State-level women’s status and psychiatric disorders among US women

Abstract

Purpose

Although greater gender equality at the state-level is associated with fewer depressive symptoms in women after controlling for individual-level confounders, the extent to which state-level women’s status is related to psychiatric disorders in women and gender differences in psychopathology has never been examined. We examined these associations in the current report.

Methods

We used data from the National Epidemiologic Survey on Alcohol and Related Conditions (n = 34,653), a national probability sample of US adults. Respondents completed structured diagnostic assessments of DSM-IV psychiatric disorders. We used generalized estimating equations to examine associations between four state-level indicators of women’s status (political participation, employment/earnings, social/economic autonomy, and reproductive rights) and odds of 12-month mood and anxiety disorders among women. We also tested whether women’s status predicted the magnitude of gender differences in psychiatric disorders.

Results

State-level political participation, employment/earnings, and social/economic autonomy were unrelated to odds of 12-month mood and anxiety disorders among women. However, the prevalence of major depression and post-traumatic stress disorder was lower in states where women have greater reproductive rights (OR 0.93–0.95), controlling for individual-level risk factors. None of the women’s status indicators predicted gender differences in mood and anxiety disorder prevalence.

Conclusions

State-level women’s status was largely unrelated to mood and anxiety disorders in women or to gender differences in these disorders. Investigation of social factors that play a role in shaping the distribution of individual-level risk factors that are associated with gender disparities in psychiatric disorders represents an important avenue for future research.

Introduction

The prevalence of mood and anxiety disorders is approximately twice as high among women as compared with men in the United States population [1–9]. This marked gender disparity in risk for mood and anxiety disorders begins in early adolescence [10, 11] and is evident throughout the remainder of the life course [7]. A wide range of individual-level factors have been posited to underlie these gender differences, ranging from sex differences in biological factors, including sex hormone changes during puberty [12, 13], to psychological factors including self-efficacy, stress reactivity, emotion regulation, and coping styles [12, 14, 15], to psychosocial and environmental factors such as exposure to interpersonal violence, child abuse, poverty, and other chronic stressors [12–15]. However, none of these individual-level factors fully explain the extant gender disparities [14, 16]. Macro-social factors are thought to play a role [17]; however, empirical tests of whether such factors explain gender differences in risk are sorely lacking. This paper is the first to examine state-level women’s status as a determinant of gender disparities in mood and anxiety disorder in the United States.

States exert a great deal of control over the construction of public policy relevant to women’s status. Substantial variability exists at the state-level in factors such as women’s labor force participation, earnings relative to men, political participation and representation, and reproductive rights [18]. Initial evidence for the importance of women’s status in shaping the distribution of women’s health and well-being came from a study of inter-marital violence that documented higher rates of partner violence directed at women by their husbands in states with lower levels of gender equality [19]. Since this initial study, state-level women’s status has been linked to a variety of health indicators including both global and cause-specific mortality rates and functional disability among both women and men, low birth weight, teen pregnancy, and infant and teen mortality [20, 21]. Recent evidence also suggests that lower women’s status at the state level is associated with higher levels of depressive symptoms among women [22]. However, to our knowledge, state-level women’s status has yet to be examined as a determinant of the prevalence of psychiatric disorders among women or of gender differences in disorder prevalence.

The current investigation addresses these two gaps in the literature. We first examined the associations between four indicators of state-level women’s status and the prevalence of mood and anxiety disorders among women in a US. national probability sample. We expected the prevalence of these disorders to be lower in states with higher women’s status. Second, we examined associations between state-level women’s status and the magnitude of gender differences in the prevalence of mood and anxiety disorders. We hypothesized that gender differences would be less pronounced in states with higher women’s status.

Methods

Sample

Data come from the 2004–2005 National Epidemiologic Survey on Alcohol and Related Conditions (NESARC), a population-based sample of psychiatric disorders in civilian non-institutionalized US adults aged 18 years and older [23, 24]. Wave 1 of the NESARC was carried out in 2001–2002 and included structured diagnostic interviews with 43,093 respondents with an overall response rate of 81.0%. Young adults, Blacks, and Hispanics were over-sampled. A second wave of interviews was conducted between 2004 and 2005 with 34,653 of the original NESARC respondents, for a conditional (on Wave 1 participation) response rate of 86.7%. A total of 20,089 women residing in all 50 states participated in the Wave 2 NESARC. The number of women per state ranged from 40 in Alaska to 2206 in California. Further information on the design and implementation of the NESARC is found elsewhere [24, 25]. Respondents provided written informed consent. The research protocol received full ethical review and approval from the US Census Bureau and the US Office of Management and Budget.

