Sage Journals: Discover world-class research

Abstract

Women differ from men in many health indicators. We tested the extent to which such differences can be explained by gender constructs in a representative Canadian cohort (N = 2423). Institutionalized gender (education, income) and seven gender domains (e.g., perceived discrimination, work strain, caregiver strain) were assessed in their association with scales assessing health and well-being, and in several self-reported health conditions. Women reported higher rates of poor mental and physical health, depressive symptoms, stress, and major depressive disorder, while men reported more cardiovascular risk factors, heart disease, stroke, and type 2 diabetes. Gender variables (in particular, perceived discrimination, caregiver strain, and low income) fully explained negative health indicators in women and strongly contributed to those in men. Social support and preparedness to take risks were protective factors. Findings suggest substantial contributions of gender variables in health disparities between women and men, highlighting modifiable prevention and intervention targets.

Keywords

gender health behavior health psychology women men well-being

Introduction

Sex and gender are important determinants in human health. “Sex” refers to biological characteristics of women and men, while “gender” denotes characteristics that are constructed by social norms and expectations (World Health Organization, 2024). Biological differences between males and females of many species such as those tied to the X or Y chromosomes (Schurz et al., 2019), or hormonal differences (Kaur and Werstuck, 2021) can significantly impact morbidity, mortality, disease course, and treatment responses (Mauvais-Jarvis et al., 2020), underscoring the importance of sex-inclusive research. However, biological sex and gender are difficult to separate in humans considering that many psychosocial differences between men and women continue to persist, creating uncertainty about solely biological interpretations of sex differences in human health. Gender is tied to many intersectional factors like income (Grotti and Scherer, 2016), education (Bertocchi and Bozzano, 2020), or caregiving and work obligations (Lee et al., 2015), and it is pertinent to measure these factors systematically to address long-standing gender biases in health research (Upchurch, 2020). Unlike biological sex, such psychosocial variables, be they unique or shared between genders, can be changed, offering public health targets to improve outcomes for all people (Bartz et al., 2020).

Gender constructs are evolving and require contemporary assessment tools. For example, past decades have shown considerable societal changes for women in taking on active roles outside of the home, with substantially fewer changes for men, accompanied by weakening of gender stereotypes for women and relative stability for men (Croft et al., 2015). Differences between men and women in stereotypically feminine traits (“gentle,”“affectionate,” etc.) have similarly diminished over past decades, while masculine traits (“analytical,”“aggressive”) have not (Bhatia and Bhatia, 2021). Thus, to evaluate relationships between gender and health indicators, it is important to use contemporary gender assessment tools.

The Stanford Gender-Related Variables for Health Research (SGVHR) questionnaire was recently developed for this purpose and assesses gender on seven domains with potential health relevance: perceived Discrimination, Work Strain, Caregiver Strain, Risk Taking, Social Support, Emotional Intelligence, and Independence (Nielsen et al., 2021). Relationships between these gender dimensions and a few self-reported health variables were also probed: Health-related quality of life (general health, physical/mental health, activity limitations), smoking/vaping, binge drinking, and overweight. Discrimination and Caregiver Strain, which were elevated in women compared to men, were tied most consistently to poor health in Nielsen et al. (2021). While we recently replicated the factor structure of this scale (Abdel-sayyed et al., 2025), the health relevance of the survey needs clarification, including with validated clinical scales. For example, the men in Nielsen et al.’s (2021) cohorts overall reported lower well-being than the women. Yet a large body of literature suggests that women’s self-reported health—especially in mental health indicators (Otten et al., 2021)—is often lower compared to men (Boerma et al., 2016). Elevated prevalence of stress, mood, and affective disorders in women are particularly well-documented (Afifi, 2007). Apart from biological differences (e.g. hormonal (Li and Graham, 2017), neuroanatomical (Bangasser and Cuarenta, 2021)), disadvantageous gender roles, expectations, and sociodemographic factors in women likely have an impact (Christiansen and Berke, 2020). Conversely, men are often more likely than women to report substance use (Keyes et al., 2008), and cardiovascular conditions, with growing evidence for interaction effects between biological sex and gender in cardiovascular disease risk and course (Connelly et al., 2021). The sensitivity of the SGVHR to substance abuse and cardiovascular health has not been examined.

Thus, we focused on health variables and conditions with known differences in prevalence between women and men (Mauvais-Jarvis et al., 2020) to test the health relevance of the SGVHR gender dimensions, along with institutionalized gender (income and education). Our working hypothesis was that women would show higher rates of stress and affective symptoms, whereas men would show higher rates of substance use. Based on Nielsen et al.’s (2021) previous findings, we expected that the SGVHR gender dimensions, in particular Discrimination and Caregiver Strain, would substantially modify these relationships and may also extend to physical health indicators such as cardiovascular health.

Methods

This study received institutional ethics approval and was conducted in accordance with the Declaration of Helsinki.

