Abstract

Gender inequalities in health have been widely described, but few studies have examined the upstream sources of these inequalities in health. The objectives of this review are 1) to identify empirical papers that assessed the effect of gender equality policies on gender inequalities in health or on women's health by using between-country (or administrative units within a country) comparisons and 2) to provide an example of published evidence on the effects of a specific policy (parental leave) on women's health. We conducted a literature search covering the period from 1970 to 2012, using several bibliographical databases. We assessed 1,238 abstracts and selected 19 papers that considered gender equality policies, compared several countries or different states in 1 country, and analyzed at least 1 health outcome among women or compared between genders. To illustrate specific policy effects, we also selected articles that assessed associations between parental leave and women's health. Our review partially supports the hypothesis that Nordic social democratic welfare regimes and dual-earner family models best promote women's health. Meanwhile, enforcement of reproductive policies, mainly studied across US states, is associated with better mental health outcomes, although less with other outcomes. Longer paid maternity leave was also generally associated with better mental health and longer duration of breastfeeding.

INTRODUCTION

Gender is a strong determinant of most health outcomes (1, 2). Gendered inequalities in health reflect the fact that women suffer more than men from a host of nonfatal, disabling physical and mental illnesses and are generally expected to live fewer years in good health (3) despite having a higher life expectancy (4, 5). For example, in the US general population, women systematically report lower health-related quality of life than men across 6 different instruments (6) and a 26% higher probability of poor self-rated health even after adjustment by demographic factors (7), differences that have been shown to be coherent with other morbidity measures and not attributable to a greater tendency to report illness (8). Moreover, poor self-rated health has been associated with higher mortality (9).

The acquisition of gender roles and stereotypes starts early, with the socialization of girls and boys, continues throughout the life course, and results in gender inequalities in power and in the unequal division of paid and unpaid work. On the one hand, patriarchy, the systematic domination of women by men, restrains women's access to social and employment-related privileges and economic resources and assigns them a larger share of domestic responsibilities (unpaid work) with consequences to women's health status (10–12). Hegemonic masculinity, on the other hand, understood as the development and maintenance of a heterosexual male identity, promotes the taking of risks that are hazardous to health and contributes to premature mortality among men compared with women (13, 14). Thus, although biology plays a part in health differences between men and women, the higher burden of suffering is related mainly to social inequalities grounded in gender (2). “Gender” is a social construct based in culture-bound conventions, roles, and behaviors for, as well as in relationships between and among, women and men (15). In this paper, gender inequalities are considered unfair, unjust, avoidable, and unnecessary differences between women and men (15).

Several epidemiologic studies have shown that gender inequalities in health outcomes (e.g., mental health, pain disorders) can be explained by inequalities between men and women in some key social determinants of health, including 1) income (when differences in income by gender are consequent of limited access to labor markets, segregation, or gendered pay gaps, and, in the elderly, limited entitlement to pensions and economic independence results among women) (7, 16, 17); 2) paid work (when women have differential access to labor markets and, once working, their labor is less valued (horizontal segregation), wield less authority (vertical segregation), or face other different working conditions) (3, 18); and 3) unpaid work (when other inequalities in participation in society are justified by the gender ideology that attaches to women primary responsibility for labor performed in the household, including care for others) (19, 20).

Although rarely studied in epidemiologic research (21, 22), macrosocial determinants of health, such as political power, the welfare state, social protection policies, and economic and labor market policies, are major drivers of the social structure and power relations within society (23–25). These power relations generate social inequalities in health that researchers are now beginning to investigate, including their causes and the policies that increase or reduce them (26, 27).

As O'Campo and Dunn (21) stated, social epidemiology has focused on describing the existence of social inequalities and their risk factors (problem identification) but much less on the causes of these inequalities (problem-focused research) or on evidence of effective interventions to reduce them (solution-focused research). In this sense, although gender inequalities in health have been widely described, few studies have examined their upstream determinants including policies. Moreover, the analysis of policies from different countries may provide further information on the implementation of gender equality policies in different contexts and their effect on health. Therefore, the objectives of this review are 1) to identify empirical papers that assessed the effect of gender equality policies on gender inequalities in health or on women's health by using between-country (or administrative units within a country) comparisons and 2) to provide an example of published evidence on the effects of a specific policy (parental leave) on women's health. To contextualize our hypothesis, we first outline frameworks of gender equality policies and some specific policy types.

Policies to reduce gender inequalities

Policy is important in enforcing the promotion of equality between men and women in society. Gender policy regimes include multiple policies that intentionally or unintentionally target gender equality. Such regimes emerge from a particular understanding of gender relations, and they reinforce rules and norms governing men's and women's behavior (28). The majority of gender-regime typologies have emerged as a critique of Esping-Andersen's welfare state typology (29) and its “gender-blind” analysis. Esping-Andersen defined a welfare-state regime as the set of institutional arrangements, rules, and understandings that guides and shapes social policy decisions, expenditure developments, problem definitions, and even the demand-and-response structures connecting states with citizens and welfare consumers. Key concepts in his typology are 1) decommodification (the degree to which individuals or families can maintain relative economic autonomy from the market); 2) social stratification (how welfare states structure social class); and 3) state-market relations (the share of public and private welfare provision). On the basis of these dimensions, he describes 3 regimes: liberal (i.e., common in North America, the United Kingdom, and Ireland, that favors the role of the private market over the role of the government and the family in the provision of goods and services); conservative/corporativist; and social democratic (i.e., northern European countries where the role of the government is important). The main critique of Esping-Andersen by gender scholars is that gender relations are central to understanding inequalities in welfare states (28, 30–32). They argue that explicit attention must be paid to patriarchy and its effects on social services and labor markets, including policies and programs addressed to gender equality (32, 33). Later welfare-state scholarship has taken up this call (34–37). However, feminist academics have gone further to propose several gender-regime typologies (28, 30–33).

Korpi (36) and Korpi et al. (37) have categorized welfare states on the basis of the degree of support they offer to a dual-earner model. That is, to what extent do policy institutions encourage women's continuous labor force participation, enable parents (men and women) to combine parenthood with paid work, and attempt to redistribute caring work within the family? Social-democratic welfare state countries embrace this “dual-earner family model.” A second type of family model is the “market-oriented model,” encouraged mainly by liberal countries, that allows market forces to dominate how gender relations are shaped, leaving individuals to find private solutions based on their market resources and family support. The “traditional family policy model,” meanwhile, is found in conservative and southern European countries like Belgium, the Netherlands, Italy, and Spain, which offer high levels of support for the traditional family and low levels of dual-earner support. Reproduction work is thereby actively allocated to the family, and low support exists for female labor-force participation. Others authors have added the “contradictory model” to refer to countries that score high in support for both traditional and dual-earner models (38).

In the United States, much has been written about how welfare, which primarily supports lone mothers, was designed without regard for gender, since these programs were modeled after unemployment insurance policies that favor a male trajectory of climbing out of poverty by providing only temporary income replacement without additional support, such as child care benefits, needed by lone mothers (25, 39).

Specific policies related to gender equality are presented in Figure 1, which is slightly modified from the proposal of Pascall and Lewis (40), where they presented the policies for each social determinant of health (voice, paid work, incomes, care work, time), and here we present them by type of policies. In the figure, policies are identified across different levels of intervention (individual, household, community, and society). These policies are discussed below.

Figure 1.

Policies for gender equality. Modified from Pascall and Lewis (J Soc Policy. 2004;33(3):373–394) (40). Levels of policy intervention: 1, individual; 2, household; 3, civic/community; 4, society/politic.

Figure 1.

Policies for gender equality. Modified from Pascall and Lewis (J Soc Policy. 2004;33(3):373–394) (40). Levels of policy intervention: 1, individual; 2, household; 3, civic/community; 4, society/politic.

Policies to promote political representation

Women, compared with men, have fewer opportunities to be socially and politically active and to influence laws and politics. Because institutions and policies are more likely to favor the interests of those with more influence, men wielding that influence have less incentive to address the institutional constraints that feed gender inequalities in political representation (41). For example, in Spain, parliamentary initiatives to address gender-based violence have been more frequently promoted by women (42). Moreover, the difference in voice between men and women also exists on smaller scales, like communities, where men usually have more power. Some policies to increase women's participation in politics include proportional representation and quotas to reach greater representational parity (43, 44). Policies aimed at developing social movements that advocate for women's interests are also important.

Employment policies

Employment conditions are not the same for men and women. Men benefit from more opportunities for paid work. When women do enter the labor market, they suffer from horizontal and vertical segregation. Thus, labor market regulation is important for influencing gender inequalities in employment and equal pay (45). Gender equality policies should also promote equal access to quality jobs. In many cases, gendered employment policies seek to give men and women a flexible work time so they can reconcile paid work with family and civil-society activities. Yet these policies do not uniformly support women. For example, part-time work opportunities can be either a means of facilitating women's integration into paid work or a means of maintaining traditional divisions of labor (46).

Policies to promote equal incomes

Men and women have unequal access to income. This relates mainly to the fact that women have less access to paid work, worse paid jobs, lower salaries for the same jobs and, consequently, lower pensions. Therefore, policies to promote equal opportunities in earnings are a high priority (47). Other policies that have been instituted in this area include equal rights to benefits (e.g., pensions) for unpaid carers taking into account that pensions for women are lower as unpaid care and work are not taken into account nor are pensions appropriately split upon divorce (48).

Family policies

This group of policies seeks to increase family well-being and to promote reconciliation between paid work and family. Family support is a cross-cutting issue that may include support for employment, transport, food, education, and so on (49, 50). Family policies include regulations that promote gender equality in the areas of, for example, divorce, same-sex partnerships, and reproductive rights (e.g., birth control) and social policies toward, for example, child and elder care. Child-care policies include state care services (e.g., public child-care services) and benefits (e.g., child-related tax allowances) (40). Parental leave policies aim to secure both women's income during leave and their return to work while promoting mother and infant well-being. Paternity leave policies can further promote gender equality by facilitating greater participation of fathers in child care and of mothers in the labor market.

Policies to promote gender-equitable use of time

Women have less time than men because they bear a greater burden of unpaid work. Policies that promote less gender-biased uses of time are policies that produce time control in the household, to develop the different activities. Anti-gender discrimination policies for the labor market are useful (47), but part-time work should have equal value to full time work. Other policies include school scheduling to match work times or services related to the household (e.g., food markets), leisure time sports facilities offered at different hours, and so on.

The policies described above target existing gender inequalities in power, income, paid and unpaid work, and use of time. The alleviation of these gender inequalities, in turn, impacts health, especially for women, through mechanisms such as a reduction in stress, discrimination, violence, financial difficulties, poverty, double burden of work, or time pressure. For example, women who have a higher burden of household unpaid work have worse health outcomes than men (19, 20, 51). Therefore, family and employment policies that reduce the burden and time pressure on women can reduce these gendered health inequalities.

