Evidence from Finland and Sweden on the relationship between early-life diseases and lifetime childlessness in men and women

The study addresses rising lifetime childlessness and the extent to which early-life diseases contribute to remaining childless by the end of the reproductive lifespan. Contextual factors include increased educational and labor market participation, economic uncertainty, partnership dynamics, and delayed childbearing that lowers fecundity and raises infertility issues. Prior work has focused on a narrow set of conditions (for example, reproductive disorders, IBD, multiple sclerosis, psychiatric disorders) and often only on women, using limited time horizons and with insufficient control for familial confounding. The authors aim to provide a comprehensive, disease-agnostic assessment across 414 diagnoses, examining sex differences, age-of-onset effects, parity stratification, and the mediating roles of partnership formation and education, using high-quality Finnish and Swedish registers and a sibling-based matched design to address unmeasured familial confounding.

Previous research largely examined singular diseases directly tied to reproductive biology (for example, recurrent miscarriage, polycystic ovary syndrome, endometriosis) or conditions such as IBD, multiple sclerosis, and major psychiatric disorders. Some evidence suggests a portion of individuals with IBD choose voluntary childlessness due to concerns about pregnancy and disease, and MS has been linked to higher childlessness via partnership and treatment considerations. Psychiatric disorders, including schizophrenia, have been associated with reduced fecundity and reproductive success. For men, evidence is comparatively limited and often clinic-based (for example, semen samples from infertility clinics). Existing studies lack a systematic, comparative evaluation across many diseases, have restricted follow-up, and limited capacity to control for familial confounding, leaving gaps on sex-specific effects and broader disease contributions to lifetime childlessness.

Design and data sources: Nationwide Finnish and Swedish population registers were used, covering index persons, their parents, spouses/partners, siblings, and children. Cohorts included all individuals born in Finland and Sweden in 1956–1968 (men) and 1956–1973 (women), alive by age 16, not emigrated, and followed to 2018 (women to age 45; men to age 50). Data included: - Childbearing from national population and Medical Birth Registers (excluding ART conceptions; 0.3% Finland, 0.8% Sweden). - Partnership/marriage from national registers (Finland: 1971–2018; Sweden: marriages and cohabiting unions with biological children, 1977–2017). - Disease diagnoses from hospital inpatient and specialist outpatient registers via ICD-8/9/10 codes (plus Finnish Cancer Registry), with age of onset defined as first recorded diagnosis. Mortality data with causes of death were also included. Exposure definitions: 414 diseases across 16 categories (infectious-parasitic and congenital anomalies defined only in Finland; others in both countries). Highly correlated endpoints (tetrachoric correlation >0.7) were pruned. Outcomes: Primary outcome was lifetime childlessness (no live-born children by 45 in women, 50 in men). Secondary outcome was singlehood (no registered marriage/partnership by these ages). Main analysis (sibling-based): Same-sex full-sibling pairs discordant for childlessness were identified (71,524 full-sister and 77,622 full-brother pairs). Within each family, one childless sibling was randomly selected as case and the closest-in-birth-order sibling with children as control. Disease diagnoses counted only if onset was at least one year before the first birth of the sibling with children (or the corresponding age in the childless sibling). Conditional logistic regression (clogit) adjusted for birth year estimated odds ratios (ORs) for disease-childlessness associations separately by sex. Bonferroni correction controlled familywise error (women: P<1.5×10^-4 for 328 diseases; men: P<1.5×10^-4 for 325 diseases). Estimates were meta-analyzed across countries. Age-of-onset analyses: For diseases significantly associated with childlessness and with adequate sample size (30 in women, 31 in men), age-of-onset categories (≤15, 16–20, 21–25, 26–30, 31–35, 36–40, 41–45, plus 46–50 for men) were modeled as fixed effects to assess onset-dependent associations. Parity stratification: Childless cases were compared with siblings having exactly one child, exactly two children, or three or more children. Mediation by singlehood: A causal mediation analysis partitioned total effects into indirect effects via singlehood and direct effects, and disease effects on singlehood were estimated. Population-based secondary analysis: A nested incident-matched case-control design matched each childless case to a control on sex, birth year, municipality of birth, and highest parental education level. Conditional logistic regression produced population-level ORs. Sensitivity analyses: - Restricted to individuals alive at end of follow-up to assess reproductive-age mortality effects. - Stratified Cox proportional hazards models treating disease as time-varying exposure (age as time scale). - Adjustment for individuals’ own highest education in sibling- and population-based analyses. Statistical software: R (v4.1.2), survival and meta packages.