Participants

Participants were recruited from several different locations across Canada through Qualtrics XM panel recruitment service using a stratified convenience sampling method to ensure national probability samples with geographic and demographic representativeness (Boas et al., 2020). Requesting equal quotas of men and women (while allowing for respondents with other gender designations), a representative distribution of 2527 adult (18+ years of age) respondents residing in Canada was recruited. After data cleaning and removal of fraudulent, incomplete, or otherwise uninterpretable datasets, a total of 2445 participants remained (see Supplemental Figure 1). Since only 22 participants identified as gender minorities (e.g. non-binary), data from individuals who did not identify as men or women were excluded from subsequent analyses, leaving 2423 participants (women = 1223, men = 1200). See Abdel-sayyed et al. (2025) for additional details.

Measures

The survey included sociodemographic questions, the SGVHR, and a series of health-related screenings and questions.

Sociodemographic questions

We included questions to ascertain age, gender, sexual orientation, ethnicity/race, highest educational attainment, employment status, caregiving status, and yearly personal income. An overview of this cohort’s sociodemographic characteristics is provided in Table 1.

Table 1.

Cohort characteristics (N = 2445).

Characteristics	Frequencies (%) or means (SD)
Age in years, mean (SD)	49.31 (16.52)
Gender, N (% of sample)
Man	1200 (49)
Woman	1223 (50)
Gender-fluid/non-binary	19 (1)
Other	3 (0.1)
Sexual orientation, N (% of sample)
Heterosexual or straight	2232 (91)
Gay or lesbian	99 (4)
Bisexual	85 (3)
Asexual	2 (<0.1)
Not sure/other	27 (1)
Ethnicity/race, multiple selections possible, N (% of sample)
White	1788 (73)
Hispanic, Latinx, or Spanish	56 (2)
Black or African Canadian/American	119 (5)
Asian	425 (17)
Indigenous	57 (2)
Middle Eastern or North African	46 (2)
Native Hawaiian or Pacific Islander	1 (<0.01)
Other	41 (2)
Highest educational attainment, N (% of sample)
High school degree or lower	563 (23)
Some college/associate degree/technical degree	876 (36)
Bachelor’s degree	708 (29)
Master’s degree	230 (9)
Professional Degree	26 (1)
Doctorate	41 (2)
Other	0 (0)
Working status, multiple selections possible, N (% of sample)
Currently working	1379 (56.4)
Retired	607 (24)
Homemaker	121 (4.9)
Volunteer	26 (1.1)
Student	98 (4.0)
Unemployed	147 (6.0)
On leave of absence/disability	102 (4.2)
Caregiving status, N (% of sample)
Currently a caregiver	1085 (45)
Not a caregiver	1360 (56)
Yearly personal income, N (% of sample)
<$15,000 or equivalent	282 (12)
$15,000–$29,999	415 (17)
$30,000–$49,999	518 (22)
$50,000–$69,999	435 (18)
>$70,000	649 (27)
Other (unknown, do not wish to answer, etc.)	146 (6)

Gender: Stanford Gender-Related Variables for Health Research (SGVHR)

The SGVHR (Nielsen et al., 2021) is a 25-item self-report survey and conceptualizes gender as a multimodal construct. Scale construction was based on extensive literature review, factor analytical reduction, and validation in multiple cohorts, arriving at an assessment of gender on seven dimensions, using 23 items with 5-point Likert scales and two items querying time estimates for weekly work and caregiving activities: The 6-item Discrimination subscale (in this cohort, Cronbach’s α = 0.92) measures frequency of gender-based unfair treatment (example: “Because of your gender, how often have you felt discriminated against when getting hired?”). The 5-item Work-Strain subscale assesses emotional and physical demands related to work (“How often do you feel emotionally exhausted from your work activities?,”α = 0.90). The 4-tem Caregiver-Strain subscale measures the negative impacts of unpaid long-term caregiving (“In the past year, how often did you feel physically exhausted because of your caretaking responsibilities?,”α = 0.89). Risk-Taking gages a person’s propensity for taking risks on 3 items (“How prepared are you to take risks when making financial decisions?,”α = 0.81). Social Support has two items evaluating the availability of social support (“In the past year, how often did you have someone to help you with daily chores?,”α = 0.74). Emotional Intelligence measures a person’s ability to recognize and use their emotions in problem-solving with 3 items (“How often do you talk to your friends about your problems?,”α = 0.65). The 2-item Independence subscale assesses how much a person values being independent (“How important is it for you to solve your problems on your own?,”α = 0.73). Scores in each of the SGVHR items were first z-scaled and then summarized into the seven subscales.

Health-Related Quality of Life Core Module (CDC HRQoL-4)

We assessed self-rated health using the four items from this module querying perceived general health, along with recent (past 30 days) physical health, mental health, and activity limitations. Answers in general health were dichotomized to represent general health as fair or poor (reference = 0) or good, very good, or excellent (1). Due to the severe skewness of days-per-month estimates, dichotomization of these three HRQoL-4 items is common (Dumas et al., 2020; Wilkins et al., 2018), and was also performed by Nielsen et al. (2021) to identify individuals affected by meaningful levels of distress or limitations. We followed this approach and used the common dichotomization threshold of 14+ days to reflect “poor” physical or mental health, or extensive activity limitations (coded as 1), compared to “good” health (13 days or less, coded as 0).