Measures

Psychiatric disorders

DSM-IV [26] mood (major depression and dysthymia) and anxiety disorders [generalized anxiety disorder (GAD), panic disorder with or without agoraphobia, social phobia, and post-traumatic stress disorder (PTSD)] were assessed with the Alcohol Use Disorder and Associated Disabilities Interview Schedule-DSM-IV Version (AUDADIS-IV) [27]. Substance-induced mood and anxiety disorders, those due to somatic illnesses, or (in the case of major depression) bereavement were ruled out as per DSM-IV definition. Diagnoses met the DSM-IV criterion requiring distress or social/occupational impairment. AUDADIS-IV mood and anxiety diagnoses have been found to be reliable and valid in a number of general population and clinical reappraisal studies [28–30]. The current analysis focuses on mood and anxiety disorders that were present in the 12 months prior to the Wave 2 interview.

Women’s status

Indicators of women’s status in four domains (reproductive rights, political participation, employment and earnings, and social and economic autonomy) were created by The Institute for Women’s Policy Research (IWPR) to examine state-level variation in women’s progress in achieving rights and opportunities and to identify barriers to gender equality [31]. These indicators have been used previously in epidemiologic research [20–22]. Indicators were selected based on relevance to women’s lives and reliability and comparability of data across states. State-level data regarding women’s status along a number of dimensions (described in more detail below) are used to give each state a score on the four indicators, and these composite scores are updated on a semi-annual basis. For all indicators, higher composite scores reflect higher levels of women’s status. We use women’s status indicators from 2002, the most recent year preceding the assessment of psychiatric disorders in which data are available for all states.

The political participation composite index reflects women’s power and participation in state-level politics and includes four components: women’s voter registration, voter turnout, representation in elected office, and the existence of institutional resources for women. The proportion of women holding political office at four levels was calculated and weighted based on prestige: (1) state representatives (weight = 1.0), (2) state senators (weight = 1.25), (3) executive officials and US representatives (weight = 1.5), and (4) US senators and governors (weight = 1.75). Institutional resources include a state commission and/or a legislative caucus for women. The four components were standardized by subtracting the mean national value from the state value and dividing by the national standard deviation. Standardized scores were summed to create the composite index. The components were given a weight of 1.0, with the exception of representation of women in elected office, which was given a weight of 4.0.

The employment and earnings composite index reflects women’s status related to occupation and income and includes four components: women’s median annual earnings, ratio of women’s to men’s earnings, women’s labor force participation, and women’s representation in managerial and professional occupations. Each component was standardized as previously described, and standardized scores were summed to create the composite. Scores for each component were weighted equally.

The social and economic autonomy composite index includes four components: proportion of women with health insurance, women’s educational attainment (proportion with a college degree), women’s business ownership, and proportion of women living above the federal poverty level. After standardizing each score as previously described, the scores were summed to create the composite. Proportion of women living above poverty was weighted 4.0, and the remaining components were weighted 1.0.

The final composite index reflects reproductive rights and includes nine components involving the presence of various legal and policy protections of women’s reproductive health. These include mandatory abortion consent/notification laws, mandatory abortion waiting period, coverage of infertility treatments, and legality of same sex couple adoption. After standardizing these components, they received a weight of 0.5 in the composite index. The composite also includes public funding for abortion, proportion of women living in counties with at least one abortion provider, contraceptive coverage laws, mandatory sexual education, and presence of a pro-choice legislature or governor. For each of these components, states were given a score of 0 for having a law that restricts reproductive rights and a score of 1 if they had no such law. In some cases, a score between 0 and 1 was given (e.g., states that required partial coverage of contraception received a score of 0.5, whereas states with full coverage received a score of 1). Detailed information regarding the four women’s status indicators and their components is available elsewhere [18, 20].

Covariates

We included the following controls in multivariate analyses: age, race-ethnicity (White, Black, Native American, Asian and Pacific Islander, and Hispanic), marital status (never married, married or cohabitating, divorced/separated), educational attainment (less than high school diploma, high school graduate, some college, and college graduate), household annual income (<$10,000, $10,000–$29,999, $30,000–$49,999, $50,000–$64,999, $70,000–$99,999, and ≥$100,000), and occupation in all models. Age was entered as series of dummy variables representing young (ages 20–34), middle-aged (ages 35–49), and older adults (ages 60–89). Occupation was coded into three categories: professional (executive, administrative, managerial, and professional specialty), technical/support/clerical (technical and related support, sales, administrative support, clerical, private household, protective services, and other services), and unskilled/manual (operators, fabricators, laborers, transportation and material moving, handlers, equipment cleaners, and military).