Study population: 1,425,640 women and 1,119,380 men included; 230,198 women (16.6%) and 279,454 men (25.4%) were childless, higher in men and in Finland. Main associations: - Mental-behavioural disorders showed the strongest average associations with childlessness (women: OR = 3.1, 95% CI 2.6–3.7; men: OR = 3.2, 95% CI 2.6–4.0) with substantial heterogeneity. • Mild intellectual disability had the strongest single-diagnosis association (men: OR = 21.7, 95% CI 10.8–43.5). • Mood disorders showed smaller effects (men: OR = 2.1, 95% CI 1.8–2.3). - Severe brain diseases were strongly associated: • Cerebral palsy (women: OR = 13.4, 95% CI 7.9–22.7; men: OR = 12.0, 95% CI 6.5–22.2). • Malignant brain neoplasms (women: OR = 5.5, 95% CI 2.8–10.8; men: OR = 5.0, 95% CI 2.8–8.9). - Endocrine–nutritional–metabolic disorders were strongly associated (women: OR = 1.4, 95% CI 1.1–1.8; men: OR = 2.0, 95% CI 1.6–2.6). • Obesity (women: OR = 1.8, 95% CI 1.5–2.4; men: OR = 2.7, 95% CI 2.0–3.6). • Type 1 diabetes (women: OR = 2.3, 95% CI 2.0–2.6; men: OR = 1.7, 95% CI 1.5–2.0). • Type 2 diabetes in women: OR = 4.2, 95% CI 2.2–8.1. - Inflammatory/autoimmune diseases across systems associated with increased childlessness in women: • Pneumonia OR = 1.4 (95% CI 1.3–1.6), myocarditis OR = 3.7 (95% CI 2.3–6.1), chronic tubulo-interstitial nephritis OR = 1.8 (95% CI 1.3–2.3), multiple sclerosis OR = 2.1 (95% CI 1.5–3.1), systemic lupus erythematosus OR = 2.7 (95% CI 1.6–4.6), juvenile idiopathic arthritis OR = 3.5 (95% CI 2.6–4.9). - Some conditions associated with reduced odds of childlessness (both sexes): chronic diseases of tonsils/adenoids and acute appendicitis (for tonsils/adenoids: women OR = 0.84, 95% CI 0.80–0.88; men OR = 0.84, 95% CI 0.80–0.89). Sex-specific effects: - Congenital anomalies showed stronger associations in women (OR = 3.5, 95% CI 2.6–4.9) than men (OR = 1.8, 95% CI 1.3–2.5). • Malformations of digestive system (women OR = 4.8, 95% CI 2.2–10.3; men OR = 1.7, 95% CI 1.0–3.0). • Musculoskeletal system (women OR = 3.5, 95% CI 2.0–6.1; men OR = 1.2, 95% CI 0.7–1.9). - Schizophrenia had stronger associations in men (women OR = 11.6, 95% CI 9.1–14.9; men OR = 20.8, 95% CI 16.1–27.0). - Acute alcohol intoxication stronger in men (women OR = 2.2, 95% CI 1.6–2.9; men OR = 4.7, 95% CI 3.8–5.7). - Type 1 diabetes stronger in women (women OR = 2.3 vs men OR = 1.7). Age-of-onset effects: - Nonlinear pattern with strongest associations when first diagnosed at ages 21–25 in women (OR = 3.1, 95% CI 2.5–3.8) and 26–30 in men (OR = 3.1, 95% CI 2.4–3.9); smaller effects at earlier and especially later onset. - Example: female obesity onset at 16–20 associated with higher childlessness (OR = 3.0, 95% CI 1.6–5.6) than onset at 26–30 (OR = 1.1, 95% CI 0.5–2.7). Parity stratification: - Comparing childless individuals to siblings with exactly one child reduced ORs for several diseases (14 in women, 6 in men). Example: schizophrenia in women decreased from OR 11.6 (95% CI 9.1–14.9) to 5.1 (95% CI 3.7–7.0). Mediation by singlehood: - 83.0% of women and 77.0% of men had registered partners by the end of reproductive age; among childless individuals, only 36.3% (women) and 29.4% (men) had partners. - Disease effects on singlehood highly correlated with those on childlessness (women R^2 = 0.71; men R^2 = 0.85). - Mediation analysis: median proportion mediated by singlehood was 29.3% in women and 37.9% in men. • Schizophrenia in women: indirect (singlehood-mediated) OR = 2.1 (95% CI 2.0–2.2); direct OR = 6.8 (95% CI 5.0–9.2). • Hypertension in women: indirect OR = 1.2 (95% CI 1.1–1.4); direct OR = 2.4 (95% CI 1.6–3.5). Partnered sub-analyses: - Among partnered individuals, several diseases remained associated with childlessness (6 in women, 11 in men), though effect sizes were generally reduced. - Notable retained associations: obesity in men (all men OR = 2.7; partnered men OR = 3.3, 95% CI 1.9–5.7); type 1 diabetes in women (all women OR = 2.3; partnered women OR = 2.5, 95% CI 2.0–3.2). Population-based and sensitivity: - Population-based estimates were broadly similar to sibling-based; stronger associations observed in men for alcohol dependence (sibling OR = 2.9 vs population OR = 3.8) and mood disorders (2.1 vs 2.7). - Restricting to those alive at end of follow-up showed some attenuation for men: acute alcohol intoxication (OR 4.7 to 3.9) and subarachnoid hemorrhage (OR 2.9 to 0.9). - Cox models yielded similar patterns; adjusting for individuals’ education in population-based models reduced ORs for many diseases, with exceptions (for example, obesity OR increased from 3.2 to 4.4). Overall, 74 of 403 diseases were significantly associated with childlessness in at least one sex after multiple-testing correction, with particularly strong effects for mental-behavioural disorders, congenital anomalies, and endocrine–nutritional–metabolic disorders, and novel associations for inflammatory/autoimmune conditions.