Based on the policies described above, the hypothesis of our review was that European social democratic welfare regimes and countries with family dual-earner family models best promote women's health and reduce gender inequalities in health due to the support they provide for men and women's participation in paid as well as unpaid work and sharing of family care work. Moreover, the specific gender equality policies described above and summarized in Figure 1 will promote improved health outcomes.

MATERIALS AND METHODS

We conducted a literature search using the following databases: Medline; Web of Science (Science Citation Index and Social Sciences Citation Index); Gender Studies Database; ProQuest databases (Applied Social Sciences Index and Abstracts (ASSIA), International Bibliography of the Social Sciences (IBSS), Public Affairs Information Service (PAIS) International, ProQuest Dissertations & Theses (PQDT), and Sociological Abstracts); PolicyFile; Embase (via Ovid); and Evidence-Based Medicine (EBM) Reviews. The search was performed in December 2012 and included papers written in English, French, Spanish, Portuguese, or Italian and published from 1970 to the search date. Subject headings and keywords included in the search are listed in Appendix Table 1, although the search was slightly modified depending on the database used; for example, gender terms were omitted when searching the Gender Studies Database. Terms were generally combined with the Boolean operator AND, that is, Policy terms AND Gender terms AND Health terms. These searches were developed in consultation with, and carried out by, a medical librarian trained in performing systematic reviews. The search produced a total of 1,238 unduplicated studies (464 in Medline, 221 in Web of Science, 142 in the Gender Studies Database, 180 in the ProQuest databases, 47 in Policy File, 132 in Embase, 1 in EBM Reviews, and 155 from related citations in PubMed). This total number was divided among the coauthors so that each abstract was reviewed by 2 authors; ties were broken by a third author's review.

Inclusion criteria for an abstract to be considered further were that the study should consider gender equality policies, gender regimes, or welfare state regimes; compare policies across countries, states, or macro-administrative units within 1 country; analyze at least 1 health outcome or health behavior (e.g., alcohol consumption); and examine outcomes among women or compare them between genders.

On the basis of these criteria, a total of 48 full-text papers were reviewed, from which 15 were finally selected that really met the criteria defined above. A review of the references cited in these 15 papers yielded 4 more papers that fulfilled the selection criteria. Thus, 19 papers entered the study. It is worth mentioning that there were numerous articles documenting gender inequalities in political, social, and economic determinants at the area level and their relationship with health outcomes. However, as they did not refer to policies specifically, we did not include them.

To accomplish our second review objective, from the same set of abstracts we also selected articles that focused specifically on parental leave in relation to women's health status or health behavior. We identified 10 articles focused on parental leave, including 1 systematic review. We retrieved 7 more articles through a review of the references of the 10 articles and author searches based on the names of the authors that were references in the 10 papers.

RESULTS

Results are structured in 2 sections. In the first section, we introduce the papers found in relation to the first objective, which was to examine the influence of macrosocial policies on women's health or gender inequalities in health. In the following section, we discuss the papers found on the specific policy chosen, namely, the relationship between parental leave and women's health outcomes.

The influence of macrosocial policies on women's health or gender inequalities in health

Table 1 shows the main characteristics of the 19 papers selected to meet our first objective. Most were published after the year 2000. Six papers analyzed 2–4 countries, 4 analyzed between 13 and 22 countries, and the other 9 compared US states. The majority of papers used an individual design only (8 papers) or individual design with a multilevel approach (7 papers), and 4 used an ecological design where data on policies and indicators of health outcomes at the state level are correlated. The policies analyzed included country classifications regarding welfare state policies (3 papers), family policy models (3 papers), and gender regimes (1 paper), but also more specific policies, such as toward reproductive rights (7 papers). The most often analyzed outcome was self-rated health (7 papers). Only 4 papers analyzed gender inequalities, while 6 of them focused on women's health; the other papers focused on women as mothers (4 papers) or on the health of men and women separately, without analyzing gender inequalities (3 papers).

Table 1.

Description of the Papers Included in the Review

 No. 
Year of publication   
 1980–1989 10.5 
 1990–1999 5.3 
 2000–2009 10 52.6 
 2010–2012 31.6 
No. of countries   
 1 (states within the country compared) 47.4 
 2–4 31.6 
 5–25 21.0 
Unit of analysis   
 Individual 42.1 
 Ecological 21.1 
 Multilevela 36.8 
Policies   
 Welfare state 15.8 
 Family policy model 15.8 
 Gender regime 5.3 
 Reproductive health policy 36.8 
 Others 26.3 
Outcomes   
 Self-rated health (alone or with LLI) 36.8 
 Mortality 15.8 
 Health behaviors 10.5 
 Variables related to mental health 15.8 
 Others 21.1 
Comparison groups   
 Men and women (gap) 21.1 
 Men and women (separate) 15.8 
 Women 31.6 
 Lone and couple mothers 21.1 
 Couples 10.5 
Total 19 100.0 
 No. 
Year of publication   
 1980–1989 10.5 
 1990–1999 5.3 
 2000–2009 10 52.6 
 2010–2012 31.6 
No. of countries   
 1 (states within the country compared) 47.4 
 2–4 31.6 
 5–25 21.0 
Unit of analysis   
 Individual 42.1 
 Ecological 21.1 
 Multilevela 36.8 
Policies   
 Welfare state 15.8 
 Family policy model 15.8 
 Gender regime 5.3 
 Reproductive health policy 36.8 
 Others 26.3 
Outcomes   
 Self-rated health (alone or with LLI) 36.8 
 Mortality 15.8 
 Health behaviors 10.5 
 Variables related to mental health 15.8 
 Others 21.1 
Comparison groups   
 Men and women (gap) 21.1 
 Men and women (separate) 15.8 
 Women 31.6 
 Lone and couple mothers 21.1 
 Couples 10.5 
Total 19 100.0 

Abbreviation: LLI, limited long-standing illness.

a In general, individuals and countries, but in one case, countries and years.

Welfare state policies

The 3 papers analyzing welfare state policies (52–54) are presented in Table 2. All of these had an individual cross-sectional design. The number of countries analyzed ranged from 2 to 13 and included Britain and Finland (which appeared in all the papers), 11 other European countries (which appeared in a sole paper), and Japan. Two of the papers analyzed self-rated health (in 1 case together with limited long-standing illness), and 1 studied mental health. One of the papers assessed the gap between men and women, another analyzed men and women separately, and 1 studied women only. Bambra et al. (52), comparing the self-assessed health among men and women living under various types of welfare regimes, found that women in social democratic and southern welfare states (quite far apart in terms of welfare-state arrangements) are more likely to report worse health than men, although this does not seem to be the case in countries classified as corporatist or liberal (52). Meanwhile, Chandola et al. (53), after analyzing the relationship between work-family conflicts and mental health, concluded that Finland's family-friendly work policies are contributing to lower levels of conflict and better mental health among Finnish working women compared with working women in Japan. Finally, Lahelma et al. (54) analyzed health status by family type (couple mothers with children and couples without children, divorced lone mothers, single lone mothers, single women) and found that lone mothers had worse health than other women, probably explainable via the multiple attachment hypothesis (having financial problems and less social support). Comparing Britain and Finland, they found that a greater part of the health disadvantage of lone mothers in Britain was explained by structural characteristics.

Table 2.

Studies Examining Welfare State Regimes or Policies Within These Regimes

First Author, Year (Reference No.) Study Aim Design, Unit of Analysis, and Type of Areas Policy Outcome Groups Examined Other Variables Key Findings 
Bambra, 2009 (52To examine the association between gender and self-assessed health in different welfare-state regimes Individual cross-sectional Welfare-state regimes: social democratic (Denmark, Norway, Sweden, Finland, Holland); corporatist (Belgium, France, Germany); liberal (England, Ireland); and southern (Italy, Spain, Portugal) Self-rated health Gap between men and women Educational status is used as a stratification variable Women in the social democratic and southern welfare states were more likely to report worse self-assessed health than men. In the corporatist and liberal (Ireland) countries, there were no gender differences in self-assessed health. In Finland and England, men had poorer self-assessed health than women. 
13 European countries 
Chandola, 2004 (53To examine whether work-to-family and family-to-work conflicts explain the effect of role strain on mental health among men and women public employees of different welfare states Individual cross-sectional Welfare-state policies Mental health (SF-36 mental health component) Men and women (separate) Multiple roles as a measure of role strain, work-to-family conflict, family-to-work conflict scales Work-to-family and family-to-work conflicts explain the effect of role strain on mental health in the 3 countries. The family-friendly work policies in Finland may contribute to its lower levels of both types of conflicts and better mental health, especially for women. The poorer mental health of working Japanese women could be attributable to their higher levels of both types of conflicts. 
2 cities and 1 area: London (United Kingdom), Helsinki (Finland), and the West Coast of Japan 
Lahelma, 2002 (54To examine whether the multiple-burden hypothesis and the multiple-attachment hypothesis explain health variations between British and Finnish women and also to assess if variations are due to socioeconomic status, educational level, and income Individual cross-sectional Welfare-state policies Self-rated health and limiting long-standing illness Women according to family type Family type (couple with children, couple without children, divorced lone mother, single lone mother, single), employment status, education, income Women living in 2-parent families with children had better health than women living in other family structures. In Britain, the disadvantaged social position of lone mothers accounts for a greater proportion of their poor health than in Finland. 
2 countries: Britain and Finland 
First Author, Year (Reference No.) Study Aim Design, Unit of Analysis, and Type of Areas Policy Outcome Groups Examined Other Variables Key Findings 
Bambra, 2009 (52To examine the association between gender and self-assessed health in different welfare-state regimes Individual cross-sectional Welfare-state regimes: social democratic (Denmark, Norway, Sweden, Finland, Holland); corporatist (Belgium, France, Germany); liberal (England, Ireland); and southern (Italy, Spain, Portugal) Self-rated health Gap between men and women Educational status is used as a stratification variable Women in the social democratic and southern welfare states were more likely to report worse self-assessed health than men. In the corporatist and liberal (Ireland) countries, there were no gender differences in self-assessed health. In Finland and England, men had poorer self-assessed health than women. 
13 European countries 
Chandola, 2004 (53To examine whether work-to-family and family-to-work conflicts explain the effect of role strain on mental health among men and women public employees of different welfare states Individual cross-sectional Welfare-state policies Mental health (SF-36 mental health component) Men and women (separate) Multiple roles as a measure of role strain, work-to-family conflict, family-to-work conflict scales Work-to-family and family-to-work conflicts explain the effect of role strain on mental health in the 3 countries. The family-friendly work policies in Finland may contribute to its lower levels of both types of conflicts and better mental health, especially for women. The poorer mental health of working Japanese women could be attributable to their higher levels of both types of conflicts. 
2 cities and 1 area: London (United Kingdom), Helsinki (Finland), and the West Coast of Japan 
Lahelma, 2002 (54To examine whether the multiple-burden hypothesis and the multiple-attachment hypothesis explain health variations between British and Finnish women and also to assess if variations are due to socioeconomic status, educational level, and income Individual cross-sectional Welfare-state policies Self-rated health and limiting long-standing illness Women according to family type Family type (couple with children, couple without children, divorced lone mother, single lone mother, single), employment status, education, income Women living in 2-parent families with children had better health than women living in other family structures. In Britain, the disadvantaged social position of lone mothers accounts for a greater proportion of their poor health than in Finland. 
2 countries: Britain and Finland 

Abbreviation:SF-36, 36-item Medical Outcomes Study Short-Form General Health Survey.