Statistical analysis

To estimate the associations between state-level women’s status and 12-month mood and anxiety disorders among women, we used logistic regression, applying generalized estimating equations (GEE). This analysis strategy accounts for both the complex sample design of the NESARC and for correlations among individuals living in the same state [32].¹ In the women sample, we first examined bivariate associations between women’s status indicators and 12-month mood and anxiety disorders, followed by the associations adjusted for (1) socio-demographics (age, race/ethnicity, marital status) and (2) both socio-demographics and socioeconomic factors (educational attainment, household income, occupation). To determine whether state-level women’s status was associated with the magnitude of gender differences in the prevalence of 12-month mood and anxiety disorders, we utilized the full sample and created interaction terms between women’s status and gender. The main effect of women’s status, gender, and their interaction were added to the model including socio-demographic and socio-economic factors to evaluate the presence of effect modification of gender by women’s status. The Wald Chi-square test statistic for the interaction term was used to evaluate the presence of statistically significant effect modification.

GEE parameter estimates and standard errors were exponentiated to generate odds ratios (ORs) and 95% confidence intervals (CIs). NESARC weights were applied to account for selection and response probabilities and to adjust the sample to be representative of the US population. Analyses were completed with SUDAAN software version 9.1 [33], which uses the Taylor series linearization method to calculate standard errors adjusted for the complex sample design. Statistical significance was tested using two-sided alpha level of 0.05.

Results

Descriptive statistics and distribution of covariates by women’s status

The distribution of socio-demographic factors in the NESARC according to state-level women’s status is presented in Table 1. We find no notable differences in the distribution of age, race/ethnicity, marital status, educational attainment, household income, and occupational prestige among women in states with high versus low women’s status. Consistent with prior research, [1–4, 8] the prevalence of major depression and anxiety disorders was higher among respondents who were younger, unmarried, White, had lower household incomes, and technical or manual occupations (Table 2).

Table 1.

Distribution of socio-demographic factors by four indicators of state-level women’s status among females (n = 20,089) in the National Epidemiologic Survey of Alcohol and Related Conditions (NESARC)

	Reproductive rightsa		Political participationa		Employment and earningsa		Social and economic autonomya
	Lowb	(%) High (%)	Low (%)	High (%)	Low (%)	High (%)	Low (%)	High (%)
Age
20–34	24.2	22.6	24.3	22.9	23.1	23.4	23.3	23.1
35–49	30.9	31.6	31.7	31.0	31.4	31.3	31.3	31.4
50–64	23.3	23.5	23.2	23.7	23.6	23.2	23.5	23.3
65–89	21.6	22.4	20.8	22.4	22.0	22.1	21.9	22.2
Race/ethnicity
Black	21.0	21.0	21.0	20.8	22.8	19.4	22.6	19.5
Native	2.1	1.4	2.0	1.5	1.7	1.6	1.8	1.6
Asian and Pacific Islander	1.8	3.3	2.3	3.2	1.8	3.5	2.1	3.2
Hispanic	15.4	20.1	18.1	17.2	15.7	20.4	17.3	19.0
White	59.7	54.2	56.6	57.3	58.0	55.0	56.2	56.7
Marital status
Never married	51.3	49.6	51.2	49.9	49.9	50.7	49.9	50.7
Widowed, divorced, or separated	16.1	18.6	17.1	18.2	17.5	17.7	17.7	17.5
Married or living with someone	32.6	31.8	31.7	31.9	32.6	31.7	32.4	31.9
Education
Less than high school diploma	17.1	16.7	17.3	16.9	17.2	16.5	17.6	16.2
High school graduate	29.4	29.2	29.3	29.1	29.4	29.2	29.5	29.2
Some college	31.1	30.7	31.2	30.3	30.9	30.8	30.7	31.0
College graduate and above	22.3	23.4	22.2	23.7	22.5	23.4	22.2	23.6
Household income
<$10,000	11.1	10.6	10.5	10.9	11.3	10.4	11.4	10.3
$10,000–$29,999	33.2	31.4	33.0	31.1	32.5	31.8	32.4	31.9
$30,000–$49,999	22.1	20.7	21.2	21.3	21.7	21.0	21.7	20.9
$50,000–$64,999	14.1	14.3	14.2	14.3	14.2	14.2	14.3	14.2
$70,000–$99,999	10.7	11.5	11.3	11.4	10.9	11.4	10.8	11.6
≥$100,000	8.7	11.5	9.7	11.0	9.5	11.2	9.4	11.2
Occupation
Technical/support/clerical	44.4	43.1	44.5	42.4	44.3	43.0	43.9	43.4
Unskilled/Manual	8.7	8.0	8.8	8.4	8.1	8.5	8.0	8.6
Occupation missing	18.3	19.8	18.0	19.8	18.9	19.4	19.4	19.1
Professional	28.5	29.1	28.7	29.4	28.7	29.0	28.8	29.0

^aState-level indicators of women’s status across four dimensions (reproductive rights, political participation, employment and earnings, and social and economic autonomy) were drawn from the Institute for Women’s Policy Research. See “Methods” for details

^bLow and high status were defined using a median split

Table 2.

