The finding does not establish that improving a marriage would prevent suicide, nor does it show that grandparent relationships are unimportant. Instead, it identifies a specific prospective correlate of suicidal ideation in a setting where maternal mental health is shaped by tightly interconnected family, economic and social pressures. The researchers assessed relationship quality when children were six and examined suicidal ideation at subsequent follow-ups when children were seven and eight. Among 817 mothers with outcome data, 79, or 9.7%, endorsed suicidal ideation at one or both later assessments.

Suicidal ideation is the occurrence of thoughts about death or self-harm. In this study, it was measured with a single question from the Patient Health Questionnaire-9 asking how often, during the preceding two weeks, respondents had been bothered by thoughts that they would be better off dead or of hurting themselves. Any endorsement was classified as ideation. Because the measure covers only a two-week window and the outcome was uncommon, investigators combined the year-seven and year-eight assessments. They also conducted a safety protocol: women who disclosed thoughts of self-harm or suicide were assessed for ideation, intent and plans, and those needing additional care were referred for further evaluation.

The Bachpan Study is a prospective cohort in rural Rawalpindi that combines a cluster-randomized trial of a maternal depression intervention with a longitudinal observational cohort. Pregnant women were recruited from 46 village clusters beginning in 2014, and mother-child dyads were followed through multiple postnatal and childhood assessments. The year-six and year-seven-to-eight waves used in this analysis were conducted from 2021 to 2024. The larger cohort contains 1,154 mother-child dyads, while the present analysis included mothers who participated in the relevant waves and had non-missing suicidal-ideation data.

Two family relationships were examined. Marital quality was measured with six culturally adapted questions covering emotional support, practical support and overall relationship quality, producing a score from 6 to 30. The scale had strong internal consistency in the analytic sample, with a Cronbach’s alpha of 0.90. Relationship quality with a grandmother was measured using 15 items, also with strong internal consistency at alpha 0.81. Mothers rated whichever grandmother was most involved in the child’s life. This measure was available only when a grandmother was alive, in good health and actively involved, whether living in the same household or elsewhere.

That restriction produced substantially different samples for the two analyses. The marital analysis included 810 mothers after seven exclusions for missing relationship-quality data. The grandmother analysis included 450 mothers because 367 women did not have an eligible, actively involved grandmother. A restricted sample of 449 women had complete data for both exposures. Researchers therefore used modified Poisson regression with robust standard errors to account for clustering by village, standardized both exposures, and applied inverse-probability-of-censoring weights to address loss to follow-up. Missing data were handled with multiple imputation using chained equations.

For each one-standard-deviation improvement in marital relationship quality, the adjusted risk ratio for suicidal ideation was 0.75, with a 95% confidence interval of 0.60 to 0.93. The result was similar in the restricted sample, where the adjusted risk ratio was 0.56, and in a model that included both relationship measures simultaneously, where it was 0.54. The corresponding adjusted risk ratio for grandmother relationship quality was 0.96, with a 95% confidence interval of 0.67 to 1.36. The grandmother estimate remained close to one in the restricted and combined models, providing no evidence of an association in the studied population.

Sensitivity analyses broadly supported the marital finding. The association remained similar when the models controlled only for suicidal ideation at the year-six assessment rather than the broader history used in the primary analysis. It also persisted when the marital scale was reduced to five items to address a differently worded question. Results were consistent when year-seven and year-eight outcomes were analyzed separately. For the grandmother measure, removing all five negatively worded items did not materially change its association. Stratified estimates involving grandmother type and co-residence were limited by sparse data, and some adjusted models did not converge, so those exploratory results cannot establish differences between paternal and maternal grandmothers or between co-residing and separate households.

The results support the hypothesis that a supportive marital bond may be especially relevant to maternal suicide risk in a patrilocal setting, where a wife often lives with or near her husband’s family and may have fewer independent social ties. The researchers drew on family-systems theory and the interpersonal theory of suicide, which links suicidal desire to thwarted belongingness and perceived burdensomeness. A positive relationship with a husband could provide connection, identity validation and protection against the belief that one is a burden, but the observational data do not identify which of these mechanisms produced the association. The null grandmother result likewise should not be read as proof of no influence in every family. Grandmothers differed in whether they were maternal or paternal, whether they lived with the mother, and whether their support was practical, emotional, conflictual or some combination.

Important limitations constrain interpretation. Relationships and suicidal ideation were self-reported, making recall, social-desirability and interviewer effects possible. Unmeasured traits such as neuroticism, acute life stresses and other structural factors could influence both relationship quality and suicidal thoughts. The outcome combined passive thoughts of death with active self-harm ideation and did not systematically quantify intent, planning or severity, while dichotomization concealed differences in frequency. A one-to-two-year gap allowed relationships to change, and transient episodes between interviews would not have been captured. Most importantly, the grandmother analysis excluded 367 women, including a disproportionately large share of those who later reported ideation. This non-random missingness limits generalizability and may have obscured an association. The findings also arise from one rural Pakistani context and may not transfer directly to settings with different residence, marriage and kinship arrangements.

Even with those caveats, the study offers a focused message for maternal suicide prevention research: the source and quality of support may matter, not merely the amount of practical help available. Depression screening and treatment remain important, but the results suggest that suicide prevention may also need to consider women’s everyday relational conditions, including how partners respond to distress and household pressures. The authors caution against importing relationship therapy directly from Western settings. Family life can be regarded as private, and outside involvement may bring shame or threaten family honor. Any intervention would therefore need community-based adaptation and input from mothers, husbands and extended family. For now, the strongest conclusion is a measured one: among mothers in this rural cohort, a more positive marital relationship was prospectively associated with a lower likelihood of reporting suicidal thoughts, while no clear association was detected for the involved-grandmother relationship.

Subject of Research: Prospective associations of marital and mother-grandmother relationship quality with maternal suicidal ideation in rural Pakistan

Article Title: Associations of marital and mother-grandmother relationship quality with maternal suicidal ideation in rural Pakistan

Article References: Lin, W.-C., Hagaman, A. K., Carias, M. E., Kachoria, A. G., Aikat, D., Frost, A., & Maselko, J. (2026). Associations of marital and mother-grandmother relationship quality with maternal suicidal ideation in rural Pakistan. SSM – Mental Health, 10, Article 100708. https://doi.org/10.1016/j.ssmmh.2026.100708

Image Credits: AI Generated

DOI: 10.1016/j.ssmmh.2026.100708

Keywords: maternal suicide, suicidal ideation, marital relationship, grandmother support, rural Pakistan, longitudinal cohort, Bachpan Study, family relationships, maternal mental health, Associations, marital, mother-grandmother