Gender regimes and family policy models

Table 3 describes the 4 papers examining gender regimes or family policy models (55–58). The only paper on gender regimes analyzed longitudinal data with a multilevel approach: First-level units were years, and second-level units were 22 countries of the Organization for Economic Cooperation and Development (OECD). The other 3 papers were cross-sectional with an individual basis. Two of these papers analyzed 3 countries (Italy, Sweden, and Britain), while the other examined 18 European countries. Outcomes under consideration were mortality, self-rated health (accompanied, in 1 of the 2 cases, by limited long-standing illness), and well-being. One paper studied the gap in mortality between men and women, 1 paper analyzed men and women separately, and the other 2 compared lone and couple mothers.

Table 3.

Studies Examining Gender Regimes or Family Policy Models

First Author, Year (Reference No.) Study Aim Design, Unit of Analysis, and Type of Areas Policy Outcome Groups Examined Other Variables Key Findings 
Backhans, 2012 (55To examine the association between gender policy and the gender gap in external causes and circulatory disease mortality Multilevel longitudinal, with years at level 1 and countries at level 2 (22 OECD countries) Gender regimes: male breadwinner, earner-carer, compensatory breadwinner, and universal citizen. Specific policies also analyzed External cause and circulatory disease mortality Gap between men and women Gender equality measure, gross domestic product, Gini coefficient, health behaviors Earner-carer cluster and policies characteristic of that cluster are associated with smaller gender differences in external causes, because of an association with increased female mortality. 
Boye, 2011 (56To examine associations between well-being and paid work and housework in different European family policy models Cross-sectional analysis of individuals in 18 European countries Korpi typologies of family policy models: dual earner (Denmark, Finland, Norway, Sweden), traditional (Austria, Belgium, France, Germany, Netherlands), market oriented (Ireland, Switzerland, United Kingdom), contradictory (dual earner: Estonia, Slovenia; traditional: Czech Republic, Slovakia; and contradictory: Hungary, Poland) Well-being index Men and women (separate) Paid work hours, housework hours, work–family conflict The traditional family policy model shows the most positive association between the women's well-being and paid working hours, when work–family conflict is taken into account, and also with housework hours. The market-oriented model shows women's paid working hours and housework hours are instead associated with decreasing well-being. 
Burström, 2010 (57To examine welfare state arrangements and social policy, living conditions, and health among lone and couple mothers in 3 contrasting policy environments Cross-sectional analysis of individuals Family policy model: dual earner (Sweden), market oriented (United Kingdom), and traditional (Italy) Self-rated health and limiting long-standing illness (only Sweden and Britain) Gap between lone and couple mothers Socioeconomic position based on occupation, low-income household, unemployment Italian lone mothers were older than their Swedish and English counterparts and less prevalent. Lone mothers had worse health than couple mothers in all countries. The dual-earner family policy model, together with the generosity of family benefits, has protected Swedish mothers in general and lone mothers in particular from poverty. 
3 countries: Italy, Sweden, and Britain 
Fritzell, 2012 (58To analyze the relationship between self-rated health and nonemployment and their potential synergistic effects among lone and couple mothers in different family policy categories Cross-sectional analysis of individuals Family policy model: dual earner (Sweden), market oriented (Britain), and traditional (Italy) Self-rated health Gap between lone and couple mothers Employment status, socioeconomic characteristics, smoking Lone motherhood was associated with increased risk of poor health in Britain and Sweden but less in Italy. Unemployment only marginally contributed to the excess risk of poor health. 
3 countries: Italy, Sweden, and Britain 
First Author, Year (Reference No.) Study Aim Design, Unit of Analysis, and Type of Areas Policy Outcome Groups Examined Other Variables Key Findings 
Backhans, 2012 (55To examine the association between gender policy and the gender gap in external causes and circulatory disease mortality Multilevel longitudinal, with years at level 1 and countries at level 2 (22 OECD countries) Gender regimes: male breadwinner, earner-carer, compensatory breadwinner, and universal citizen. Specific policies also analyzed External cause and circulatory disease mortality Gap between men and women Gender equality measure, gross domestic product, Gini coefficient, health behaviors Earner-carer cluster and policies characteristic of that cluster are associated with smaller gender differences in external causes, because of an association with increased female mortality. 
Boye, 2011 (56To examine associations between well-being and paid work and housework in different European family policy models Cross-sectional analysis of individuals in 18 European countries Korpi typologies of family policy models: dual earner (Denmark, Finland, Norway, Sweden), traditional (Austria, Belgium, France, Germany, Netherlands), market oriented (Ireland, Switzerland, United Kingdom), contradictory (dual earner: Estonia, Slovenia; traditional: Czech Republic, Slovakia; and contradictory: Hungary, Poland) Well-being index Men and women (separate) Paid work hours, housework hours, work–family conflict The traditional family policy model shows the most positive association between the women's well-being and paid working hours, when work–family conflict is taken into account, and also with housework hours. The market-oriented model shows women's paid working hours and housework hours are instead associated with decreasing well-being. 
Burström, 2010 (57To examine welfare state arrangements and social policy, living conditions, and health among lone and couple mothers in 3 contrasting policy environments Cross-sectional analysis of individuals Family policy model: dual earner (Sweden), market oriented (United Kingdom), and traditional (Italy) Self-rated health and limiting long-standing illness (only Sweden and Britain) Gap between lone and couple mothers Socioeconomic position based on occupation, low-income household, unemployment Italian lone mothers were older than their Swedish and English counterparts and less prevalent. Lone mothers had worse health than couple mothers in all countries. The dual-earner family policy model, together with the generosity of family benefits, has protected Swedish mothers in general and lone mothers in particular from poverty. 
3 countries: Italy, Sweden, and Britain 
Fritzell, 2012 (58To analyze the relationship between self-rated health and nonemployment and their potential synergistic effects among lone and couple mothers in different family policy categories Cross-sectional analysis of individuals Family policy model: dual earner (Sweden), market oriented (Britain), and traditional (Italy) Self-rated health Gap between lone and couple mothers Employment status, socioeconomic characteristics, smoking Lone motherhood was associated with increased risk of poor health in Britain and Sweden but less in Italy. Unemployment only marginally contributed to the excess risk of poor health. 
3 countries: Italy, Sweden, and Britain 

Abbreviations: Gini coefficient, the most commonly used measure of inequality in the distribution of family income; OECD, Organization for Economic Cooperation and Development.

The paper examining gender regimes analyzing the data of 22 countries showed that the most gender-egalitarian welfare-state approach, called the earner-carer model, presents smaller gender differences in mortality by external causes, mainly due to an association with increased female mortality (55). Boye (56) analyzed 18 countries and showed that a traditional family policy model presents the most positive association between women's well-being and both paid and housework hours, while in the market-oriented model, women's paid and housework hours are associated with decreasing well-being, the former association appearing to be caused by work-family conflict. The papers that compared the health of lone versus couple mothers showed that, in general, the health of lone mothers was worse than that of couple mothers, and that this was true regardless of the family policy model (57, 58). However, 1 paper (58) found that this association was lower for Italy, which operates a traditional family policy model. In addition, 1 study found that the prevalence of lone motherhood and the characteristics of that population (e.g., teen mothers vs. older widowed mothers), joblessness rates, poverty, and health status all varied across countries with different family policy models (57), and that Sweden's dual-earner model, together with its generous family benefits, protected mothers in general and lone mothers in particular from the degree of poverty experienced elsewhere.

Reproductive policies

The 7 papers that examined reproductive policies (59–65) are presented in Table 4. One of these, considering 22 countries worldwide and using a multilevel approach, analyzed the effect of an index of reproductive autonomy and the context of violence against women on the gap in frequency of drinking between men and women. The remaining 6 papers compared the 50 US states. These were either ecological or individual analyses using a multilevel approach. They used indices measuring reproductive rights, and the common policies and indicators considered in these indices covered several dimensions: support by governors and legislatures for bans or restrictions on abortion; state funding for abortion; state-mandated coverage for contraception and infertility treatment; access to abortion for minors without parental notification; mandatory waiting periods for abortions; the percentage of women living in counties with at least 1 abortion provider; legality of same-sex couple adoption; mandatory sexual education; prochoice government in the state; and state requirements for private insurers to cover Papanicolaou and cervical cancer screening and chlamydia screening. One of the papers also analyzed policies in the economic, access to care, and community domains (65).

Table 4.