Associations (odds ratios) between socio-demographic factors and 12-month mood and anxiety disorders among females (n = 20,089) in the National Epidemiologic Survey of Alcohol and Related Conditions (NESARC)

	Major depression		Dysthymia		GAD		Social phobia		PTSD		Panic disorder		Any mood disorder		Any anxiety disorder
Age
20–34	2.8*	(2.3–3.3)	2.2*	(1.4–3.5)	1.8*	(1.3–2.4)	3.9*	(2.5–5.8)	2.0*	(1.7–2.5)	4.3*	(3.0–6.3)	2.7*	(2.2–3.2)	2.3*	(1.9–2.7)
35–49	2.4*	(2.0–2.9)	2.3*	(1.5–3.7)	2.2*	(1.7–2.8)	3.6*	(2.4–5.4)	2.4*	(1.9–2.9)	4.5*	(2.9–5.9)	2.4*	(2.0–2.8)	2.5*	(2.1–2.9)
50–64	2.2*	(1.8–2.7)	2.2*	(1.5–3.4)	2.0*	(1.5–2.7)	2.5*	(1.7–3.8)	1.9*	(1.5–2.3)	2.7*	(1.9–4.0)	2.2*	(1.8–2.7)	1.9*	(1.6–2.3)
65–89	1.0	–	1.0	–	1.0	–	1.0	–	1.0	–	1.0	–	1.0	–	1.0	–
${\chi }_3^2$ (p value)	177.0*	(<0.001)	36.5*	(<0.001)	79.3*	(<0.001)	77.6*	(<0.001)	128.5*	(<0.001)	112.3*	(<0.001)	179.8*	(<0.001)	236.9*	(<0.001)
Race/ethnicity
Black	0.9	(0.8–1.1)	1.3	(1.0–1.7)	0.8	(0.7–1.0)	0.8	(0.7–1.1)	1.1	(0.9–1.3)	0.9	(0.7–1.2)	0.9	(0.8–1.1)	1.0	(0.8–1.1)
Native	1.6*	(1.1–2.2)	1.3	(0.5–3.5)	0.8	(0.4–1.7)	1.9	(1.1–3.3)	1.7*	(1.2–2.5)	1.5	(0.9–2.7)	1.5*	(1.1–2.1)	1.3	(0.9–1.8)
Asian and Pacific Islander	0.7	(0.5–1.1)	1.5	(0.7–3.1)	0.7	(0.4–1.4)	0.6	(0.3–1.2)	0.6*	(0.4–0.9)	0.8	(0.3–1.9)	0.7	(0.5–1.1)	0.7*	(0.5–0.9)
Hispanic	1.0	(0.8–1.1)	1.8*	(1.3–2.7)	0.7*	(0.5–0.9)	0.9	(0.6–1.3)	1.0	(0.8–1.2)	1.0	(0.8–1.3)	1.0	(0.9–1.2)	0.9	(0.7–1.1)
White	1.0	–	1.0	–	1.0	–	1.0	–	1.0	–	1.0	–	1.0	–	1.0	–
${\chi }_4^2$ (p value)	21.3*	(<0.001)	13.7*	(<0.001)	17.3*	(<0.001)	8.2	(.08)	28.1*	(<0.001)	6.6	(.16)	22.4*	(<0.001)	25.8*	(<0.001)
Marital status
Never married	1.4*	(1.3–1.6)	1.9*	(1.4–2.6)	1.3*	(1.1–1.6)	1.5*	(1.2–1.9)	1.4*	(1.2–1.6)	1.2	(1.0–1.5)	1.4*	(1.3–1.6)	1.4*	(1.2–1.5)
Widowed, divorced, or separated	1.6*	(1.4–1.9)	2.0*	(1.3–2.9)	1.0	(0.8–1.4)	1.6*	(1.2–2.1)	1.1	(0.9–1.3)	1.1	(0.9–1.5)	1.6*	(1.3–1.8)	1.2*	(1.0–1.4)
Married or living with someone	1.0	–	1.0	–	1.0	–	1.0	–	1.0	–	1.0	–	1.0	–	1.0	–
${\chi }_2^2$ (p value)	54.7*	(<0.001)	22.9*	(<0.001)	11.9*	(<0.001)	27.9*	(<0.001)	33.2*	(<0.001)	5.9*	(0.05)	54.7*	(<0.001)	39.6*	(<0.001)
Education
Less than high school diploma	1.0	(0.9–1.2)	0.7	(0.5–1.0)	0.8	(0.6–1.0)	1.3	(1.0–1.8)	0.9	(0.7–1.0)	0.7*	(0.5–1.0)	1.0	(0.9–1.2)	0.9*	(0.8–1.0)
High school graduate	1.1	(0.9–1.3)	0.6*	(0.4–0.9)	1.0	(0.8–1.3)	1.2	(0.9–1.7)	1.0	(0.8–1.2)	0.7*	(0.5–0.9)	1.1	(0.9–1.3)	1.0	(0.9–1.2)
