Do perceived control and time orientation mediate the effect of early life adversity on reproductive behaviour and health status? Insights from the European Value Study and the European Social Survey

The study investigates whether perceived control and time orientation mediate the relationship between early life adversity and a proposed trade-off between reproductive investments (earlier age at first childbirth, higher number of children) and maintenance (poorer health). Against a backdrop of rising socioeconomic inequality and cumulative adversity (economic scarcity, family instability, violence, pollutants, malnutrition) that shapes biological and behavioral development, prior work suggests individuals from adverse environments show poorer health and more present-oriented, risky behaviors, alongside earlier and higher reproduction. Life History theory posits limited resources force trade-offs between reproduction and somatic maintenance, potentially differing by sex due to differential reproductive costs. The authors hypothesize that lower perceived control and greater present orientation, arising from harsh and unpredictable early environments, drive a shift toward immediate reproductive payoffs at the expense of long-term health. They test two hypotheses across sexes: (1) early life adversity relates to a reproduction–maintenance trade-off (poorer health, earlier reproduction, more children); (2) this association is mediated by perceived control (EVS) or time orientation (ESS).

Prior research links early adversity to long-term health risks (e.g., obesity, diabetes, cardiovascular, autoimmune diseases) and to behaviors detrimental to health (higher tobacco, alcohol, cannabis use; less exercise). Early adversity is also associated with accelerated/reprioritized reproduction (earlier puberty, earlier first birth, more children). Life History theory provides a framework for energy allocation trade-offs. Perceived control and locus of control vary with social class and adversity, with lower SES linked to more external locus and lower perceived control; perceived extrinsic mortality risk can mediate SES effects on health effort. Early-life harshness and unpredictability foster schemas of unpredictability (e.g., future discounting relations with risk taking) and may shape parental and mating efforts. However, an integrated empirical test linking early adversity, perceived control, time orientation, reproductive timing/fertility, and health within a single multivariate framework has been lacking. The study situates its hypotheses within these literatures, noting possible sex differences and the need to model cumulative adversity via composite indicators.

Design and data: Cross-validated structural equation modeling (SEM) using two large-scale surveys: EVS 4th wave (2008; 46 European countries; final N = 43,084; females 25,341; males 17,743) and ESS 9th round (2018; 40 countries; final N = 31,065; females 17,639; males 13,426). Respondents with excessive missingness (+2 SD), missing age, or no children were excluded in main analyses; control analyses later included childless respondents. Missing data: EVS used multiple imputation (mice in R): 20 datasets via fully conditional specification (logistic for categorical, proportional odds for ordinal). ESS used survey weights with a custom lavaan.survey-based approach. SEM specification: Two latent constructs: (1) Early life adversity (modeled as an emergent composite latent via unknown-weight composite) with three indicators; (2) Reproduction–maintenance trade-off (reflective latent) with three indicators. Mediators measured directly: Perceived control (EVS) and Time orientation (ESS). Ordinal items with ≥5 points were treated as continuous. Indicators: - Early life adversity (both datasets): Economic capital deprivation; Human capital deprivation; Experienced mortality in childhood. EVS economic deprivation from parental financial difficulty items at age ~14; human capital deprivation from parental reading/news/politics engagement plus parents’ education; experienced mortality from parental death before age 14. ESS economic deprivation from parental employment/occupational status at 14 (with recoding); human capital from parents’ education; experienced mortality from parental absence/death proxies. All three indicators z-transformed and normalized 0–10. - Reproduction–maintenance trade-off: Health (subjective self-rated health, 1 very good to 5 very poor), Age at first reproduction (first child’s birth year minus respondent birth year), Number of children. A residual covariance was allowed between age at first reproduction and number of children. - Mediators: EVS perceived control (10-point freedom of choice/control item; higher = more control). ESS time orientation (11-point item, 0 = plan for future, 10 = take each day; higher = more present orientation). Adjustment variables: Health, age at first reproduction, number of children, perceived control, and time orientation were adjusted for current age and household income at interview. Model identification and scaling: Reproduction–maintenance trade-off latent variance fixed to 1. Early life adversity composite scaled by fixing the path from economic capital deprivation to 1. Structural paths: Early life adversity → Mediator; Mediator → Reproduction–maintenance trade-off; Early life adversity → Reproduction–maintenance trade-off. This enables estimation of direct, indirect (mediated), and total effects. Models estimated separately for males and females. Estimation and fit: WLSMV estimator; fit assessed by scaled χ², CFI, RMSEA, SRMR. ESS incorporated complex survey weights per ESS guidelines. Cross-validation: Stratified 5-fold CV repeated 10 times (50 iterations). At each iteration, model trained on 80% and evaluated on 20% test data. A permutation test (indicator-wise randomization) assessed whether model structure fails to predict when internal structure is destroyed. Results summarized by medians (and MADs) across iterations. Visualization and additional analyses: For visualization, models were estimated per imputed dataset (ignoring CV) to extract latent scores and then averaged. Exploratory mixed-effects models assessed nonlinearity in associations between mediators and adversity/trade-off (linear vs quadratic vs cubic). Control analyses included childless respondents using ML with FIML (due to FIML requirement), re-estimating models with the same pipeline.