Studies Examining Reproductive Health Policies

First Author, Year (Reference No.) Study Aim Design, Unit of Analysis, and Type of Areas Policy Outcome Groups Examined Other Variables Key Findings 
Bond, 2010 (59To examine the association between country-level gender equality and differences in the frequency of women's compared with men's drinking Multilevel cross-sectional, with individuals at level 1 and countries at level 2 Index of women's reproductive autonomy (including abortion laws, maternity leave) and context of violence against women (including legislation and other activities to address violence against women) Frequency of drinking in public and private settings Gap between men and women Gender equality at the country level: gender empowerment measure, the GGI, GGI education, and GGI political participation subindices Neither of the 2 indices predicts gender differences in the frequency of drinking in public or private settings in the model adjusted by gross domestic product; however, in nonadjusted models, more reproductive autonomy and less violence against women predicted smaller gender differences in drinking in public settings. 
22 countries in Europe (n = 8), the Americas (n = 7), Asia (n = 3), Oceania (n = 2), and Africa (n = 2) Economic status: gross domestic product and human development index of the country 
Age, gender, and marital status at the individual level 
Chen, 2005 (60To analyze the association between state-level women's status and women's depressive symptoms Multilevel cross-sectional, with women of reproductive age at level 1 and states at level 2 Composite index measuring reproductive rights Depressive symptomatology (CES-D Scale) Women 3 other indices measuring women's political participation, employment, and earnings, as well as economic autonomy at the state level. Gini coefficient. Individual variables Women living in states guaranteeing more reproductive rights had lower levels of depressive symptoms. Indices of employment and earnings, as well as economic autonomy, were also associated with lower levels of depressive symptoms. 
The 50 US states 
Jun, 2004 (61To examine the association between state-level women's status and women's health Multilevel cross-sectional, with women at level 1 and states at level 2 Composite index measuring reproductive rights Self-rated health Women 3 other indices measuring women's political participation, employment, and earnings, as well as economic autonomy. State median income, income ratios of top and bottom 20%. Individual variables Reproductive rights were not associated with self-rated health. 
The 50 US states 
Kawachi, 1999 (62To analyze the association between state-level women's status and women's and men's mortality and morbidity rates Ecological, cross-sectional Composite index measuring reproductive rights Total mortality rates, cause-specific death rates, mean days of activity limitations, infant mortality Men and women (separate) for total mortality rates; women for the cause-specific rates and limited activity 3 other indices measuring women's political participation, employment and earnings, and economic autonomy. Gini coefficient, median household income, and poverty rates The reproductive rights index was the only index that was not associated with women's total mortality or any specific cause of death except infant mortality. Associations were not found with activity limitations. 
The 50 US states 
McLaughlin, 2011 (63To analyze the associations between state-level women's status and women's mood and anxiety disorders and gender inequalities in the presence of the disorders Multilevel cross-sectional, with women at level 1 and states at level 2 Composite index measuring reproductive rights Presence of mood and anxiety disorders Women and gap between men and women 3 other indices measuring women's political participation, employment and earnings, and social/economic autonomy. Individual variables Women living in states with greater reproductive rights had lower 12-month prevalences of any mood disorder and specifically major depression and also of 1 anxiety disorder (post–traumatic stress). 
The 50 US states 
       Gender inequalities were not associated with reproductive rights. Other indices were not related with mood and anxiety disorders. 
Roberts, 2012 (64To examine the associations between state-level women's status and alcohol consumption by women and men Multilevel cross-sectional, with women at level 1 and states at level 2 Composite indices measuring policies of violence against women and reproductive rights 5 alcohol consumption measures Men and women (separate) 3 other indices measuring women's political participation, women's socioeconomic status, and gender equality in socioeconomic status. Income inequality and religion. Individual variables. Nearly all associations between reproductive rights or violence policy and alcohol consumption were either negative or nonsignificant in adjusted models. All other associations between gender equality and alcohol consumption were either negative or nonsignificant. 
The 50 US states 
Wisdom, 2005 (65To examine the associations between state-level policies related to women's health and women's health outcomes Ecological, cross-sectional Policies in 4 domains: reproduction, economic security, access to care (Medicaid eligibility and efforts to expand Medicaid), and community (environmental health and violence against women) Mortality rates for heart disease, lung cancer, stroke, and breast cancer, as well as infant mortality Women Percentages of population African American and also Hispanic, percentage of women in poverty, and average earnings Access to care and community were most strongly associated with health outcomes. Reproductive issues were related to breast cancer and infant mortality. 
The 50 US states Mental health Better mental health was associated with policies related to violence against women, economic security, and family planning. 
First Author, Year (Reference No.) Study Aim Design, Unit of Analysis, and Type of Areas Policy Outcome Groups Examined Other Variables Key Findings 
Bond, 2010 (59To examine the association between country-level gender equality and differences in the frequency of women's compared with men's drinking Multilevel cross-sectional, with individuals at level 1 and countries at level 2 Index of women's reproductive autonomy (including abortion laws, maternity leave) and context of violence against women (including legislation and other activities to address violence against women) Frequency of drinking in public and private settings Gap between men and women Gender equality at the country level: gender empowerment measure, the GGI, GGI education, and GGI political participation subindices Neither of the 2 indices predicts gender differences in the frequency of drinking in public or private settings in the model adjusted by gross domestic product; however, in nonadjusted models, more reproductive autonomy and less violence against women predicted smaller gender differences in drinking in public settings. 
22 countries in Europe (n = 8), the Americas (n = 7), Asia (n = 3), Oceania (n = 2), and Africa (n = 2) Economic status: gross domestic product and human development index of the country 
Age, gender, and marital status at the individual level 
Chen, 2005 (60To analyze the association between state-level women's status and women's depressive symptoms Multilevel cross-sectional, with women of reproductive age at level 1 and states at level 2 Composite index measuring reproductive rights Depressive symptomatology (CES-D Scale) Women 3 other indices measuring women's political participation, employment, and earnings, as well as economic autonomy at the state level. Gini coefficient. Individual variables Women living in states guaranteeing more reproductive rights had lower levels of depressive symptoms. Indices of employment and earnings, as well as economic autonomy, were also associated with lower levels of depressive symptoms. 
The 50 US states 
Jun, 2004 (61To examine the association between state-level women's status and women's health Multilevel cross-sectional, with women at level 1 and states at level 2 Composite index measuring reproductive rights Self-rated health Women 3 other indices measuring women's political participation, employment, and earnings, as well as economic autonomy. State median income, income ratios of top and bottom 20%. Individual variables Reproductive rights were not associated with self-rated health. 
The 50 US states 
Kawachi, 1999 (62To analyze the association between state-level women's status and women's and men's mortality and morbidity rates Ecological, cross-sectional Composite index measuring reproductive rights Total mortality rates, cause-specific death rates, mean days of activity limitations, infant mortality Men and women (separate) for total mortality rates; women for the cause-specific rates and limited activity 3 other indices measuring women's political participation, employment and earnings, and economic autonomy. Gini coefficient, median household income, and poverty rates The reproductive rights index was the only index that was not associated with women's total mortality or any specific cause of death except infant mortality. Associations were not found with activity limitations. 
The 50 US states 
McLaughlin, 2011 (63To analyze the associations between state-level women's status and women's mood and anxiety disorders and gender inequalities in the presence of the disorders Multilevel cross-sectional, with women at level 1 and states at level 2 Composite index measuring reproductive rights Presence of mood and anxiety disorders Women and gap between men and women 3 other indices measuring women's political participation, employment and earnings, and social/economic autonomy. Individual variables Women living in states with greater reproductive rights had lower 12-month prevalences of any mood disorder and specifically major depression and also of 1 anxiety disorder (post–traumatic stress). 
The 50 US states 
       Gender inequalities were not associated with reproductive rights. Other indices were not related with mood and anxiety disorders. 
Roberts, 2012 (64To examine the associations between state-level women's status and alcohol consumption by women and men Multilevel cross-sectional, with women at level 1 and states at level 2 Composite indices measuring policies of violence against women and reproductive rights 5 alcohol consumption measures Men and women (separate) 3 other indices measuring women's political participation, women's socioeconomic status, and gender equality in socioeconomic status. Income inequality and religion. Individual variables. Nearly all associations between reproductive rights or violence policy and alcohol consumption were either negative or nonsignificant in adjusted models. All other associations between gender equality and alcohol consumption were either negative or nonsignificant. 
The 50 US states 
Wisdom, 2005 (65To examine the associations between state-level policies related to women's health and women's health outcomes Ecological, cross-sectional Policies in 4 domains: reproduction, economic security, access to care (Medicaid eligibility and efforts to expand Medicaid), and community (environmental health and violence against women) Mortality rates for heart disease, lung cancer, stroke, and breast cancer, as well as infant mortality Women Percentages of population African American and also Hispanic, percentage of women in poverty, and average earnings Access to care and community were most strongly associated with health outcomes. Reproductive issues were related to breast cancer and infant mortality. 
The 50 US states Mental health Better mental health was associated with policies related to violence against women, economic security, and family planning. 

Abbreviations: CES-D,Center for Epidemiologic Studies Depression; GGI, Global Gender Gap Index; Gini coefficient, the most commonly used measure of inequality in the distribution of family income.

Diverse outcomes were assessed by this set of papers, including depressive symptoms, self-rated health, mortality, and alcohol consumption, and they were mainly focused on women's health. Results showed that policies related to reproductive rights were not associated with women's self-rated health or mortality (61, 62), although 1 paper found an association with breast cancer and infant mortality (65). On the other hand, an association was found between such policies and lower levels of women's depressive symptoms (60), prevalence of major depression and posttraumatic stress (63), and lower levels of alcohol consumption among men and women (64). Moreover, study participants in US states with more family-planning policies had fewer days described as “not good” in terms of mental health (65). However, the paper that compared countries worldwide did not find an association between reproductive rights or context of violence and the gap in drinking between men and women in private settings, and in public settings the associations were found only when gross domestic product was not included in the statistical models (59). It is worth mentioning that the majority of these studies also analyzed indices of gender equality. They found an association between these indices and better health outcomes. For example, in the drinking study, the indices predicted smaller gender differences in drinking, mainly in public settings (59).

Other policies

The last table, Table 5, shows the papers in our review that analyzed an amalgam of different policies and policy contexts (66–70), including the following: social transfers received by mothers; policy frameworks targeting lone mothers; policies in 6 different domains (economic security, access to health insurance and services, protection for the insured, health education and supplies, antidiscrimination, and smoking); and the legal dimension of the “Status of Women Index,” which includes laws related to equality in paid and unpaid work and rape conviction and which is combined with 3 other dimensions of the Status of Women Index. The outcomes under analysis were also diverse, including self-rated health and long-standing limiting illness, as well as obesity, high blood pressure, diabetes and smoking, and violence against women. Two papers compared 2 countries (Canada and Norway; Britain and Sweden), and the other 3 analyzed US states.

Table 5.