Some college	0.8*	(0.6–0.9)	0.5*	(0.3–0.7)	0.7	(0.5–1.0)	0.7	(0.5–1.1)	0.7*	(0.6–0.8)	0.4*	(0.3–0.6)	0.7*	(0.6–0.9)	0.7*	(0.6–0.8)
College graduate and above	1.0	–	1.0	–	1.0	–	1.0	–	1.0	–	1.0	–	1.0	–	1.0	–
${\chi }_3^2$ (p value)	26.1*	(<0.001)	21.1*	(<0.001)	14.0*	(<0.001)	20.3*	(<0.001)	36.3*	(<0.001)	32.9*	(<0.001)	27.8*	(<0.001)	48.1*	(<0.001)
Household Income
<$10,000	2.1*	(1.6–2.8)	9.7*	(4.7–20.1)	2.0*	(1.4–3.0)	2.5*	(1.6–3.8)	1.8*	(1.4–2.4)	2.3*	(1.5–3.5)	2.3*	(1.7–3.0)	1.9*	(1.5–2.4)
$10,000–$29,999	1.7*	(1.4–2.1)	7.3*	(3.5–14.9)	1.7*	(1.2–2.2)	1.7*	(1.2–2.6)	1.5*	(1.1–1.9)	1.7*	(1.2–2.4)	1.7*	(1.4–2.2)	1.4*	(1.2–1.8)
$30,000–$49,999	1.4*	(1.1–1.8)	4.9*	(2.4–10.2)	1.4	(1.0–1.9)	1.5	(1.0–2.3)	1.3	(1.0–1.7)	1.3	(0.9–2.0)	1.5*	(1.2–1.8)	1.3*	(1.1–1.6)
$50,000–$64,999	1.2	(0.9–1.6)	4.0*	(1.7–9.5)	1.2	(0.8–1.8)	2.1*	(1.3–3.2)	1.0	(0.7–1.4)	1.2	(0.8–1.8)	1.2	(1.0–1.6)	1.2	(1.0–1.5)
$70,000–$99,999	1.1	(0.9–1.4)	2.5*	(1.1–5.9)	1.2	(0.8–1.9)	1.3	(0.8–2.1)	0.9	(0.7–1.3)	1.1	(0.7–1.9)	1.1	(0.9–1.4)	1.0	(0.9–1.3)
≥$100,000	1.0	–	1.0	–	1.0	–	1.0	–	1.0	–	1.0	–	1.0	–	1.0	–
${\chi }_5^2$ (p value)	90.5*	(<0.001)	79.9*	(<0.001)	37.6*	(<0.001)	44.5*	(<0.001)	83.5*	(<0.001)	48.7*	(<0.001)	99.1*	(<0.001)	88.4*	(<0.001)
Occupation
Technical/support/clerical	1.3*	(1.2–1.5)	1.9*	(1.3–2.9)	1.1	(0.9–1.3)	1.3	(1.0–1.6)	1.1	(1.0–1.3)	1.4*	(1.1–1.7)	1.3*	(1.2–1.5)	1.2*	(1.1–1.3)
Unskilled/manual	1.1	(0.9–1.4)	2.2*	(1.3–3.9)	1.4	(1.0–2.0)	1.3	(0.9–2.0)	1.3*	(1.1–1.7)	1.7*	(1.1–2.6)	1.1	(0.9–1.4)	1.4*	(1.1–1.7)
Occupation missing	0.9	(0.8–1.1)	2.2*	(1.4–3.5)	1.1	(0.9–1.4)	1.0	(0.7–1.3)	0.9	(0.7–1.0)	1.0	(0.7–1.4)	1.0	(0.8–1.2)	0.9	(0.8–1.0)
Professional	1.0	–	1.0	–	1.0	–	1.0	–	1.0	–	1.0	–	1.0	–	1.0	–
${\chi }_3^2$ (p value)	34.8*	(<0.001)	19.9*	(<0.001)	3.7	(0.29)	8.3*	(0.04)	15.1*	(<0.001)	23.1*	(<0.001)	30.8*	(<0.001)	37.0*	(<0.001)

^*p <0.05, 2-sided test

Women’s status and 12-month mood and anxiety disorders

In bivariate models, women’s state-level political participation was associated with only one of the mood and anxiety disorder outcomes, and this association was not in the expected direction (Table 3). Greater state-level political participation was associated with elevated odds of 12-month dysthymia (OR 1.05), and this association remained statistically significant after adjustment for socio-demographic and socio-economic factors.