Alcohol Use Disorders Identification Test Consumption (AUDIT-C) and Drug Abuse Screening Test (DAST-10)

To assess alcohol consumption, the three-question AUDIT-C (Bush et al., 1998) was used, with scores ranging from 0 to 12. Scores were dichotomized as 4 or more indicating potentially problematic alcohol consumption for men, and scores of 3 or more for women. To assess substance abuse in the past 12 months we used the 10-question DAST-10 (Skinner, 1982) with scores ranging from 0 to 10, indicating presence and severity of substance use. Outcomes in the DAST-10 were dichotomized to represent positive indicators of substance use at scores of 3 or above (“moderate level” of substance use).

Patient-Health Questionnaire (PHQ-2) and Perceived Stress Scale (PSS-4)

The two-item PHQ-2 is the short version of the PHQ-9 (Kroenke et al., 2003). It assesses low mood and anhedonia as core symptoms of depression within the past 2 weeks (example: “Over the last 2 weeks, how often have you been bothered by the following problems? Little interest or pleasure in doing things.” Answers range from 0: “not at all” to 3: “nearly every day”). PHQ-2 totals range from 0 to 6 and were dichotomized at the clinical threshold of 3 or above indicating a high likelihood of depression (Kroenke et al., 2003). The four-item PSS-4 is the short version of the Perceived Stress Scale which measures the frequency and severity of perceived stressors within the past month (Cohen et al., 1983; example: “In the last month, how often have you felt that you were unable to control the important things in your life?” Answers range from 0: “never” to 4 “very often”). Total PSS-4 scores range from 0 to 16. The PSS-4 has no clinical threshold, but considering 8 points as 50% of the scale’s maximum score we dichotomized the PSS-4 total to indicate high stress at 8 or more points following previous studies (Campbell et al., 2020; Giovenco et al., 2022).

Health-related variables and conditions

A number of health-relevant variables and conditions with potential differences between men and women (Mauvais-Jarvis et al., 2020) were queried. These included weight and height to calculate body-mass index (BMI = kg/m²), dichotomized as scores of 30 and above indicating obesity versus scores below 30 indicating non-obese BMI. Blood pressure, cholesterol, and fasting glucose levels were probed and dichotomized as positive (“high” responses) or negative indicators (“low” and “I don’t know” responses). Cigarette smoking was queried as average cigarettes smoked/day and years of smoking, dichotomized into a single score for any indication of past or present cigarette smoking. Specific health conditions were queried with “yes,”“no,”“I don’t know,”“do not wish to answer” as response options: Major depressive disorder, type-2 diabetes, heart disease, stroke, chronic obstructive pulmonary disease (COPD), cancers, Alzheimer’s disease, influenza, kidney diseases. Endorsement was coded as “1,” all other responses were coded “0”.

Statistical analyses

The institutionalized gender variables education and personal income were simplified into three categories and dummy-coded. This categorization resulted in 1334 individuals in the educational attainment reference group (less than a bachelor’s degree), 666 individuals with a bachelor’s degree, and 227 individuals with postsecondary education. A total of 146 participants did not report an income such that the final numbers within income brackets were 686 individuals that reported ≤$29,999 personal income per year, 514 individuals with $30,000–$69,99, and 1077 individuals with $70,000 or more personal income per year.

We first conducted univariate comparisons between men and women using χ²-tests for dichotomous variables or t-tests for continuous variables (SGVHR dimensions). Health variables that showed univariate differences between men and women were then further tested with hierarchical logistic regressions, conducted on 2277 participants (1127 women, 1150 men) without imputing missing values in income. To reduce the number of analyses, we summarized high blood pressure, high cholesterol, high blood glucose, or smoking as cardiovascular risk factors (CVR), coded as “1” for positive indications for any one of these four variables and “0” if none of them were endorsed. Furthermore, we summarized heart disease and stroke as cardiovascular diseases (CVD), again coded as “1” if either were endorsed and “0” for negative responses in both.

We tested two models for each health indicator: Simple models included age and being a woman (vs man = 0). In the complete models, we added the SGVHR, income, and educational attainment in a second predictor block. Analyses were repeated in women and men separately to test if the gender variables in block 2 contributed similarly or differently to health in women than in men. Bonferroni-corrections were applied to correct for the number of models we tested. All statistical analyses were carried out with SPSS v.29 (IBM SPSS Statistics, 2023).

Results

Cohort characteristics are summarized in Table 1. We obtained data from 1200 men (49%), 1223 women (50%), and 22 participants who identified as gender-fluid, non-binary, or another gender. The cohort was middle-aged (mean age = 49.31 years, SD = 16.52) and predominantly “White” with heterosexual orientation. The majority’s (59%) highest educational attainment fell below a bachelor’s degree, and 45% reported an annual income of at least CAD $30,000. Over half of the cohort (56.79%) reported they are currently working, 24% reported being retired, and almost half (45%) reported having some caregiving responsibilities.