Studies Examining Miscellaneous Policies

First Author, Year (Reference No.) Study Aim Design, Unit of Analysis, and Type of Areas Policy Outcome Groups Examined Other Variables Key Findings 
Curtis, 2004 (66To investigate the potential health consequences of giving Canadian mothers the social transfers available in Norway Individual cross-sectional Probability of receiving transfers (for Canadian mothers; all Norwegian mothers receive them) and level of transfers received Self-rated health Gap between lone and couple mothers Household income, percentage of income in form of transfers, employment status. Other: marital status, age, number of children, maternal education, smoking, exercise Health status of lone mothers is worse than that of married mothers in Canada but not in Norway. Giving Canadian mothers Norwegian transfers would provide significant gains in socioeconomic and health status of lone mothers. 
2 countries: Canada and Norway 
Whitehead, 2000 (67To perform a comparative analysis of social policies and their points of potential impact on the pathways leading from lone motherhood to ill health in Britain and Sweden Individual cross-sectional Policy context in relation to lone parents Self-rated health, long-standing limiting illness Gap between lone and couple mothers Poverty, employment status, age, ethnicity Lone mothers had poorer health than couple mothers, and the difference is similar in Sweden and Britain. However, more favorable social policies in Sweden have protected lone mothers from poverty and insecurity in the labor market to a higher degree. 
2 countries: Britain and Sweden 
Wisdom, 2008 (68To examine the association between state-level policies related to women's health and obesity, diabetes, high blood pressure, and smoking in women Ecological cross-sectional Policies in 6 domains: economic security, access to health insurance and services, protection for the insured, health education and supplies, antidiscrimination, smoking Obesity, high blood pressure, diabetes, and smoking Women Percentage of population African American and also percentage of Hispanic, percentage of women in poverty, average earnings In general, state-level policies were associated with the outcomes. Antidiscrimination policies were associated with 3 outcomes: high blood pressure, smoking, and obesity. 
The 50 US states 
Yllö, 1983 (69To analyze the association between the status of women and intimate partner violence beating in each American state Ecological cross-sectional Legal dimension of the Status of Women Index (laws for equal rights for men and women) Proportion of couples who indicated that the husband had used violence against his wife (and also wife against husband) Individuals living with a person of the opposite sex 3 other dimensions of the Status of Women Index: economic, educational, and political. Violent crime, urbanization, education, state per capita income The correlation between legal status and violence against women was negative and almost statistically significant. In addition, the index of women's status had a curvilinear relationship with violence, which increased where the index was lowest and highest. 
30 US states 
Yllö, 1984 (70To examine whether structured sexual inequality affects the relationship between marital inequality and intimate partner violence Multilevel cross-sectional with individuals at level 1 and states at level 2 Legal dimension of the Status of Women Index (laws for equal rights for men and women) Proportion of couples who indicated that the husband had used violence against his wife Individuals living with a person of the opposite sex 3 other dimensions of the Status of Women Index: economic, educational, and political. Type of couple: wife dominant, separate, shared, husband dominant Violence against wives is most severe in families residing in high-status states but in which the husband dominates decision making. Also, wife beating seems to be high in wife-dominated couples residing in low-status states. 
30 US states 
First Author, Year (Reference No.) Study Aim Design, Unit of Analysis, and Type of Areas Policy Outcome Groups Examined Other Variables Key Findings 
Curtis, 2004 (66To investigate the potential health consequences of giving Canadian mothers the social transfers available in Norway Individual cross-sectional Probability of receiving transfers (for Canadian mothers; all Norwegian mothers receive them) and level of transfers received Self-rated health Gap between lone and couple mothers Household income, percentage of income in form of transfers, employment status. Other: marital status, age, number of children, maternal education, smoking, exercise Health status of lone mothers is worse than that of married mothers in Canada but not in Norway. Giving Canadian mothers Norwegian transfers would provide significant gains in socioeconomic and health status of lone mothers. 
2 countries: Canada and Norway 
Whitehead, 2000 (67To perform a comparative analysis of social policies and their points of potential impact on the pathways leading from lone motherhood to ill health in Britain and Sweden Individual cross-sectional Policy context in relation to lone parents Self-rated health, long-standing limiting illness Gap between lone and couple mothers Poverty, employment status, age, ethnicity Lone mothers had poorer health than couple mothers, and the difference is similar in Sweden and Britain. However, more favorable social policies in Sweden have protected lone mothers from poverty and insecurity in the labor market to a higher degree. 
2 countries: Britain and Sweden 
Wisdom, 2008 (68To examine the association between state-level policies related to women's health and obesity, diabetes, high blood pressure, and smoking in women Ecological cross-sectional Policies in 6 domains: economic security, access to health insurance and services, protection for the insured, health education and supplies, antidiscrimination, smoking Obesity, high blood pressure, diabetes, and smoking Women Percentage of population African American and also percentage of Hispanic, percentage of women in poverty, average earnings In general, state-level policies were associated with the outcomes. Antidiscrimination policies were associated with 3 outcomes: high blood pressure, smoking, and obesity. 
The 50 US states 
Yllö, 1983 (69To analyze the association between the status of women and intimate partner violence beating in each American state Ecological cross-sectional Legal dimension of the Status of Women Index (laws for equal rights for men and women) Proportion of couples who indicated that the husband had used violence against his wife (and also wife against husband) Individuals living with a person of the opposite sex 3 other dimensions of the Status of Women Index: economic, educational, and political. Violent crime, urbanization, education, state per capita income The correlation between legal status and violence against women was negative and almost statistically significant. In addition, the index of women's status had a curvilinear relationship with violence, which increased where the index was lowest and highest. 
30 US states 
Yllö, 1984 (70To examine whether structured sexual inequality affects the relationship between marital inequality and intimate partner violence Multilevel cross-sectional with individuals at level 1 and states at level 2 Legal dimension of the Status of Women Index (laws for equal rights for men and women) Proportion of couples who indicated that the husband had used violence against his wife Individuals living with a person of the opposite sex 3 other dimensions of the Status of Women Index: economic, educational, and political. Type of couple: wife dominant, separate, shared, husband dominant Violence against wives is most severe in families residing in high-status states but in which the husband dominates decision making. Also, wife beating seems to be high in wife-dominated couples residing in low-status states. 
30 US states 

One of the papers analyzing lone and couple mothers found that the health status of lone mothers was worse than that of married mothers in Canada but not in Norway, and that giving Canadian mothers transfers similar to those offered in Norway would provide significant gains in both the socioeconomic status and the health status of lone mothers (66). Meanwhile, Whitehead et al. (67) found that the health of lone mothers was worse than that of couple mothers in both Sweden and Britain, although the pathways leading from lone motherhood to poor health varied between these countries: Sweden's more favorable social policies have protected lone mothers from poverty and insecurity in relation to the labor market to a much greater degree. The remaining 3 papers were conducted in the US states. Results showed that antidiscrimination policies (including domestic violence and sexual assault policies) were related to decreased high blood pressure, smoking, and obesity among women (68). In addition, the above-mentioned legal dimension of the Status of Women Index was negatively correlated with violence in women, and the Status of Women Index had a curvilinear relationship (violence increases in states with a low and high index) (69). Moreover, in another study, an interaction took place between gender inequality at the state level (also using the Status of Women Index), inequality in the couple, and violence against wives (70).

A specific example: parental leave and women's health

In this section, we summarize the papers on the effects of the length of parental leave on maternal health or health behaviors (71–87). While most countries have some sort of parental leave policy (88), countries (and jurisdictions within countries) differ in the length of this leave. As a consequence, researchers have focused mainly on length of leave rather than on its presence or absence.

Research into mental health in the postpartum period is extensive. With regard to the relationship between this issue and maternity leave, findings from cross-sectional (76, 79), as well as longitudinal (71, 78, 80), studies consistently associate longer leave with reductions in depressive symptoms. In the United States, Chatterji and Markowitz (71) found that taking more than 12 weeks of maternity leave would reduce maternal depressive symptoms on the Centre for Epidemiologic Studies Depression Scale by 15%. In addition, extending paid leave over 8 weeks would increase the probability of reporting excellent health by 3.5 percentage points. Similar findings regarding postpartum depressive symptoms were obtained by using data from the 1988 National Maternal and Infant Health Survey (79). However, a study based on the Wisconsin Maternity Leave and Health Project did not observe a relationship between length of leave and mental health (84).

It is important to note that leave lengths are shorter in the United States than in Canada and most European countries. This raises questions about nonlinear effects and the impact of optimal parental leave length on various outcomes in mothers, fathers, and children. For example, McGovern et al. (76) found that, in the United States, maternal mental health had a U-shaped association with time off work and that the positive effect of time off work on mental health was observed to begin at 15 weeks postpartum, when only 15% of mothers were still on maternity leave.

Findings regarding maternity leave in relation to overall maternal health status are less consistent. Self-reported health status has also been shown to improve with longer maternity leave (71), but no associations were found with other measures, such as outpatient physician visits (79), clinical symptoms (81), backache, tiredness, or lack of sleep (82). A study evaluating the impact of the policies extending the parental leave in Canada in 2000 found that, following the reform, the health of women and depression did not change (75).

The effects of maternity leave may also vary according to maternal characteristics. Chatterji and Markowitz (71) noted that the positive effects of maternity leave were more pronounced among mothers who returned to work full-time and who were married and of non-Latino white ethnicity. Fathers' leave may also impact on maternal health. Chatterji and Markowitz (71) found that married mothers whose spouses did not take any paternal leave had higher depressive symptoms than those whose spouses did.

Three Swedish studies that assessed gender inequality in the couples' total share of parental leave found that women who took less than 20% of the couple's total leave (i.e., at least 4 times less than their partners) were at higher risk of mortality and sickness absence compared with women in equal-couples relationships (40%–60% of total leave), but more traditional couples (women took the majority of the leave) also had increased risks (73). However, another study did not find associations for women when analyzing sickness absence and gender inequality in the private sphere, including sharing of leave (85). No association was found with alcohol harm for women, but fathers who took paternity leave ran a lower risk (74).

Two studies considered the effects of antenatal leave on delivery outcomes and found that antenatal leave and its duration reduce the odds of delivering by cesarean section and preterm delivery (77), as well as the presence of an intervention or complication during delivery (83).

Breastfeeding is known to be associated with better maternal and infant outcomes (89). In our search, a literature review found that longer maternity leave has been consistently associated with duration of breastfeeding in 4 studies in the United States (72). More recent studies confirm these findings. A longitudinal study (87) found a dose-response association between length of maternity leave and duration of breastfeeding. The study from Canada found that, following the reform of extending parental leave, the average length of breastfeeding increased over a month for eligible mothers and that the proportion of women attaining 6 months of exclusive breastfeeding increased (75).

DISCUSSION

The current body of evidence comparing country- or state-level gender policy regimes and women's health or gender inequalities in health is small and limited by methodological challenges. The studies we reviewed cover several policies, ranging from analysis of state-level reproductive policies and violence-protection policies to welfare gender regimes and family support at the country level, covering only a small spectrum of gender equality policies described in the Introduction. Yet, with a few notable exceptions, the weight of evidence in this growing area of research suggests that policies intended to explicitly support women and families are associated with a variety of improved health outcomes for women or diminished gender inequalities in health. Our hypothesis that Nordic social democratic welfare regimes and dual-earner family models promote mostly women's health was partially supported by our review, especially regarding the health of mothers, although gender inequalities in self-rated health and a higher mortality risk for women were also found in these countries. The enforcement of reproductive policies, an issue mainly studied in the US states, shows associations with better mental health outcomes, although not with other outcomes. Taking longer paid maternity leave was also generally associated with better mental health and longer duration of breastfeeding, although cross-country comparisons were lacking in this field.

