Eight Dimensions of Gender Reshape How Scientists Predict Health Outcomes

For decades, health research has treated gender as little more than a checkbox, collapsing a richly layered social reality into the binary categories of male and female. A new study published in Discover Mental Health challenges that simplification by refining and validating a Spanish-language version of the Gender-Related Variables for Health Research, or GVHR, an instrument designed to capture the norms, traits, and relational patterns that shape health far beyond biological sex. The work, led by Sara Esteban-Gonzalo together with M. Inmaculada López-Núñez and Juan F. Díaz-Morales at Complutense University of Madrid, offers some of the strongest statistical evidence yet that gender operates as a multidimensional construct with measurable consequences for physical and mental wellbeing.

The research team recruited 646 adults ranging in age from 18 to 78 years, a sample broad enough to span working-age adults through older retirement years. Each participant completed the GVHR alongside a battery of comparison instruments: the Personality Traits Questionnaire, the Gender Determinism scale, and validated measures of health-related quality of life, psychological health, and health-risk behaviours. This design allowed the investigators to test not only whether the refined questionnaire held together statistically, but whether its dimensions carried unique predictive information about health once conventional demographic variables were taken into account.

At the heart of the study lies a technical exercise in psychometrics known as confirmatory factor analysis. Rather than allowing factors to emerge freely from the data, the researchers specified in advance an eight-factor structure and tested how well the observed responses fit that theoretical model. The eight dimensions that survived scrutiny were caregiver strain, work strain, independence, risk taking, emotional intelligence, social support, discrimination, and competition. The fit indices were respectable for research of this kind: a root mean square error of approximation of 0.052 with a 95 percent confidence interval spanning 0.048 to 0.056, a comparative fit index of 0.91, a Tucker-Lewis index of 0.90, and a goodness-of-fit index of 0.96. In practical terms, these numbers indicate that the eight-factor model reproduces the pattern of responses in the sample with adequate precision, a nontrivial achievement for an instrument measuring constructs as diffuse as gendered expectations.

Equally important was the demonstration of factorial invariance, the statistical property that a questionnaire measures the same underlying constructs in different groups. The team tested whether the instrument behaved equivalently across sex and across age, and the data supported scalar invariance in both cases. Scalar invariance is the stricter and more demanding level of equivalence, because it establishes not only that the same factors exist in each group but that the response scales are calibrated identically. Without it, comparing mean scores between men and women, or between younger and older adults, would be statistically indefensible. Its presence means the refined GVHR can legitimately be used to compare gender-related experiences across these demographic strata in future studies.

The refinement itself represents a meaningful evolution of the previous Spanish adaptation of the Stanford-developed instrument. The authors reformulated the independence and emotional intelligence dimensions and added an entirely new factor, competition, measured through items probing whether respondents see themselves as competitive, feel compelled to compete even when unnecessary, and act competitively across many areas of life. The independence items now blend the value placed on self-reliance with the frequency of depending on others for help and the discomfort of failing to meet others’ expectations, capturing a relational texture that a simple autonomy scale would miss. Emotional intelligence items tap attention to one’s own moods, awareness of feelings toward others, and the ability to self-soothe when ruminating.

What elevates the study beyond a psychometric exercise is the systematic link it draws between gendered experience and health. Using multivariate logistic regressions adjusted for sex, age, education, marital status, and income, the researchers quantified how each dimension relates to health outcomes. Caregiver strain emerged as the most consequential factor. People reporting higher caregiver strain had 2.29 times the odds of poorer general health, 1.59 times the odds of poorer mental health, 1.34 times the odds of poorer physical health, and 1.68 times the odds of health limitations. The same dimension predicted 1.85 times greater odds of depression, lower life satisfaction, with an odds ratio of 0.45 indicating reduced odds of higher satisfaction, and 2.22 times greater work-family conflict.

Work strain told a parallel story. Higher work strain was associated with poorer mental health, with an odds ratio of 1.67, poorer physical health at 1.50, greater odds of depression at 1.61, and the single largest association in the study, a 2.79-fold increase in the odds of work-family conflict. Discrimination based on gender, measured across school, healthcare, and public settings, was linked to poorer physical health, with an odds ratio of 1.36, and greater health limitations at 1.62. Crucially, these associations persisted after statistical adjustment for sex and socioeconomic characteristics, which means the gendered dimensions of experience carry health information that binary sex classification and income alone cannot supply. The burden of unpaid caregiving, the pressure of demanding work, and the experience of gender-based discrimination are not proxies for being a woman or a man; they are health-relevant exposures in their own right.

The methodological rigour of the study deserves emphasis because gender-health research has often been criticised for relying on ad hoc measures with untested psychometric properties. Here, the questionnaire items were published in full as an appendix, allowing other researchers to inspect exactly what is being asked, from the frequency of feeling physically and emotionally exhausted by caregiving tasks to the frequency of feeling discriminated against when receiving medical care. The study protocol received ethics approval from the Faculty of Psychology at the Universidad Complutense de Madrid, participants provided written informed consent, and data were handled anonymously in compliance with the Declaration of Helsinki. The work was supported by a grant from the Spanish Ministerio de Ciencia e Innovación, reference PID2024-159240NB-I00, and the article is published open access under a Creative Commons licence.

For clinicians and public health planners, the implications are tangible. Screening for caregiver strain in primary care, for example, could identify patients at elevated risk of depression and deteriorating general health long before a binary sex variable would flag them, since the study shows the strain itself, not the gender category, drives the association. Similarly, recognising work strain and gender-based discrimination as modifiable or at least measurable exposures opens the door to interventions targeted at the social conditions that produce illness, rather than at demographic categories that merely correlate with them. The inclusion of social support as a measured dimension, spanning advice, affection, and practical help with daily tasks, gives researchers a lever for studying protective factors alongside risks.

Limitations remain, as they do in any single-sample validation study. The cross-sectional design cannot establish whether caregiver strain causes depression or whether depressed individuals perceive their caregiving burden more heavily, and the Spanish adult sample may not generalise automatically to other linguistic and cultural contexts, although the underlying instrument was designed for cross-cultural adaptation. The authors also note that the version shared ahead of final publication is subject to further editorial processing. Even so, the combination of an adequate eight-factor fit, demonstrated scalar invariance across sex and age, and robust adjusted associations with multiple health indicators positions the refined GVHR as a serious candidate for the measurement standard that gender-sensitive health research has lacked. As the field moves toward precision public health, instruments that quantify the social architecture of gender, from caregiving and competition to discrimination and support, may prove as essential to predicting health as any biomarker.

Subject of Research: Validation of the eight-factor Gender-Related Variables for Health Research model and its associations with health outcomes in Spanish adults

Article Title: Refining the 8-factor GVHR model to improve gender-related health prediction

Article References: Esteban-Gonzalo, S., López-Núñez, M. I., & Díaz-Morales, J. F. (2026). Refining the 8-factor GVHR model to improve gender-related health prediction. Discover Mental Health. https://doi.org/10.1007/s44192-026-00614-1

Image Credits: AI Generated

DOI: 10.1007/s44192-026-00614-1

Keywords: gender and health, GVHR, psychometric validation, confirmatory factor analysis, caregiver strain, work strain, gender discrimination, social support, mental health, health-related quality of life, factorial invariance, health prediction