We tested univariate differences in health variables comparing women (N =1223) with men (N = 1200), omitting the small number of individuals identifying as gender-fluid/non-binary or another gender. Compared to men, women reported more problems with physical health, mental health, and activity limitations, depressive symptoms (PHQ-2), stress (PSS-4), and rates of major depressive disorder (all p’s < 0.05). Conversely, men reported higher rates of type 2 diabetes, cardiovascular diseases (heart disease, stroke), and cardiovascular risk factors (high blood pressure, high cholesterol, high blood glucose, and smoking). Apart from these nine health indicators, none of the other variables showed univariate differences between women and men (Supplemental Table S2 for statistical details).

We then followed up with nine hierarchical logistic regressions, one for each health indicator that had shown differences between men and women, probing effects of age and being a woman in simple models first, and then adding the SGVHR, income, and education in block 2 in the complete models. Note that most of the predictors in block 2 differed between women and men in the expected directions, that is, to the disadvantage of women compared to men (Supplemental Figure 2). As shown in Table 2, all simple models (block 1) were significant, explaining between 1% (activity limitations) and 14% (PSS-4) of the variance in the health indicators. However, including the SGVHR and institutionalized gender variables in the second block increased the variance explanation substantially, ranging from 9.8% (type 2 diabetes) to 27% (PSS-4) and all final models were significant after Bonferroni-correction by a factor of nine.

Table 2.

Hierarchical logistic regressions (N = 2277) on nine health indicators that showed univariate differences between men and women (cf. Table S2).

Test statistic	Physical health	Mental health	Activity limitations	PHQ	PSS	MDD	T2D	CVD	CVR
Simple models: women (vs. men), age
Model χ²	39.605***	47.436***	12.414**¹	129.312***	252.236***	39.671***	83.180***	60.865***	92.950***
Pseudo R²	0.035	0.036	0.011	0.083	0.141	0.028	0.076	0.072	0.053
Complete models: women (vs. men), age, income, education, SGVHR
Block χ²	109.933***	102.185***	127.751***	227.060***	254.757***	109.119***	24.616*¹	37.512***	66.275***
Model χ²	149.539***	149.621***	140.164***	356.371***	506.992***	148.791***	107.796***	98.377***	159.226***
Pseudo R²	0.128	0.110	0.125	0.218	0.269	0.103	0.098	0.116	0.090

PHQ: depressive symptoms in the PHQ-2; PSS: high stress in the PSS-4; MDD: major depressive disorder; T2D: type-2 diabetes; CVD: cardiovascular disorders (heart disease, stroke); CVR: cardiovascular risk factors (high blood pressure, high cholesterol, high blood glucose, smoking).

p < 0.05. **p < 0.01. ***p < 0.001. ¹Not significant after Bonferroni-correction.

The direction and magnitude of the odds-ratios for each predictor and the negative health indicators are illustrated in Figure 1(a). Being a woman was no longer related to any of the health variables that had shown univariate disadvantages in women (physical health, mental health, activity limitations, depressive symptoms, stress, and major depressive disorder). These were better explained by the gender variables in block 2 (SGVHR, education, income). Of the three health variables that had shown univariate disadvantages in men, the odds for cardiovascular diseases and cardiovascular risk factors (but not type-2 diabetes) were still elevated in men in the complete models.

Figure 1.

Significant odds-ratios of the complete models from Table 2. (a) All participants (b) Women, (c) Men. See Supplemental Materials for all odds ratios and their 95%-confidence intervals.

Within block 2, higher educational attainment and higher personal income reduced the odds for almost all health variables, with up to 66% reduced odds for activity limitations in participants in the highest (compared to lowest) income bracket.

Among the SGVHR variables, higher scores in perceived Discrimination were the most important and predicted all nine adverse health variables, followed by Caregiver Strain, which was linked to seven variables. The strongest relationships with the SGVHR were seen in some of the mental health variables. For example, perceived Discrimination increased the odds for depressive symptoms in the PHQ-9 by 88%, and high stress in the PSS-4 by 67%. These two mental health indicators interacted also with Caregiver Strain, each associated with 48% increased odds, and were the only two outcomes interacting with Work Strain (36% increased odds of depressive symptoms, 23% increased odds of high stress, respectively).

Inspecting the outcomes in women (Figure 1(b)) and men (Figure 1(c)) separately, many of the outcomes were similar, but with some variations (see Supplemental Tables for additional details). For example, perceived Discrimination most consistently increased the odds of adverse health, especially stress and depressive symptoms, in both women and men but interacted with self-reported poor health/activity limitations only in women, and with type-2 diabetes/cardiovascular conditions only in men. Particularly for men, Caregiver Strain increased, and higher income decreased the odds for negative health indicators, compared to more widespread positive relationships between health and higher education levels, Social Support, and Risk-Taking in women.