Several studies reveal better health among women in Nordic countries (53, 54, 57). This supports the literature emphasizing that it is the “dual-earner policy model”—the model supported by Scandinavian welfare states (37, 90)—that leads to better women's health outcomes, as opposed to models that support the family in general (36, 38). In dual-earner models, public policies support women's participation in the labor force and enforce more equitable sharing of responsibilities for unpaid work (caring and reproductive work within the family). For example, our review has shown that these models have protected lone mothers from poverty (57, 66, 67). Reduction in poverty, in turn, is a powerful promoter of health. However, it is worth noting that women still have the larger burden for unpaid work compared with men in these countries along with other gender inequalities, such as low political participation among women, gender-segregated employment, and lower pay for the same jobs (91).

Although we found just 1 article using a gender regime typology (55), its results highlight the benefits of using this approach or family policies models, as opposed to the broad category of welfare state regimes, prevalent in this field of inquiry, when studying women's health (92, 93). Thus, our results also give credibility to the critiques of “black box” welfare state typologies that neglect policies, specifically gender policies (92, 93), and further support the benefits of linking welfare regime theory to specific gender policy instruments to understand their effect on women's health. The consideration of other welfare state types (Asian, east European, Latin, US) represents a future challenge to the heuristics of these classifications with regard to the impact of welfare state type upon women's health.

Many studies from the United States examined state-level policies and also compared the relationship between variations in support for gender equality or women's well-being and rates of adverse outcomes, mostly using ecological analyses, with some studies taking a multilevel approach. Almost all of the US studies included here utilized national samples, such as the Behavioral Risk Factor Survey or the National Maternal Infant Health Survey, and linked these to state-level databases, such as the National Report Card on States, which evaluates policies supporting women, or to composite indices impacting women's status published by the Institute for Women's Policy Research. Across a range of outcomes (depressive symptoms, chronic conditions, mortality), gender equality or women's autonomy had positive impacts on women's health status, with a few exceptions. Three studies examining mental well-being provided contrasting findings. The studies examining depressive symptoms using the Center for Epidemiologic Studies Depression Scale (60) and the Behavioral Risk Factor Survey reports of number of days that mental health was “not good” (65) were influenced by state-level equality indices in the predicted direction. However, data on mood and anxiety disorders from the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC) reported by McLaughlin et al. (63) were not associated with state-level indices of women's political participation, employment and earnings, or social/economic autonomy. Two US-based studies examined partner violence as an outcome (69, 70). These studies used data collected 3 decades before the most recent studies and may not be comparable to the rest, in part because of fewer public services available to victims at that time, as well as a completely different policy context from more contemporary studies.

In several, but not all, of the state-level studies that examined associations between reproductive policies and health outcomes, these associations were not established. Health outcomes evaluated in these studies ranged from self-rated health, depressive symptoms, and alcohol consumption to mortality. However, the 3 papers concerning mental health outcomes found positive associations analyzing depressive symptomatology (60), mood and anxiety disorders (63), and the number of days when mental health was “not good” (65). In these cases, mental health was better in the states guaranteeing better reproductive rights. A possible explanation of these findings can be that these policies can strengthen women's autonomy in their reproductive decisions and therefore improve women's life opportunities and achieve a better standard of living (60, 63). The literature on the impact on health of reproductive policies has an abundance of studies examining infant and birth outcomes that were not included here because this was a study of women's health. That literature, however, may shed further light on whether and how reproductive policies impact population well-being.

With regard to the specific example of parental leave, evidence to date generally presents a positive portrait of its effect on maternal mental health and breastfeeding, while the evidence on physical health is less conclusive. However, not all the papers reviewed included a specific policy; they mainly compared length of maternal leave. Variability in the length of leave may be determined in part by policies at the state, county, or workplace levels. The literature on the effects of parental leave and maternal health is incipient, and, as such, the few studies were rather exploratory and based on self-reported measures. With a few exceptions (71, 87), longitudinal studies were conducted in the United States with small samples ranging from 141 to 436 mothers (78, 80, 81). Longitudinal designs with larger and representative samples and in diverse settings would be beneficial. We have not identified research using multilevel analyses to study parental leave, although these are most appropriate to draw inferences involving multiple units of analyses. Studies modeling nonlinear effects of length may help to establish the optimal length and inform decision making. A limitation of this literature is that we could find no studies analyzing leave among same-sex couples.

A variety of study designs were used in the studies reviewed. Earlier studies tended to use ecological designs (e.g., correlating state-level differences in the United States with women's health outcomes), which are appropriate for generating hypotheses about how and why policies might impact upon population well-being. More recent studies use multilevel models, usually with secondary data, which are stronger in design as the policies and health outcomes are included in the statistical models at the appropriate “levels” and sample sizes are larger than those for ecological studies. In turn, hierarchical models take into account the similarity and thus dependence of observations within a group (e.g., country), and their estimations are more precise than those of individual models using ordinary least-squares regression (94). However, in this review, we did not find systematic differences in the results according to the type of approach used, but, as said before, the type of policies and outcomes analyzed are very heterogeneous. Ideal designs would include longitudinal studies spanning multiple years (i.e., before and after policy changes) to enable the study of lag times required for a policy to impact population well-being, but, unfortunately, only 1 article used longitudinal data, and it analyzed the evolution of gender inequalities in different gender regimes (55).

The tendency in the majority of articles is still to document associations, with few examining pathways or mechanisms by which policies impact health and well-being. One study, by Burström et al. (57), proposed hypotheses to conceptualize various pathways of influence for different types of policies on health and well-being and tested each pathway accordingly. The use of theory and testable hypotheses to guide examination of specific pathways as this article illustrated would strengthen the current evidence. Several studies also examined the impact of gender equality–promoting policies on men's health, as well as women's, or on subgroups of women, like lone mothers. This approach strengthens the evidence by showing whether policy impacts are specific to women or general to the population. However, only 4 studies (52, 55, 59, 63) analyzed gender inequalities in health. Although gender relations affect these inequalities via power inequalities, none of these studies took this into account. Meanwhile, only 1 study in the review (55) examined the relationship between regimes favoring gender equality (“earner-carer”) and the gender gap. Moreover, the intersectionality with other axes of inequality, such as social class or race, should also be emphasized, because policies should have different impacts on populations of different social classes or races (95).

Recent evidence is also emerging on the lag time of the impact of macrolevel polices (96), suggesting that future research should consider the time needed for impacts on population well-being to be felt. This is a gap in existing studies. In the area of parental leave, self-selection and longer benefits may be associated with unmeasured characteristics that confer protection for adverse outcomes: Future work could benefit from comparisons across countries with different policies.

Finally, as noted in Figure 1, the number and types of policies that influence gender equality are much broader than those covered in studies thus far. Future research should use strong study designs to capture the full range of macrolevel policies on gender inequalities, as well as on women's and men's health.

ACKNOWLEDGMENTS

Author affiliations: Health Information Systems Unit, Agència de Salut Pública de Barcelona, Barcelona, Spain (Carme Borrell, Laia Palència, Davide Malmusi); CIBER de Epidemiología y Salud Pública (CIBERESP), Madrid, Spain (Carme Borrell, Laia Palència, Davide Malmusi); Institut d'Investigació Biomèdica de Sant Pau (IIB Sant Pau), Barcelona, Spain (Carme Borrell, Laia Palència, Davide Malmusi); Department of Experimental and Health Sciences, Universitat Pompeu Fabra, Barcelona, Spain (Carme Borrell); Bloomberg Faculty of Nursing, University of Toronto, Toronto, Canada (Carles Muntaner); Center for Research in Inner City Health, Keenan Research Center, Li Ka Shing Knowledge Institute, St. Michael's Hospital, Toronto, Canada (Carles Muntaner, Marcelo Urquía, Patricia O'Campo); Divisions of Social and Behavioral Sciences and Global Health, Dalla Lana School of Public Health, University of Toronto, Toronto, Canada (Carles Muntaner, Marcelo Urquía, Patricia O'Campo); and Department of Population, Family, and Reproductive Health, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland (Patricia O'Campo).

The authors form part of the research team of the “Evaluating the impact of structural policies on health inequalities and their social determinants and fostering change (Sophie)” project.

This research was supported by the European Community's Seventh Framework Programme (FP7/2007–2013, grant agreement 278173): “Evaluating the impact of structural policies on health inequalities and their social determinants and fostering change (Sophie)” project; and by the Plan Nacional de Investigación Científica, Desarrollo e Innovación Tecnológica (I+D+I) and the Instituto de Salud Carlos III-Subdirección General de Evaluación y Fomento de la Investigación (grant PI12/03114).

We thank Dr. Carolyn Ziegler, the medical librarian who collaborated in the literature search. We are also grateful to Dr. Lucía Artazcoz for her comments on a previous version of the manuscript.

We will present this review to the meeting of the Spanish Society of Epidemiology, Granada, Spain, in September 2013.

Conflict of interest: none declared.

REFERENCES

1
Krieger
N
Genders, sexes, and health: what are the connections—and why does it matter?
Int J Epidemiol
 , 
2003
, vol. 
32
 
4
(pg. 
652
-
657
)
2
Sen
G
Östlin
P
Gender inequity in health: why it exists and how we can change it
Glob Public Health
 , 
2008
, vol. 
3
 
suppl 1
(pg. 
1
-
12
)
3
Verbrugge
LM
Gender and health: an update on hypotheses and evidence
J Health Soc Behav
 , 
1985
, vol. 
26
 
3
(pg. 
156
-
182
)
4
Glei
DA
Horiuchi
S
The narrowing sex differential in life expectancy in high-income populations: effects of differences in the age pattern of mortality
Popul Stud (Camb)
 , 
2007
, vol. 
61
 
2
(pg. 
141
-
159
)
5
Wang
H
Dwyer-Lindgren
L
Lofgren
KT
, et al.  . 
Age-specific and sex-specific mortality in 187 countries, 1970–2010: a systematic analysis for the Global Burden of Disease Study 2010
Lancet
 , 
2013
, vol. 
380
 
9859
(pg. 
2071
-
2094
)
6
Fryback
DG
Dunham
NC
Palta
M
, et al.  . 
US norms for six generic health-related quality-of-life indexes from the National Health Measurement Study
Med Care
 , 
2007
, vol. 
45
 
12
pg. 
1162
 
7
Prus
SG
Comparing social determinants of self-rated health across the United States and Canada
Soc Sci Med
 , 
2011
, vol. 
73
 
1
(pg. 
50
-
59
)
8
Malmusi
D
Artazcoz
L
Benach
J
, et al.  . 
Perception or real illness? How chronic conditions contribute to gender inequalities in self-rated health
Eur J Public Health
 , 
2012
, vol. 
22
 