The findings demonstrate that a broad set of early-life diseases are associated with increased lifetime childlessness, beyond conditions directly tied to reproductive biology. The sibling-based design reduces confounding from shared familial factors, and results were consistent across two countries. Strong sex differences emerged: mental-behavioural disorders and alcohol-related diagnoses were more strongly linked to male childlessness, whereas diabetes-related diseases and congenital anomalies were more impactful for women. The age at first diagnosis showed peak associations around young adulthood (women 21–25; men 26–30), aligning with sex differences in union formation timing. Partnership formation played a major mediating role—especially for men—indicating that disease may reduce partnership chances, thereby increasing childlessness. Nevertheless, several conditions (for example, obesity in men, type 1 diabetes in women) remained associated among partnered individuals, suggesting direct biological effects on fertility or pregnancy, or effects on timing and stability of family formation. The results enrich understanding of mechanisms underlying involuntary childlessness and point to targets for public health interventions and counseling, including mental-health and metabolic conditions, and attention to partnership dynamics.

This study offers a comprehensive, disease-agnostic assessment of associations between early-life diseases and lifetime childlessness across Finnish and Swedish populations, identifying strong and sex-specific links—particularly in mental-behavioural, congenital, and endocrine–nutritional–metabolic categories—and novel associations with inflammatory and autoimmune diseases. Age-of-onset patterns and mediation by singlehood underscore the interplay between health, partnership, and reproduction. These findings provide an evidence base to prioritize preventive and therapeutic interventions to reduce involuntary childlessness and inform clinical counseling. Future research should disentangle voluntary vs involuntary pathways with direct data on fertility intentions, examine treatment effects and institutionalization, and assess generalizability to more recent cohorts and other countries with different healthcare and social contexts.

- Disease data primarily from secondary/tertiary care registers, likely capturing more severe cases and lacking primary care information (for example, milder metabolic conditions). - Diagnostic practices and ICD coding changed over time; harmonization across ICD-8/9/10 may affect accuracy. - Cohorts born 1956–1973; results may not fully generalize to later cohorts with changing partnering, reproductive behavior, and medical treatments. - Inability to distinguish voluntary from involuntary childlessness due to lack of data on intentions; estimates should consider that a minority are voluntarily childfree. - Possible bias from assortative mating among partnered couples. - Residual confounding and reverse causality cannot be completely excluded despite sibling design; siblings may experience different environments, and selection to same-sex sibling pairs may introduce bias. - Generalizability beyond Nordic countries may be limited, though patterns align with trends in other industrialized nations. - Some analyses limited by rarity of diagnoses among partnered individuals; medication and treatment effects (for example, antiepileptics, immunosuppressants) not fully captured.