Table 3.

Bivariate and adjusted associations (odds ratios) between state-level women’s political participation and 12-month mood and anxiety disorders among women (n = 20,089)

	Bivariate		Adjusteda		Adjustedb
	OR	(95% CI)	OR	(95% CI)	OR	(95% CI)
Mood disorders	1.01	(1.0–1.03)	1.01	(1.0–10.3)	1.02	(1.0–1.03)
Major depression	1.01	(0.99–1.02)	1.01	(1.0–1.03)	1.01	(1.0–1.03)
Dysthymia	1.05*	(1.01–1.09)	1.05*	(1.01–1.09)	1.05*	(1.01–1.10)
Anxiety disorders	1.00	(0.98–1.01)	1.00	(0.98–1.01)	1.00	(0.98–1.01)
Panic Disorder	1.00	(0.98–1.03)	1.01	(0.98–1.04)	1.01	(0.98–1.04)
Generalized anxiety disorder	0.99	(0.97–1.02)	0.99	(0.97–1.02)	0.99	(0.97–1.02)
Social phobia	1.02	(0.99–1.05)	1.03	(1.0–1.06)	1.03	(1.0–1.06)
Posttraumatic stress disorder	1.00	(0.99–1.02)	1.01	(0.99–1.02)	1.01	(0.99–1.02)

State-level indicators of women’s political participation were drawn from the Institute for Women’s Policy Research. See “Methods” for details. Analysis conducted using generalized estimating equations

^*p <0.05, 2-sided test

^aControlling for age, race/ethnicity, and marital status

^bControlling for age, race/ethnicity, and marital status, household income, educational attainment, and occupation

Women’s employment and earnings at the state level was not associated with 12-month prevalence of any of the mood and anxiety disorder outcomes in bivariate models or in the adjusted multivariate models (Table 4).

Table 4.

Bivariate and adjusted associations (odds ratios) between state-level women’s employment and earnings and 12-month mood and anxiety disorders among women (n = 20,089)

	Bivariate		Adjusteda		Adjustedb
	OR	(95% CI)	OR	(95% CI)	OR	(95% CI)
Mood disorders	1.22	(0.70–2.16)	1.16	(0.66–2.01)	1.16	(0.67–1.99)
Major depression	1.30	(0.72–2.34)	1.24	(0.70–2.22)	1.25	(0.71–2.19)
Dysthymia	0.45	(0.12–1.69)	0.31	(0.09–1.15)	0.30	(0.08–1.14)
Anxiety disorders	1.34	(0.84–2.14)	1.37	(0.84–2.22)	1.38	(0.82–2.30)
Panic disorder	1.07	(0.45–2.52)	1.01	(0.44–2.35)	1.01	(0.43–2.37)
Generalized anxiety disorder	1.25	(0.58–2.67)	1.32	(0.61–2.84)	1.31	(0.60–2.86)
Social phobia	0.65	(0.26–1.66)	0.62	(0.25–1.57)	0.64	(0.25–1.66)
Posttraumatic stress disorder	1.38	(0.77–2.47)	1.34	(0.74–2.44)	1.34	(0.71–2.52)

State-level indicators of women’s social and economic autonomy were drawn from the Institute for Women’s Policy Research. See “Methods” for details. Analysis conducted using generalized estimating equations

^aControlling for age, race/ethnicity, and marital status

^bControlling for age, race/ethnicity, and marital status, household income, educational attainment, and occupation

Associations between women’s state-level social and economic autonomy and 12-month mood and anxiety disorders were not significant in bivariate or adjusted multivariate models (Table 5).

Table 5.

Bivariate and adjusted associations (odds ratios) between state-level women’s social and economic autonomy and 12-month mood and anxiety disorders among women (n = 20,089)

	Bivariate		Adjusteda		Adjustedb
	OR	(95% CI)	OR	(95% CI)	OR	(95% CI)
Mood disorders	1.00	(0.70–1.44)	1.02	(0.72–1.45)	1.05	(0.75–1.46)
Major depression	0.97	(0.66–1.42)	0.97	(0.67–1.39)	0.99	(0.70–1.40)
Dysthymia	1.74	(0.63–4.79)	2.32	(0.84–6.45)	2.50	(0.88–7.08)
Anxiety disorders	0.97	(0.69–1.36)	0.93	(0.66–1.30)	0.95	(0.67–1.35)
Panic disorder	0.92	(0.49–1.74)	0.92	(0.50–1.71)	0.96	(0.52–1.80)
Generalized anxiety disorder	1.05	(0.61–1.80)	0.97	(0.72–1.31)	1.01	(0.59–1.72)
Social phobia	1.35	(0.68–2.68)	1.33	(0.67–2.64)	1.34	(0.66–2.69)
Posttraumatic stress disorder	0.99	(0.68–1.44)	0.99	(0.67–1.45)	1.02	(0.68–1.53)