6
(pg. 
781
-
786
)
9
DeSalvo
KB
Bloser
N
Reynolds
K
, et al.  . 
Mortality prediction with a single general self-rated health question
J Gen Intern Med
 , 
2006
, vol. 
21
 
3
(pg. 
267
-
275
)
10
Doyal
L
What Makes Women Sick: Gender and the Political Economy of Health
 , 
1995
London, United Kingdom
Macmillan Press, Ltd
11
Ross
CE
Bird
CE
Sex stratification and health lifestyle: consequences for men's and women's perceived health
J Health Soc Behav
 , 
1994
, vol. 
35
 
2
(pg. 
161
-
178
)
12
Bird
CE
Fremont
AM
Gender, time use, and health
J Health Soc Behav
 , 
1991
, vol. 
32
 
2
(pg. 
114
-
129
)
13
Doyal
L
Sex, gender, and health: the need for a new approach
BMJ
 , 
2001
, vol. 
323
 
7320
(pg. 
1061
-
1063
)
14
Stanistreet
D
Bambra
C
Scott-Samuel
A
Is patriarchy the source of men's higher mortality?
J Epidemiol Community Health
 , 
2005
, vol. 
59
 
10
(pg. 
873
-
876
)
15
Krieger
N
A glossary for social epidemiology
J Epidemiol Community Health
 , 
2001
, vol. 
55
 
10
(pg. 
693
-
700
)
16
Cherepanov
D
Palta
M
Fryback
DG
, et al.  . 
Gender differences in health-related quality-of-life are partly explained by sociodemographic and socioeconomic variation between adult men and women in the US: evidence from four US nationally representative data sets
Qual Life Res
 , 
2010
, vol. 
19
 
8
(pg. 
1115
-
1124
)
17
Roy
K
Chaudhuri
A
Influence of socioeconomic status, wealth and financial empowerment on gender differences in health and healthcare utilization in later life: evidence from India
Soc Sci Med
 , 
2008
, vol. 
66
 
9
(pg. 
1951
-
1962
)
18
Hosseinpoor
AR
Stewart Williams
J
Amin
A
, et al.  . 
Social determinants of self-reported health in women and men: understanding the role of gender in population health
PLoS One
 , 
2012
, vol. 
7
 
4
pg. 
e34799
 
19
Artazcoz
L
Borrell
C
Benach
J
Gender inequalities in health among workers: the relation with family demands
J Epidemiol Community Health
 , 
2001
, vol. 
55
 
9
(pg. 
639
-
647
)
20
Harryson
L
Novo
M
Hammarström
A
Is gender inequality in the domestic sphere associated with psychological distress among women and men? Results from the Northern Swedish Cohort
J Epidemiol Community Health
 , 
2012
, vol. 
66
 
3
(pg. 
271
-
276
)
21
O'Campo
P
Dunn
JR
Rethinking Social Epidemiology: Towards a Science of Change
 , 
2011
London, United Kingdom
Springer
22
Putnam
S
Galea
S
Epidemiology and the macrosocial determinants of health
J Public Health Policy
 , 
2008
, vol. 
29
 
3
(pg. 
275
-
289
)
23
Commission on Social Determinants of Health
Closing the Gap in a Generation: Health Equity Through Action on the Social Determinants of Health
 , 
2008
Geneva, Switzerland
World Health Organization
 
Final Report
24
Commission to Reduce Social Inequalities in Health in Spain. A proposal of policies and interventions to reduce social inequalities in health in Spain [in Spanish].
Gac Sanit
 , 
2012
, vol. 
26
 
2
(pg. 
182
-
189
)
25
O'Campo
P
Rojas-Smith
L
Welfare reform and women's health: review of the literature and implications for state policy
J Public Health Policy
 , 
1998
, vol. 
19
 
4
(pg. 
420
-
446
)
26
CDC health disparities and inequalities report—United States, 2011
MMWR
 , 
2011
, vol. 
60
 
suppl
(pg. 
1
-
109
)
27
Marmot
M
Allen
J
Goldblatt
P
, et al.  . 
Fair Society, Healthy Lives: Strategic Review of Health Inequalities in England post-2010
 , 
2010
London, United Kingdom
The Marmot Review
28
Sainsbury
D
Gender and Welfare State Regimes
 , 
1999
New York, NY
Oxford University Press
29
Esping-Andersen
G
The Three Worlds of Welfare Capitalism
 , 
1990
Cambridge, United Kingdom
Polity Press
30
Orloff
AS
Gender and the social rights of citizenship: the comparative analysis of gender relations and welfare states
Am Sociol Rev
 , 
1993
, vol. 
58
 (pg. 
303
-
328
)
31
Orloff
AS
Gendering the comparative analysis of welfare states: an unfinished agenda
Sociological Theory
 , 
2009
, vol. 
27
 
3
(pg. 
317
-
343
)
32
Lewis
J
Gender and the development of welfare regimes
J Eur Soc Policy
 , 
1992
, vol. 
2
 
3
(pg. 
159
-
173
)
33
O'Connor
JS
Kennet
P
Gender, citizenship and welfare state regimes
Handbook of Comparative Social Policy
 , 
2004
Cheltenham, United Kingdom:
Edward Elgar Publishing, Limited
34
Esping-Andersen
G
Social Foundations of Postindustrial Economies
 , 
1999
New York, NY
Oxford University Press
35
Huber
E
Stephens
JD
Development and Crisis of the Welfare State: Parties and Policies in Global Markets
 , 
2001
Chicago, IL
University of Chicago Press
36
Korpi
W
Faces of inequality: gender, class, and patterns of inequalities in different types of welfare states
Soc Pol
 , 
2000
, vol. 
7
 
2
(pg. 
127
-
191
)
37
Korpi
W
Ferrarini
T
Englund
S
Women's opportunities under different constellations of family policies in Western countries
2010
Luxembourg, Luxembourg
Luxembourg Income Study (LIS);
 
A comparative analysis. In: Luxembourg Income Study Working Paper Series (No. 556)
38
Ferrarini
T
Families, States and Labour Markets: Institutions, Causes and Consequences of Family Policy in Post-war Welfare States
 , 
2006
Cheltenham, United Kingdom
Edward Elgar Publishing
39
Gordon
L
Women, the State, and Welfare
 , 
1990
Madison, WI
University of Wisconsin Press
40
Pascall
G
Lewis
J
Emerging gender regimes and policies for gender equality in a wider Europe
J Soc Policy
 , 
2004
, vol. 
33
 
3
(pg. 
373
-
394
)
41
The World Development Report 2012: Gender Equality and Development
 , 
2012
Washington, DC
The World Bank
42
Vives-Cases
C
Gil-González
D
Carrasco-Portiño
M
, et al.  . 
Gender-based violence on the agenda of the Spanish Parliament (1979–2004) [in Spanish]
Gac Sanit
 , 
2006
, vol. 
20
 
2
(pg. 
142
-
148
)
43
Ballington
J
Karam
AM
Women in Parliament: Beyond Numbers
 , 
2005
Stockholm, Sweden
International Idea
44
Verge
T
Gendering representation in Spain: opportunities and limits of gender quotas
J Women Politics Policy
 , 
2010
, vol. 
31
 
2
(pg. 
166
-
190
)
45
Rubery
J
Towards a gendering of the labour market regulation debate
Cambridge J Econ
 , 
2011
, vol. 
35
 
6
(pg. 
1103
-
1126
)
46
Stier
H
Lewin-Epstein
N
Women's part-time employment and gender inequality in the family
J Fam Issues
 , 
2000
, vol. 
21
 
3
(pg. 
390
-
410
)
47
Duncan
S
Policy discourses on ‘reconciling work and life’ in the EU
Social Policy and Society
 , 
2002
, vol. 
1
 
4
(pg. 
305
-
314
)
48
Muller
C
Volkov
O
Older women: work and caregiving in conflict? A study of four countries
Soc Work Health Care
 , 
2009
, vol. 
48
 
7
(pg. 
665
-
695
)
49
Gauthier
AH
Family policies in industrialized countries: is there convergence?
Population
 , 
2002
, vol. 
57
 
3
(pg. 
447
-
474
)
50
Kuronen
M
Family Platform. Research on Families and Family Policies in Europe. State of the Art
 , 
2010
Jyväskylän, Finland
Family Research Centre
 
Final Report
51
Arcas
MM
Novoa
AM
Artazcoz
L
Gender inequalities in the association between demands of family and domestic life and health in Spanish workers
Eur J Public Health
  
[published online ahead of print August 8, 2012] (doi:10.1093/eurpub/cks095)
52
Bambra
C
Pope
D
Swami
V
, et al.  . 
Gender, health inequalities and welfare state regimes: a cross-national study of 13 European countries
J Epidemiol Community Health
 , 
2009
, vol. 
63
 
1
(pg. 
38
-
44
)
53
Chandola
T
Martikainen
P
Bartley
M
, et al.  . 
Does conflict between home and work explain the effect of multiple roles on mental health? A comparative study of Finland, Japan, and the UK
Int J Epidemiol
 , 
2004
, vol. 
33
 
4
(pg. 
884
-
893
)
54
Lahelma
E
Arber
S
Kivela
K
, et al.  . 
Multiple roles and health among British and Finnish women: the influence of socioeconomic circumstances
Soc Sci Med
 , 
2002
, vol. 
54
 
5
(pg. 
727
-
740
)
55
Backhans
M
Burström
B
Leon
AP
, et al.  . 
Is gender policy related to the gender gap in external cause and circulatory disease mortality? A mixed effects model of 22 OECD countries 1973–2008
BMC Public Health
 , 
2012
, vol. 
12
 
1
pg. 
969
 
56
Boye
K
Work and well-being in a comparative perspective—the role of family policy
Eur Sociol Rev
 , 
2011
, vol. 
27
 
1
(pg. 
16
-
30
)
57
Burström
B
Whitehead
M
Clayton
S
, et al.  . 
Health inequalities between lone and couple mothers and policy under different welfare regimes—the example of Italy, Sweden and Britain
Soc Sci Med
 , 
2010
, vol. 
70
 
6
(pg. 
912
-
920
)
58
Fritzell
S
Vannoni
F
Whitehead
M
, et al.  . 
Does non-employment contribute to the health disadvantage among lone mothers in Britain, Italy and Sweden? Synergy effects and the meaning of family policy
Health Place
 , 
2012
, vol. 
18
 
2
(pg. 
199
-
208
)
59
Bond
JC
Roberts
SC
Greenfield
TK
, et al.  . 
Gender differences in public and private drinking contexts: a multi-level GENACIS analysis
Int J Environ Res Public Health
 , 
2010
, vol. 
7
 