Discussion

In this Canadian cohort, women reported lower health-related quality of life, more stress, depressive symptoms, and higher rates of major depressive disorder than men. Men reported higher rates of cardiovascular risk factors (high blood pressure, high cholesterol, high blood glucose, and smoking), cardiovascular diseases (heart disease, stroke), and type 2 diabetes. Many of these differences were better explained by the assessed gender variables in the SGVHR, along with education and income. Our findings suggest that these gendered variables are critical in understanding health risks, emphasizing the potential benefits of modifying these factors as part of healthcare and prevention.

Education and income interacted with health in the expected ways, with higher levels reducing the odds of negative indicators. Relationships with educational attainment were more pronounced for women; relationships with income were evident in multiple health indicators for women and men, but more pronounced in men. For example, only in men, higher income significantly reduced the odds of stress and depressive symptoms.

Among the SGVHR variables, perceived Discrimination played a particularly important role and was negatively related to all health variables. Gender-based discrimination is a well-described risk factor for mental and physical well-being (Hackett et al., 2019), and similar effects have been documented for racial (Kirkinis et al., 2021), body weight (Gerend et al., 2023), workplace (Xu and Chopik, 2020), or socioeconomic (Angerer et al., 2019) discrimination. The adverse effects of perceived discrimination on health are multifactorial, including systemic biases in health resource allocation and access (Krieger, 2020), trust in healthcare providers (Hong et al., 2018), and disengagement from available healthcare resources (Burgess et al., 2008). Perceived discrimination has been robustly linked to stress (Pascoe and Richman, 2009), and the perpetuated psychological trauma resulting from long-term discrimination (Kirkinis et al., 2021) can lead to chronic overactivation of the hypothalamic-pituitary-adrenal (HPA) axis, the body’s stress response system. This affects multiple downstream metabolic and inflammatory pathways, which can later manifest as physical or mental health problems. Indeed, multiple forms of discrimination have been associated with HPA axis overactivity (Busse et al., 2017), also raising the likelihood of engaging in maladaptive stress coping strategies (Crockett et al., 2018). Thus, in addition to the direct physiological effects of chronic stress caused by discrimination, participation in unhealthy behaviors and non-participation in healthy behaviors are more likely with higher perceived discrimination.

The odds of reporting cardiovascular diseases were reduced by 58% and the odds of cardiovascular risk factors were reduced by 36% in women compared to men, and these simple differences persisted in the complete models (see Figure 1(a)). Nevertheless, Discrimination, Caregiver Strain, and less so, Independence significantly increased the odds, while Social Support and higher education decreased the odds, with some variation between women and men (cf. Figure 1(b)): Discrimination increased the odds of cardiovascular diseases in men and of cardiovascular risk factors in women. Social Support and higher education decreased the odds of cardiovascular risk factors only in women. Conversely, higher Caregiver Strain and Independence increased the odds of cardiovascular risk factors only in men.

Cardiovascular disease has historically been considered a disease of men, and current prevalence rates still show higher rates in men than in women across all ages (Martin et al., 2024). Biological factors play a role in these differences, for example, higher average visceral fat in men than women (Nakamura et al., 1994) and so do differences in lifestyle and stress coping (Chumaeva et al., 2009). However, gender biases in diagnosis, treatment, and prevention of cardiovascular risk factors and diseases are likely also reflected in our results. For example, cardiovascular conditions in women are underdiagnosed, with treatment delays or under-treatment compared to men with similar or more severe symptoms (Todorov et al., 2021). Such biases may also contribute to women underrecognizing their own risk for cardiovascular disease, and missing or misinterpreting their symptoms (Smith et al., 2018). Thus, the robustly higher rates of cardiovascular conditions and risk factors we observed in men here are likely a combination of biological, personal, and societal factors. With increasing recognition of the complex interplay between sex and gender in cardiovascular health (Regitz-Zagrosek and Gebhard, 2023), our study also pointed to important nuances in which of the assessed gender-related variables were important, in the entire cohort as well as particularly in women or in men. Similar to previous findings predominantly from studies of racial discrimination (Panza et al., 2019), higher perceived Discrimination elevated the odds of cardiovascular risk factors (in women) and diseases (in men). Bachelors-level education reduced the odds of cardiovascular risk factors by 30% in the entire cohort, particularly in women. This resembles previous metanalytical findings (Backholer et al., 2017) where lower education increased the risk for coronary heart disease in women and men, but women with the lowest education were at a 24%–31% excess risk compared to men at the same low education levels. Higher Caregiver Strain increased the odds of cardiovascular risk factors, and this was particularly evident in men, adding to the limited cardiovascular health literature on this variable (Lambrias et al., 2023). Generally, caregiver strain has robust relationships to stress and mental health (Kayaalp et al., 2021) as replicated in our findings. We can only speculate about why Caregiver Strain interacted more strongly with men’s health here. Taking on caregiving responsibilities could mark a reversal of traditional male gender roles (Montgomery and Kosloski, 2009) and a loss of previous identities with shifts in power and authority (Kim and Moen, 2002). Caregiver Strain in men, if reported, could then reflect a combination of factors that interact more strongly with health, a hypothesis that should be tested in future studies.