State-level indicators of women’s social and economic autonomy were drawn from the Institute for Women’s Policy Research. See “Methods” for details. Analysis conducted using generalized estimating equations

^aControlling for age, race/ethnicity, and marital status

^bControlling for age, race/ethnicity, and marital status, household income, educational attainment, and occupation

State-level reproductive rights were significantly associated with several mood and anxiety disorder outcomes in bivariate models (Table 6). The odds of having 12-month major depression (OR 0.94) and PTSD (OR 0.92) was lower among women living in states with greater reproductive rights. Greater reproductive rights at the state level also were associated with reduced odds of having any mood disorder (OR 0.93) and any anxiety disorder (OR 0.95) among women. The associations between women’s reproductive rights and major depression, PTSD, and any mood disorder remained significant after adjustment for socio-demographic and socio-economic factors.

Table 6.

Bivariate and adjusted associations (odds ratios) between state-level women’s reproductive rights and 12-month mood and anxiety disorders among women (n = 20,089)

	Bivariate		Adjusteda		Adjustedb
	OR	(95% CI)	OR	(95% CI)	OR	(95% CI)
Mood disorders	0.93*	(0.89–0.97)	0.94*	(0.90–0.98)	0.94*	(0.90–0.98)
Major depression	0.94*	(0.90–0.98)	0.95*	(0.90–0.99)	0.95*	(0.91–0.99)
Dysthymia	0.92	(0.83–1.03)	0.91	(0.80–1.02)	0.91	(0.81–1.02)
Anxiety disorders	0.95*	(0.91–0.99)	0.96	(0.92–1.01)	0.96	(0.92–1.00)
Panic disorder	0.94	(0.87–1.02)	0.95	(0.88–1.03)	0.95	(0.88–1.03)
Generalized anxiety disorder	0.93	(0.87–1.0)	0.95	(0.88–1.02)	0.95	(0.88–1.02)
Social phobia	0.93	(0.86–1.02)	0.95	(0.87–1.03)	0.95	(0.87–1.03)
Posttraumatic stress disorder	0.92*	(0.88–0.97)	0.93*	(0.88–0.98)	0.93*	(0.88–0.98)

State-level indicators of women’s reproductive rights were drawn from the Institute for Women’s Policy Research. See “Methods” for details. Analysis conducted using generalized estimating equations

^*p <0.05, 2-sided test

^aControlling for age, race/ethnicity, and marital status

^bControlling for age, race/ethnicity, and marital status, household income, educational attainment, and occupation

Women’s status and gender differences in mood and anxiety disorders

To determine whether gender differences in the prevalence of 12-month mood and anxiety disorders were greater in states with lower women’s status, we examined interactions between gender and each of the four women’s status indicators. In contrast to the previous analyses, these models included the total sample of men and women. Of 32 possible interactions of this sort, none were significantly associated with odds of past-year mood or anxiety disorders, indicating no effect modification of gender on 12-month mood and anxiety disorders by state-level women’s status (detailed results not shown but available on request).

Discussion

Our results indicated state-level women’s status was largely unrelated to the distribution of mood and anxiety disorders among women and to gender disparities in these disorders. The single exception was our finding that women living in states with greater reproductive rights had a lower 12-month prevalence of major depression and PTSD. This finding is consistent with prior work on depressive symptoms [22]. Also consistent with this work, we found no negative association between women’s political participation and the prevalence of mood and anxiety disorders. In contrast, however, we also found no association between state-level employment and earnings or social and economic autonomy and the prevalence of mood and anxiety disorders among women. Moreover, we found no evidence to suggest that gender differences in the prevalence of mood and anxiety disorders are related to state-level variation in women’s status. However, we found strong associations between individual-level factors including age, educational attainment, income, and occupational prestige and the prevalence of mood and anxiety disorders among women, which is consistent with previous epidemiologic research [1–3, 8, 9].

The significant association between state-level reproductive rights and lower mood and anxiety disorder prevalence among women after accounting for individual-level characteristics merits further investigation. These data suggest mood and anxiety disorder prevalence is lower in states that have enacted policies that provide funding or mandate insurance coverage for contraception and reduce barriers to obtaining an abortion. Although the mechanisms underlying this association are unclear, increased autonomy over reproduction may provide one explanation. Indeed, lower rates of unintended pregnancy have been documented in communities with lower gender inequality [34]. Increased autonomy over reproduction may reduce anxiety about unintended pregnancy and lead to actual reductions in the prevalence of unintended pregnancy, which has been linked to risk for both anxiety and depression among women [35, 36].