5
(pg. 
2136
-
2160
)
60
Chen
YY
Subramanian
SV
Acevedo-Garcia
D
, et al.  . 
Women's status and depressive symptoms: a multilevel analysis
Soc Sci Med
 , 
2005
, vol. 
60
 
1
(pg. 
49
-
60
)
61
Jun
HJ
Subramanian
SV
Gortmaker
S
, et al.  . 
A multilevel analysis of women's status and self-rated health in the United States
J Am Med Womens Assoc
 , 
2004
, vol. 
59
 
3
(pg. 
172
-
180
)
62
Kawachi
I
Kennedy
BP
Gupta
V
, et al.  . 
Women's status and the health of women and men: a view from the States
Soc Sci Med
 , 
1999
, vol. 
48
 
1
(pg. 
21
-
32
)
63
McLaughlin
KA
Xuan
Z
Subramanian
SV
, et al.  . 
State-level women's status and psychiatric disorders among US women
Soc Psychiatry Psychiatr Epidemiol
 , 
2011
, vol. 
46
 
11
(pg. 
1161
-
1171
)
64
Roberts
SC
Macro-level gender equality and alcohol consumption: a multi-level analysis across U.S. States
Soc Sci Med
 , 
2012
, vol. 
75
 
1
(pg. 
60
-
68
)
65
Wisdom
JP
Berlin
M
Lapidus
JA
Relating health policy to women's health outcomes
Soc Sci Med
 , 
2005
, vol. 
61
 
8
(pg. 
1776
-
1784
)
66
Curtis
L
Phipps
S
Social transfers and the health status of mothers in Norway and Canada
Soc Sci Med
 , 
2004
, vol. 
58
 
12
(pg. 
2499
-
2507
)
67
Whitehead
M
Burström
B
Diderichsen
F
Social policies and the pathways to inequalities in health: a comparative analysis of lone mothers in Britain and Sweden
Soc Sci Med
 , 
2000
, vol. 
50
 
2
(pg. 
255
-
270
)
68
Wisdom
JP
Michael
YL
Ramsey
K
, et al.  . 
Women's health policies associated with obesity, diabetes, high blood pressure, and smoking: a follow-up on the women's health report card
Women Health
 , 
2008
, vol. 
48
 
1
(pg. 
103
-
122
)
69
Yllö
K
Sexual equality and violence against wives in American states
J Comp Fam Stud
 , 
1983
, vol. 
14
 
1
(pg. 
67
-
86
)
70
Yllö
K
The status of women, marital equality, and violence against wives a contextual analysis
J Fam Issues
 , 
1984
, vol. 
5
 
3
(pg. 
307
-
320
)
71
Chatterji
P
Markowitz
S
Family leave after childbirth and the mental health of new mothers
J Ment Health Policy Econ
 , 
2012
, vol. 
15
 
2
(pg. 
61
-
76
)
72
Staehelin
K
Bertea
PC
Stutz
EZ
Length of maternity leave and health of mother and child—a review
Int J Public Health
 , 
2007
, vol. 
52
 
4
(pg. 
202
-
209
)
73
Månsdotter
A
Lindholm
L
Lundberg
M
, et al.  . 
Parental share in public and domestic spheres: a population study on gender equality, death, and sickness
J Epidemiol Community Health
 , 
2006
, vol. 
60
 
7
(pg. 
616
-
620
)
74
Månsdotter
A
Lundberg
M
Lindholm
L
How does gender equality progress link to alcohol care and death? A registry study of the Swedish parental cohort of 1988/1989
J Public Health Policy
 , 
2012
, vol. 
33
 
1
(pg. 
105
-
118
)
75
Baker
M
Milligan
K
Maternal employment, breastfeeding, and health: evidence from maternity leave mandates
J Health Econ
 , 
2008
, vol. 
27
 
4
(pg. 
871
-
887
)
76
McGovern
P
Dowd
B
Gjerdingen
D
, et al.  . 
Time off work and the postpartum health of employed women
Med Care
 , 
1997
, vol. 
35
 
5
(pg. 
507
-
521
)
77
Guendelman
S
Pearl
M
Graham
S
, et al.  . 
Maternity leave in the ninth month of pregnancy and birth outcomes among working women
Womens Health Issues
 , 
2009
, vol. 
19
 
1
(pg. 
30
-
37
)
78
Gjerdingen
DK
Chaloner
KM
The relationship of women's postpartum mental health to employment, childbirth, and social support
J Fam Pract
 , 
1994
, vol. 
38
 
5
(pg. 
465
-
472
)
79
Chatterji
P
Markowitz
S
Does the length of maternity leave affect maternal health?
South Econ J
 , 
2005
, vol. 
72
 
1
(pg. 
16
-
41
)
80
Hyde
JS
Klein
MH
Essex
MJ
, et al.  . 
Maternity leave and womens mental-health
Psychol Women Q
 , 
1995
, vol. 
19
 
2
(pg. 
257
-
285
)
81
Killien
DMG
Habermann
B
Jarrett
M
Influence of employment characteristics on postpartum mothers' health
Women Health
 , 
2001
, vol. 
33
 
1-2
(pg. 
63
-
81
)
82
Romito
P
Saurel-Cubizolles
MJ
Cuttini
M
et al
Mothers' health after the birth of the first child
Women Health
 , 
1994
, vol. 
21
 
2-3
(pg. 
1
-
22
)
83
Xu
Q
Seguin
L
Goulet
L
Beneficial effect of maternity leave on delivery
Can J Public Health
 , 
2002
, vol. 
93
 
1
(pg. 
72
-
77
)
84
Klein
MH
Hyde
JS
Essex
MJ
, et al.  . 
Maternity leave, role quality, work involvement, and mental health one year after delivery
Psychol Women Q
 , 
1998
, vol. 
22
 
2
(pg. 
239
-
266
)
85
Backhans
MC
Burström
B
Lindholm
L
, et al.  . 
Pioneers and laggards—is the effect of gender equality on health dependent on context?
Soc Sci Med
 , 
2009
, vol. 
68
 
8
(pg. 
1388
-
1395
)
86
Visness
CM
Kennedy
KI
Maternal employment and breast-feeding: findings from the 1988 National Maternal and Infant Health Survey
Am J Public Health
 , 
1997
, vol. 
87
 
6
(pg. 
945
-
950
)
87
Ogbuanu
C
Glover
S
Probst
J
, et al.  . 
The effect of maternity leave length and time of return to work on breastfeeding
Pediatrics
 , 
2011
, vol. 
127
 
6
(pg. 
e1414
-
e1427
)
88
Heymann
J
Raub
A
Earle
A
Creating and using new data sources to analyze the relationship between social policy and global health: the case of maternal leave
Public Health Rep
 , 
2011
, vol. 
126
 
suppl 3
(pg. 
127
-
134
)
89
León-Cava
N
Lutter
C
Ross
J
, et al.  . 
Quantifying the Benefits of Breastfeeding: A Summary of the Evidence
 , 
2002
Washington, DC
Pan American Health Organization
90
Esping-Andersen
G
Incomplete Revolution: Adapting Welfare States to Women's New Roles
 , 
2009
Cambridge, United Kingdom
Polity Press
91
Hausmann
R
Tyson
LD
Zahidi
S
The Global Gender Gap Report 2012
 , 
2012
Geneva, Switzerland
World Economic Forum
92
Lundberg
O
Yngwe
MA
Stjärne
MK
, et al.  . 
The role of welfare state principles and generosity in social policy programmes for public health: an international comparative study
Lancet
 , 
2008
, vol. 
372
 
9650
(pg. 
1633
-
1640
)
93
Espelt
A
Borrell
C
Rodríguez-Sanz
M
, et al.  . 
Inequalities in health by social class dimensions in European countries of different political traditions
Int J Epidemiol
 , 
2008
, vol. 
37
 
5
(pg. 
1095
-
1105
)
94
Osborne
JW
Advantages of hierarchical linear modeling
Pract Assess Res Eval
 , 
2000
, vol. 
7
 
1
(pg. 
1
-
3
)
95
Weber
L
Schulz
AJ
Mullings
L
Reconstructing the landscape of health disparities research: promoting dialogue and collaboration between feminist intersectional and biomedical paradigm
Gender, Race, Class, and Health: Intersectional Approaches
 , 
2006
San Francisco, CA
Jossey-Bass
96
Zheng
H
Do people die from income inequality of a decade ago?
Soc Sci Med
 , 
2012
, vol. 
75
 
1
(pg. 
36
-
45
)

Appendix Table 1.

Search Terms in the Literature Research

Policy Gender Health 
“aid to families”, “cash transfer*”, “child care policy”, “employment assistance”, “employment insurance”, “family allowance”, “family benefit”, “family benefits”, “family policy”, “food stamps”, “gender policies”, “gender policy”, “gender regime”, “health policy”, “housing subsid*”, “maternity leave”, “old age assistance”, “parental leave”, “pension*”, “policy”, “public assistance”, “public housing”, “public preschool”, “social assistance”, “social housing”, “social insurance”, “social program*”, “social protection”, “social protection policies”, “social protection policy”, “social security”, “social transfers”, “unemployment insurance”, “welfare”, “welfare program”, “welfare state” “female”, “gender”, “mothers”, “sex differences”, “women”, “sex factors”, “maternal”, “women's health” “epidemiology”, “equality”, “equity”, “health”, “mental health”, “outcomes”, “quality of life”, “health status indicators”, “health disparities”, “health surveys”, “life expectancy” 
“gender gap”, “gender equality”, “gender equity”, “gender disparity”, “gender inequality” 
Policy Gender Health 
“aid to families”, “cash transfer*”, “child care policy”, “employment assistance”, “employment insurance”, “family allowance”, “family benefit”, “family benefits”, “family policy”, “food stamps”, “gender policies”, “gender policy”, “gender regime”, “health policy”, “housing subsid*”, “maternity leave”, “old age assistance”, “parental leave”, “pension*”, “policy”, “public assistance”, “public housing”, “public preschool”, “social assistance”, “social housing”, “social insurance”, “social program*”, “social protection”, “social protection policies”, “social protection policy”, “social security”, “social transfers”, “unemployment insurance”, “welfare”, “welfare program”, “welfare state” “female”, “gender”, “mothers”, “sex differences”, “women”, “sex factors”, “maternal”, “women's health” “epidemiology”, “equality”, “equity”, “health”, “mental health”, “outcomes”, “quality of life”, “health status indicators”, “health disparities”, “health surveys”, “life expectancy” 
“gender gap”, “gender equality”, “gender equity”, “gender disparity”, “gender inequality”