Some unexpected findings were also observed and will be briefly discussed in the following. Higher scores in the Independence subscale increased the odds of cardiovascular risk factors (in men) and higher scores in the Risk-Taking subscale of the SGVHR decreased the odds of several negative health indicators. We suggest that the content and wording of these subscales may have contributed to these outcomes. As described in the Methods section, most of the SGVHR subscales ask participants to estimate the frequency of specific behaviors or experiences. However, Risk-Taking inquires how “prepared” someone feels about engaging in risky activities (“How prepared are you to take risks when making financial decisions?”) and Independence asks about how “important” someone’s personal independence is to them (“How important is it for you to solve your problems on your own?”). Women showed markedly lower Risk-Taking scores than men (Figure S2C), which emphasizes the gender-relevance of this subscale and may reflect higher risk aversion in women than men (Dawson, 2023). However, feeling highly prepared to take on risks in various scenarios, rather than actually engaging in risky activities could also reflect optimism, with positive implications for some health indicators (Rozanski et al., 2019), and perhaps specifically so for women. Furthermore, based on previous findings (Keyes et al., 2008; Mauvais-Jarvis et al., 2020), we expected higher rates of alcohol and substance use in men than women, but because no differences emerged in the univariate outcomes, relationships with the gender variables were not further probed. However, addictions may still show positive relationships with Risk-Taking and should be examined in future studies.

Some limitations of the current study should be mentioned. First, we did not inquire about our participants’ sex at birth but asked them to identify as men, women, or another gender. Although gender identity closely matched sex at birth in similar studies that assessed both (e.g. >98% concordance in Nielsen et al., 2021), this shortcoming should be considered when interpreting the differences between women and men in this study. Furthermore, very few participants identified as gender-fluid/non-binary, precluding a meaningful subgroup analysis of these individuals. Future targeted studies should focus on understanding the health relevance of gendered variables specifically in individuals who do not identify as women or men, as well as in otherwise more diverse populations than a predominantly white cohort of heterosexual individuals. Finally, based on limited distributional ranges in many of the health scales and to compare outcomes with those on presence/absence of health conditions, we dichotomized the scores on the health scales. Even though we followed previous studies in this practice and used available clinical cut-off scores to motivate this choice, scores on these scales may also be analyzed with linear regressions. Such a re-analysis of the PSS-4 (in the entire cohort) can be found in the Supplemental Material, with similar outcomes to the present logistic regression results.

Conclusions

In addition to biological sex, gender is now well recognized as a critical determinant of human health and many health-related differences between women and men may be driven by gender, representing potentially modifiable public health targets. Our findings demonstrate the potential this approach holds, with the most compelling example that the initial disadvantages in mental and physical health we observed in women relative to men were fully explained by gender variables. To assess health-relevant gender variables in a climate of evolving perceptions of gender, novel validated assessment tools like the SGVHR are needed. Our findings suggest that among the assessed gender facets in this tool, perceived Discrimination, Caregiver Strain, followed by Work Strain might be the most promising dimensions associated with health risks. Their assessment should be targeted to develop and optimize interventions promoting optimal health for all individuals.

Supplemental Material

sj-docx-1-hpq-10.1177_13591053251341188 – Supplemental material for Gender predictors of adverse health

Supplemental material, sj-docx-1-hpq-10.1177_13591053251341188 for Gender predictors of adverse health by Yousef Jallad, Ahmed Abdel-sayyed, Tarek Turk, Kim Ngan Hoang, Lujie Xu and Esther Fujiwara in Journal of Health Psychology

Footnotes

Acknowledgements

We would like to acknowledge Sasha Kanevets for her help with data cleaning.

Data sharing statement

The current article is accompanied by the relevant raw data generated during and/or analyzed during the study. These files are available and accessible on the OSF platform: . Ethics approval, participant permissions, and all other relevant approvals were granted for this data sharing.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by the Canadian Institutes of Health Research. FRN 201803.

Ethics approval

This study received ethics approval from the University of Alberta Research Ethics Board (Pro00086704: “Gender Roles and Health”).

Informed consent

All participants provided informed consent prior to data collection. Participation was voluntary, and participants could withdraw at any time without penalty.

Consent to publication

All participants consented to the publication of de-identified findings from this research.

ORCID iDs

Yousef Jallad

Ahmed Abdel-sayyed

Tarek Turk

Kim Ngan Hoang

Lujie Xu

Esther Fujiwara

Supplemental material

Supplemental material for this article is available online.

References

Abdel-sayyed

Hoang

Turk

, et al. (2025) Validating the Stanford gender-related variables for Health Research (SGVHR) in a Canadian population. Journal of Health Psychology 30(6): 13591053241247376. DOI: 10.1177/13591053241247376.

Afifi

(2007) Gender differences in mental health. Singapore Medical Journal 48(5): 385–391.

Angerer

Waibel

Stummer

(2019) Discrimination in health care: A field experiment on the impact of patients’ socioeconomic status on access to care. American Journal of Health Economics 5(4): 407–427.