One potential explanation for the lack of association between the other indicators of women’s status and female psychiatric disorders concerns the difference in level between the putative social determinants and psychiatric outcomes. State-level women’s political participation, social and economic autonomy, and employment and earnings are contextual determinants that reflect the cumulative participation or representation of individual women in a state. Reproductive rights, on the other hand, reflect policies that both confer protection and determine the reproductive choices that are available to individual women and therefore come closer to the definition of a true contextual determinant. These protective policies are likely to have a more direct impact on individual women’s lives than, for example, the number of women who participate in politics in their state and, as a result, may be more strongly related to mental health. However, the association between reproductive rights and female mood and anxiety disorders warrants replication in future research given that this was the only significant association we found between an indicator of state-level women’s status and female psychiatric disorders. This association should be interpreted cautiously, as the large sample size used in the current analyses and number of tests conducted increases the likelihood of finding a significant association in the absence of a true relationship.

Although variations in state-level women’s status are associated with depressive symptoms among women [22], on the whole these variations are not important determinants of the prevalence of mood and anxiety disorders among women or gender differences in disorder prevalence. There are at least two possible explanations for the discrepancy between our findings and previous work. The first involves differences in the definition of mental health problems. Prior work used a self-report measure of depressive symptoms; we relied on structured diagnostic interviews of psychiatric diagnoses. Self-report measures of symptoms tend to measure non-specific psychological distress or ‘demoralization’ rather than psychiatric diagnoses [37] and perform relatively poorly at identifying cases of major depression as compared to diagnostic interviews [38–40]. These measurement problems with self-report scales arise as a result of incomplete coverage of diagnostic criteria (resulting in low sensitivity), inclusion of items that are not part of the diagnostic criteria (resulting in low specificity), and inability to ascertain clinical significance and functional impairment associated with symptoms [41]. As a result, symptom scales and diagnostic interviews assess different underlying constructs.

A second possibility is that women living in states with lower gender equality have higher levels of depressive symptoms, but these symptom elevations do not cross the threshold to psychiatric disorder. Consistent with this interpretation, depressive symptoms among women were less than one point higher in states with low status relative to states with high status in the one prior study on women’s status and mental health [22]. Importantly, we used measures of women’s status that are identical to those used in prior work on mental and physical health outcomes [20–22], ruling out potential methodological explanations resulting from differences in the measurement of women’s status.

Contrary to our predictions, we found that dysthymia was more common among women in states with greater female political participation. Although greater state-level political participation among women has been associated with decreased female mortality [21], no relationship was found between this indicator and female depressive symptoms in previous research [22] or with any of the other psychiatric disorders in the current analysis. It is difficult to provide an explanation for why dysthymia was more common in states with greater political participation of women. Dysthymia is the least common of the disorders considered here, with fewer than 2% of women in the NESARC meeting criteria for the disorder [42]. Dysthymia also is distinct from the other mood and anxiety disorders because it must be present for at least 2 years to be diagnosed and is thus thought to reflect, in part, chronic depressive personality characteristics [43]. We are thus inclined to believe that this association is a chance finding, particularly given the large number of tests conducted.

Several limitations of the current study warrant discussion. First, the study design was cross-sectional, which did not allow us to examine potential selection effects based on mental health status. Women with better mental health may be more likely to move to states that have more equitable laws and policies and that provide better opportunities for women. Such selection effects would inflate associations between state-level women’s status and psychiatric disorder prevalence. Second, although women’s status was assessed 2–3 years before psychiatric disorders it is possible that there is a latency period before these contextual factors impact risk for mental disorders. Finally, because information on women’s status is not updated on an annual basis for each state [18] we could not examine the impact of changes in women’s status on trends in psychiatric disorders over time. Prospective investigation of the effect of such changes on psychiatric disorder prevalence among women represents an important avenue for future research.

Despite these limitations, the current study has a number of strengths that extend the literature on social determinants of mental health outcomes in the US. To our knowledge, we provide the first empirical evaluation of the association between state-level women’s status and both the prevalence of psychiatric disorders among women and gender differences in prevalence in a national probability sample. Our findings suggest that state-level women’s status has no association with gender differences in mood and anxiety disorder prevalence and, with the exception of reproductive rights, is unrelated to disorder prevalence among women. Investigation of other social and contextual factors that play a role in shaping the distribution of individual- and community-level risk factors that are associated with gender differences in mental health outcomes, such as exposure to inter-personal violence, control over sexual decision-making and use of contraception, pay inequity, income, and educational attainment [1, 4, 34, 44–46], represents an important avenue for future research.