need for future studies to shed light on the questions of whether there is indeed a causal link between use of hormonal contraception in adolescence and the risk of developing psychopathologies in the short and long term, why such a relationship might exist, and which women may be

particularly sensitive to such adverse effects related to using hormonal contraceptives. In informing decision-making about the prescription of hormonal contraceptives to adolescent women for both contraceptive and noncontraceptive reasons, this line of research may have a major impact on the health of women worldwide. 5 Methods Data and population. We analyzed data from all women in NHANES12 for whom information on mental health and age at first OC use was publicly available (NHANES 19992004). NHANES is a continuous program of cross-sectional surveys which combine interviews and physical examinations to assess the health and nutritional status of the U.S population; it uses a complex, multistage, probability sampling design to select participants representative of the non-institutionalized population in the United States and a sample weight is assigned to each sample person to control for the unequal probability of selection, nonresponse adjustment, and adjustment to independent

population controls12. From 19992004, half of the NHANES sample aged 20-39 years received interview modules assessing psychiatric disorders (n = 1243). We excluded all women from analyses who had incomplete data for either the depression module (n = 3) or in questions assessing OC use (n = 4), resulting in a final sample of N = 1236 women (age: M = 29.11; SD = 570) Oral contraceptive use. Current and past OC use was determined via self-report in the reproductive health module of NHANES 1999-2004. Participants reported on whether they had ever used OC, and if so, at what age they had first started using OC, and whether they were currently using OC. If they reported that they were not currently using OC, they were asked at what age they had last taken OC. We defined first OC use in adolescence as first intake of OC at an age ≤ 19 years2 (n = 561) and first OC use in adulthood as first intake of OC at an age > 19 years (n = 353). Women who reported that they had never taken any birth

control pills were defined as never users of OC (n = 322). Following the procedures described in prior research with the same data set13, we defined women as current OC users if they reported currently taking birth control pills, or having recently taken birth control pills but stopped at an age equal to or one year lower than their current age; this definition of “current use” thus covered the same time period as the MDD diagnosis measure (i.e, 1-year prevalence). 6 Depression. A computer-assisted version of the World Health Organization Composite International Diagnostic Interview (CIDI-Auto 2.1) was administered to selected participants to assess depression in the past 12 months12. For all our analyses, diagnostic status for clinical depression was defined as meeting DSM-5 diagnostic criteria for MDD. None of the main results changed substantially when we used DSM-IV criteria instead (all Ps < .10) Data analysis. We used binary logistic regression models to predict 1-year

prevalence of MDD from onset of OC use. Women with first OC use in adolescence were used as reference category to allow for concurrent comparison of these women with both women who had never used OC and women who had their first use of OC in adulthood within the same regression analysis. To test the robustness of our findings, we then repeated these analyses while controlling for central demographic characteristics (NHANES cohort, age, educational level (high school graduate [reference]; less than high school education; education beyond high school), poverty income ratio (a ratio of family income to poverty threshold), and ethnicity (non-Hispanic white [reference]; non-Hispanic black; Hispanic; other)) and additional variables that have been associated with depression in prior research (body mass index (weight in kilograms divided by height in meters squared), smoking (never smoker [reference]; former smoker; current smoker), age at first period, age at first sexual intercourse,

marital status (married [reference]; never married; other), and pregnancy in the past year (no [reference]; yes)). Analyses were further adjusted for presence of other mental disorders available in the data set (general anxiety and panic disorder)12,13, and OC use in the past year. Finally, to rule out that the presented results were an artefact of our specific age-based categories, we tested whether early OC use predicted MDD by using age of first OC use in years as a continuous (as opposed to categorical) predictor. All analyses were performed using the “survey” package15 in R to account for the complex NHANES sampling design and sampling weights12. 7 Acknowledgments The authors thank V. Chu, J LeMoult, L Horng, and V Huang for feedback on the manuscript and B. Zareian for assistance in data preparation CA is supported by a Feodor Lynen Research Fellowship from the Alexander von Humboldt-Foundation (DEU 1187856 FLF-P). Author Contributions C.A and FSC developed the study