Backholer

Peters

SAE

Bots

, et al. (2017) Sex differences in the relationship between socioeconomic status and cardiovascular disease: A systematic review and meta-analysis. Journal of Epidemiology and Community Health 71(6): 550–557.

Bangasser

Cuarenta

(2021) Sex differences in anxiety and depression: Circuits and mechanisms. Nature Neuroscience Reviews 22(11): 674–684.

Bartz

Chitnis

Kaiser

, et al. (2020) Clinical advances in sex- and gender-informed medicine to improve the health of all: A review. JAMA Internal Medicine 180(4): 574–583.

Bertocchi

Bozzano

(2020) Gender gaps in education. In: Zimmermann

(ed.) Handbook of Labor, Human Resources and Population Economics. Cham: Springer, pp.1–31.

Bhatia

(2021) Changes in gender stereotypes over time: A computational analysis. Psychology of Women Quarterly 45: 106–125.

Boas

Christenson

Glick

(2020) Recruiting large online samples in the United States and India: Facebook, Mechanical Turk, and Qualtrics. Political Science Research and Methods 8(2): 232–250.

10.

Boerma

Hosseinpoor

Verdes

, et al. (2016) A global assessment of the gender gap in self-reported health with survey data from 59 countries. BMC Public Health 16: 675.

11.

Burgess

Ding

Hargreaves

, et al. (2008) The association between perceived discrimination and underutilization of needed medical and mental health care in a multi-ethnic community sample. Journal of Health Care for the Poor and Underserved 19(3): 894–911.

12.

Bush

Kivlahan

McDonell

, et al. (1998) The AUDIT alcohol consumption questions (AUDIT-C): An effective brief screening test for problem drinking. Ambulatory Care Quality Improvement Project (ACQUIP). alcohol Use Disorders Identification Test. Archive of Internal Medicine 158(16): 1789–1795.

13.

Busse

Yim

Campos

, et al. (2017) Discrimination and the HPA axis: current evidence and future directions. Journal of Behavioral Medicine 40(4): 539–552.

14.

Campbell

Tamayo-Ortiz

Cantoral

, et al. (2020) Maternal prenatal psychosocial stress and prepregnancy BMI associations with fetal iron status. Current Developments in Nutrition 4(2): nzaa018.

15.

Christiansen

Berke

(2020) Gender- and sex-based contributors to sex differences in PTSD. Current Psychiatry Reports 22(4): 19.

16.

Chumaeva

Hintsanen

Ravaja

, et al. (2009) Interactive effect of long-term mental stress and cardiac stress reactivity on carotid intima-media thickness: The cardiovascular risk in Young Finns study. Stress 12(4): 283–293.

17.

Cohen

Kamarck

Mermelstein

(1983) A global measure of perceived stress. Journal of Health and Social Behavior 24(4): 385–396.

18.

Connelly

Azizi

Alipour

, et al. (2021) The importance of gender to understand sex differences in cardiovascular disease. Canadian Journal of Cardiology 37(5): 699–710.

19.

Crockett

Rice

Turan

(2018) Associations between multiple forms of discrimination and tobacco use among people living with HIV: The mediating role of avoidance coping. JAIDS Journal of Acquired Immune Deficiency Syndromes 78(1): 9–15.

20.

Croft

Schmader

Block

(2015) An underexamined inequality: Cultural and psychological barriers to men’s engagement with communal roles. Personality and Social Psychology Review 19(4): 343–370.

21.

Dawson

(2023) Gender differences in optimism, loss aversion and attitudes towards risk. British Journal of Psychology 114(4): 928–944.

22.

Dumas

Dongchung

Sanderson

, et al. (2020) A comparison of the four healthy days measures (HRQOL-4) with a single measure of self-rated general health in a population-based health survey in New York City. Health and Quality of Life Outcomes 18(1): 315.

23.

Gerend

Zetrenne

Sutin

, et al. (2023) Weight discrimination and health risk behavior in racial, ethnic, and sexual minority adults. Annals of Behavioral Medicine 57(7): 571–581.

24.

Giovenco

Shook-Sa

Hutson

, et al. (2022) Social isolation and psychological distress among southern U.S. college students in the era of COVID-19. PLoS One 17(12): e0279485.

25.

Grotti

Scherer

(2016) Does gender equality increase economic inequality? Evidence from five countries. Research in Social Stratification and Mobility 45: 13–26.

26.

Hackett

Steptoe

Jackson

(2019) Sex discrimination and mental health in women: A prospective analysis. Health Psychology 38(11): 1014–1024.

27.

Hong

Lee

Collins

, et al. (2018) Factors affecting trust in healthcare among middle-aged to older Korean American women. BMC Women’s Health 18(1): 109.

28.

IBM Corp (2023) IBM SPSS Statistics for Windows (Version 29.0) [Computer software]. IBM Corp. Available at: https://www.ibm.com/products/spss-statistics

29.

Kaur

Werstuck

(2021) The effect of testosterone on cardiovascular disease and cardiovascular risk factors in men: A review of clinical and preclinical data. Canadian Journal of Cardiology Open 3(10): 1238–1248.