Model fit and robustness: - EVS (test sets): Excellent fit in both sexes: CFI 0.970 (males), 0.972 (females); RMSEA 0.020 (males), 0.022 (females); SRMR 0.020 (males), 0.021 (females). Randomized data fits were poor, supporting validity of latent structure. - ESS (test sets): Excellent fit: CFI 0.971 (males), 0.989 (females); RMSEA 0.014 (males), 0.010 (females); SRMR 0.025 (males), 0.022 (females). Randomized fits poor. Female models showed overall better fit indices. Measurement model highlights: - Early life adversity composite: Economic and human capital deprivation contributed strongly. Experienced mortality had weak or nonsignificant weight in EVS (both sexes) and in ESS males; it contributed significantly in ESS females. - Reproduction–maintenance trade-off: Positively associated with poorer subjective health and younger age at first reproduction in all models. Number of children loaded significantly in EVS females and in ESS males and females; EVS males showed no significant loading of number of children in main analysis. Structural paths and mediation: - EVS (Perceived control mediator): - Males: Early adversity → lower perceived control (std negative); perceived control → trade-off (lower control associated with more reproduction/poorer health). Indirect effect significant but small (~8.4% of total effect). Total effect of adversity on trade-off positive. - Females: Similar directions with generally stronger or equal effect sizes; indirect effect significant but small (~4.5% of total effect). Early adversity associated with lower perceived control; lower control associated with greater reproduction–maintenance trade-off. - ESS (Time orientation mediator): - Males: Early adversity not significantly related to time orientation or trade-off. However, present orientation was positively associated with the trade-off (more present orientation → more reproduction/poorer health). Indirect/total effects not meaningful due to absent adversity effects. - Females: Early adversity → more present orientation; present orientation → higher trade-off. Indirect effect significant but small (~2.9% of total effect). Total effect positive. Nonlinearities: - Best-fitting associations: Perceived control with adversity and trade-off followed cubic (degree 3) trends; time orientation with adversity and trade-off followed quadratic (degree 2) trends, especially pronounced at extreme mediator values. Control analyses (including childless respondents with FIML): - Fit indices similar to main results. In EVS males, number of children significantly loaded on the trade-off latent. In ESS males, experienced mortality contributed to adversity and adversity effects on time orientation and trade-off reached significance; indirect and total effects became significant with somewhat larger effect sizes. Overall findings robust to inclusion of childless respondents. Overall: Early life adversity, perceived control, and time orientation each associated with a reproduction–maintenance trade-off favoring reproduction over health, but mediating effects of perceived control/time orientation on the adversity–trade-off link are statistically significant yet small.

The study tested whether perceived control (EVS) and time orientation (ESS) mediate the link between early life adversity and a reproduction–maintenance trade-off. Results support the existence of the trade-off and its association with adversity, particularly via economic and human capital deprivation. Perceived control and time orientation relate to the trade-off as predicted (less control or more present orientation associated with earlier reproduction, more children, poorer health), but they account for only a small fraction of the adversity effect. This suggests relative independence among these constructs or that the measures used do not capture the full psychological mechanisms linking adversity to life-history-related behaviors. Sex differences emerged: models fit better in females and number of children was more consistently part of the trade-off latent in women, consistent with higher direct maintenance costs of reproduction for women (e.g., parity-related health costs). Experienced mortality contributed less consistently to adversity in European samples, possibly due to restricted variance or differences between objective mortality events and subjective deprivation measures. Unexpected nonlinear associations between mediators and adversity/trade-off may reflect either methodological artifacts (e.g., scale-end responses) or meaningful patterns (e.g., potential competencies developed under adversity that influence perceived control or present orientation at extremes). The findings refine life history-inspired hypotheses by showing that while perceived control and time orientation align with trade-off behaviors, they do not strongly mediate adversity’s impact in these European datasets, indicating other mechanisms or more nuanced constructs may be involved.

Using cross-validated SEMs on EVS and ESS, the study found that early life adversity, perceived control, and time orientation each relate to a reproduction–maintenance trade-off characterized by earlier reproduction, higher fertility, and poorer health. However, perceived control (EVS) and time orientation (ESS) mediate only a small portion of the adversity–trade-off association, with stronger and more consistent effects in women. These results partially support life history-inspired predictions but highlight that commonly used single-item measures of perceived control and time orientation and European samples may limit detection of stronger mediation. Future work should develop richer, multidimensional measures of these psychological constructs, incorporate broader and more diverse populations with greater adversity variability, and leverage longitudinal/causal designs to establish temporal precedence and mechanisms.

Key limitations include: (1) Correlational, cross-sectional survey data preclude causal inference; reverse or bidirectional associations are possible. (2) Measures of perceived control and time orientation are single-item proxies that may not capture multidimensional constructs (e.g., locus of control, self-efficacy, learned helplessness), potentially attenuating associations. (3) Early adversity composite relies on retrospective and limited indicators (especially experienced mortality tied only to parental death/absence), possibly underrepresenting extrinsic mortality and unpredictability. (4) Samples are restricted to European, industrialized contexts (WEIRD), limiting generalizability; restricted variance in some indicators may weaken effects. (5) Within-person variability over time exists in perceived control and time orientation, reducing cross-sectional mediation strength. (6) Potential methodological artifacts could contribute to observed nonlinearities (e.g., end-of-scale responding). (7) Exclusion of childless respondents in main analyses may affect fertility-related measurement; control analyses mitigate this concern but employed different estimators (ML/FIML).