concept and planned the data analysis, which was performed by C.A Both authors contributed to the interpretation of the data CA drafted the manuscript and F.SC provided critical revisions Author Information The authors declare no competing financial interests. Correspondence and requests for materials should be addressed to C.A (anderl@psychubcca) 8 References 1. World Health Organization Depression and other common mental disorders: Global health estimates. (2017) at <http://appswhoint/iris/bitstream/10665/254610/1/WHOMSD-MER-20172-engpdf?ua=1> 2. Skovlund, C W, Mørch, L S, Kessing, L V & Lidegaard, Ø Association of hormonal contraception with depression. JAMA Psychiatry 73, 1154–1162 (2016) 3. Montoya, E R & Bos, P A How oral contraceptives impact social-emotional behavior and brain function. Trends Cogn Sci 21, 125–136 (2017) 4. Naninck, E F G, Lucassen, P J & Bakker, J Sex differences in adolescent depression: do sex hormones determine vulnerability?. J

Neuroendocrinol 23, 383–392 (2011) 5. Schulz, K M, Molenda-Figueira, H A & Sisk, C L Back to the future: the organizational–activational hypothesis adapted to puberty and adolescence. Horm Behav 55, 597–604 (2009). 6. Karina, I M & Sivakumaran, P Hormonal contraception and its association with depression. JAMA Psychiatry 74, 301–301 (2017) 7. Joinson, C, Heron, J, Lewis, G, Croudace, T & Araya, R Timing of menarche and depressive symptoms in adolescent girls from a UK cohort. Br J Psychiatry 198, 17–23 (2011). 8. Kaltiala-Heino, R, Kosunen, E & Rimpelä, M Pubertal timing, sexual behaviour and self-reported depression in middle adolescence. J Adolesc 26, 531–545 (2003) 9. Fleischman, D S, Navarrete, C D & Fessler, D M Oral contraceptives suppress ovarian hormone production. Psychol Sci 21, 750–752 (2010) 10. Jones, J, Mosher, W & Daniels, K Current contraceptive use in the United States, 2006– 2010, and changes in patterns of use since 1995. Natl

Health Stat Report 60, 1–25 (2012). 9 11. Jones, R K Beyond birth control: The overlooked benefits of oral contraceptive pills (Guttmacher Institute, 2011). 12. CDC National Center for Health Statistics National Health and Nutrition Examination Survey. (2015) at <http://wwwcdcgov/nchs/nhanes/nhanes questionnaireshtm> 13. Cheslack-Postava, K, Keyes, K M, Lowe, S R & Koenen, K C Oral contraceptive use and psychiatric disorders in a nationally representative sample of women. Arch Womens Ment. Health 18, 103–111 (2015) 14. Temmerman, M, Khosla, R, Laski, L, Mathews, Z & Say, L Women’s health priorities and interventions. BMJ 351, h4147 (2015) 15. Lumley T Survey: analysis of complex survey samples R package version 332-1 (2017). at <https://cranr-projectorg/web/packages/survey/indexhtml> 10 Table. Weighed Vulnerability to Depression Depending on Oral Contraceptive Use Never Users 1st OC Use 1st OC Use of OC in Adulthood in Adolescence (n = 322) (n =

353) (n = 561) 4.6 (17-75) 8.7 (50-124) 17.1 (132-209) ORunadjusted (95% CI) 1.00 [Reference] 1.99 (091-435) 4.27 (203-896)] ORcovariates (95% CI) 1.00 [Reference] 3.51 (074-1674) 8.04 (191-3375) ORother disorders (95% CI) 1.00 [Reference] 2.18 (078-606) 4.03 (159-1021) 1.00 [Reference] 2.12 (115-393) Major Depressive Disorder 1-Year Prevalence, % (95% CI) ORcurrent OC (95% CI)a Abbreviations: OC, oral contraceptives; CI, confidence interval; ORunadjusted, odds ratio unadjusted for covariates; ORcovariates, odds ratio adjusted for the following covariates: NHANES cohort, age, educational level (high school graduate [reference]; less than high school education; education beyond high school), poverty income ratio (a ratio of family income to poverty threshold), ethnicity (non-Hispanic white [reference]; non-Hispanic black; Hispanic; other), body mass index (weight in kilograms divided by height in meters squared), smoking (never smoker [reference]; former smoker;

current smoker), age at first period, age at first sexual intercourse, marital status (married [reference]; never married; other), and pregnancy in the past year (no [reference]; yes); ORother disorders, odds ratio adjusted presence of general anxiety or panic disorder; ORcurrent OC, odds ratio adjusted for OC use in the past 12 months. a Analysis only possible for subsample of women with some lifetime use of OC