30.

Kayaalp

Page

Rospenda

(2021) Caregiver burden, work-family conflict, family-work conflict, and mental health of caregivers: A mediational longitudinal study. Work and Stress 35(3): 217–240.

31.

Keyes

Grant

Hasin

(2008) Evidence for a closing gender gap in alcohol use, abuse, and dependence in the United States population. Drug and Alcohol Dependence 93(1–2): 21–29.

32.

Kim

Moen

(2002) Retirement transitions, gender, and psychological well-being: a life-course, ecological model. The Journals of Gerontology. Series B, Psychological Sciences and Social Sciences 57(3): 212–222.

33.

Kirkinis

Pieterse

Martin

, et al. (2021) Racism, racial discrimination, and trauma: A systematic review of the social science literature. Ethnicity and Health 26(3): 392–412.

34.

Krieger

(2020) Measures of racism, sexism, heterosexism, and gender binarism for health equity research: From structural injustice to embodied harm-an ecosocial analysis. Annual Review of Public Health 41: 37–62.

35.

Kroenke

Spitzer

Williams

(2003) The Patient Health Questionnaire-2: Validity of a two-item depression screener. Medical Care 41(11): 1284–1292.

36.

Lambrias

Ervin

Taouk

, et al. (2023) A systematic review comparing cardiovascular disease among informal carers and non-carers. International Journal of Cardiology Cardiovascular Risk and Prevention 16: 200174.

37.

Lee

Tang

Kim

, et al. (2015) Exploring gender differences in the relationships between eldercare and labour force participation. Canadian Journal on Aging 34(1): 14–25.

38.

Graham

(2017) Why are women so vulnerable to anxiety, trauma-related and stress-related disorders? The potential role of sex hormones. Lancet Psychiatry 4(1): 73–82.

39.

Martin

Aday

Allen

, et al. (2024) 2024 heart disease and stroke statistics: A report of US and global data from the American Heart Association. Circulation 149(8): e41–e913.

40.

Mauvais-Jarvis

Bairey Merz

Barnes

, et al. (2020) Sex and gender: Modifiers of health, disease, and medicine. Lancet 396(10250): 565–582.

41.

Montgomery

Kosloski

(2009) Caregiving as a process of changing identity: Implications for caregiver support. Generations 33(1): 47–52.

42.

Nakamura

Tokunaga

Shimomura

, et al. (1994) Contribution of visceral fat accumulation to the development of coronary artery disease in non-obese men. Atherosclerosis 107(2): 239–246.

43.

Nielsen

Stefanick

Peragine

, et al. (2021) Gender-related variables for health research. Biology of Sex Differences 12(1): 23.

44.

Otten

Tibubos

Schomerus

, et al. (2021) Similarities and differences of mental health in women and men: A systematic review of findings in three large German cohorts. Frontiers in Public Health 9: 553071.

45.

Panza

Puhl

Taylor

, et al. (2019) Links between discrimination and cardiovascular health among socially stigmatized groups: A systematic review. PLoS One 14(6): e0217623.

46.

Pascoe

Richman

(2009) Perceived discrimination and health: A meta-analytic review. Psychological Bulletin 135(4): 531–554.

47.

Regitz-Zagrosek

Gebhard

(2023) Gender medicine: Effects of sex and gender on cardiovascular disease manifestation and outcomes. Nature Reviews Cardiology 20(4): 236–247.

48.

Rozanski

Bavishi

Kubzansky

, et al. (2019) Association of optimism with cardiovascular events and all-cause mortality: A systematic review and meta-analysis. JAMA Network Open 2(9): e1912200.

49.

Schurz

Salie

Tromp

, et al. (2019) The X chromosome and sex-specific effects in infectious disease susceptibility. Human Genomics 13(1): 2.

50.

Skinner

(1982) The drug abuse screening test. Addictive Behaviors 7(4): 363–371.

51.

Smith

Frazer

Hyde

, et al. (2018) “Heart disease never entered my head”: Women’s understanding of coronary heart disease risk factors. Journal of Clinical Nursing 27(21–22): 3953–3967.

52.

Todorov

Kaufmann

Arslani

, et al. (2021) Gender differences in the provision of intensive care: A Bayesian approach. Intensive Care Medicine 47(5): 577–587.

53.

Upchurch

(2020) Gender bias in research. In: Naples

(ed.) Companion to Women’s and Gender Studies. Hoboken, NJ: Wiley Blackwell, pp.139–154.

54.

Wilkins

Ghosh

Vivar

, et al. (2018) Exploring the associations between systemic inflammation, obesity and healthy days: A health related quality of life (HRQOL) analysis of NHANES 2005-2008. BMC Obesity 5: 21.

55.

World Health Organization (2024) Gender and health. Available at: https://www.who.int/health-topics/gender#tab=tab_1 (accessed 15 August 2024).

56.

Chopik

(2020) Identifying moderators in the link between workplace discrimination and health/well-being. Frontiers in Psychology 11: 458.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.46 MB