Childhood disadvantage leaves a health mark at 60 that adult success cannot erase

…

The 1958 National Child Development Study stands among the most valuable longitudinal resources available for studying how early-life circumstances shape health decades later. Because its members were all born during a single week in Britain, the cohort minimises the seasonal and generational variation that can complicate comparisons across birth dates, while its repeated sweeps at birth, age 33, age 55, and ages 60–66 allow researchers to reconstruct socioeconomic trajectories with unusual temporal precision. This matters for mediation analysis in particular, since distinguishing cause from correlation requires that potential mediators be measured after the exposure but before the outcome. By anchoring childhood socioeconomic position in paternal occupational class at birth and adult socioeconomic position in circumstances recorded at age 55, the study design preserves the chronological ordering that causal inference demands.

The choice of health outcomes reflects a deliberate balance between subjective and objective indicators. Self-rated health, despite its simplicity, has accumulated decades of evidence showing that it independently predicts subsequent morbidity and mortality, capturing an integrated assessment of bodily and mental wellbeing that clinical measures may miss. The four biomarkers each probe a distinct physiological system relevant to ageing. C-reactive protein serves as a marker of systemic low-grade inflammation, a process repeatedly implicated in the biological embedding of early adversity. The cholesterol ratio captures lipid-related cardiovascular risk, glycated haemoglobin reflects longer-term glycaemic regulation, and NT-proBNP indexes cardiac strain, a hormone fragment released when the heart muscle is under haemodynamic stress. Together these measures span inflammatory, metabolic, lipid, and cardiovascular pathways, offering a broad physiological portrait of health in early older age.

Several methodological refinements strengthen the credibility of the biomarker findings. Excluding C-reactive protein values above 10 mg/L guards against the distortion that acute infections introduce into inflammation measurements. Add-back corrections for participants taking lipid-lowering or glucose-regulating medications address a subtle but important bias: without such corrections, treated individuals would appear healthier than they truly are, since medication artificially lowers the measured biomarker while the underlying disease process persists. Log-transforming NT-proBNP accommodates its strongly right-skewed distribution. These decisions reflect growing recognition within social epidemiology that pharmacological treatment can mask, rather than reverse, socioeconomic gradients in underlying physiological risk.

The interventional mediation framework employed in the study represents a meaningful advance over conventional mediation approaches. Traditional path-based methods typically assume that there is no interaction between the exposure and the mediator and no confounding of the mediator–outcome relationship other than variables measured before the exposure. Both assumptions are difficult to satisfy in life-course research. Socioeconomic circumstances at different life stages plausibly interact, with the health consequences of adult disadvantage differing depending on childhood origins. Moreover, adult SEP itself can influence later confounders such as health behaviours, which in turn affect health outcomes, creating exposure-induced mediator–outcome confounding that violates standard assumptions. Interventional effects sidestep these problems by defining the indirect effect as the difference in outcomes that would occur if, hypothetically, the mediator distribution among the exposed were set to what it would be under the unexposed, holding other factors fixed. This yields policy-relevant quantities: the expected change in health outcomes if adult socioeconomic circumstances were equalised across childhood origins, without requiring the stronger and less plausible assumptions of conventional mediation.

The finding that mediation through adult socioeconomic position is evident for self-rated health but modest for C-reactive protein and minimal for the remaining biomarkers invites careful interpretation. Self-rated health is a holistic judgement that incorporates lived experience, functional limitations, psychological wellbeing, and expectations about health, all domains plausibly shaped by socioeconomic circumstances throughout life. Biomarkers, by contrast, may reflect physiological processes laid down earlier or influenced by pathways not captured by the four measured adult SEP dimensions. The divergence between subjective and objective outcomes suggests that the mechanisms linking childhood disadvantage to later health are heterogeneous, operating partly through attainable adult circumstances for some outcomes and through other channels, including direct biological embedding, health behaviours, psychosocial stress, or unmeasured adult conditions, for others.

The substantial direct effects that persist after accounting for adult socioeconomic pathways carry significant policy implications. If adult occupational class, education, housing tenure, and income explained most of the childhood gradient, then interventions targeting adult circumstances could largely compensate for early disadvantage. The evidence points in a different direction: equalising adult socioeconomic circumstances would reduce but not eliminate health inequalities rooted in childhood. This supports a life-course policy perspective in which early-childhood investments, such as reducing child poverty, supporting families, and improving early educational environments, complement rather than substitute for adult-focused interventions. Neither alone appears sufficient.

The four dimensions of adult SEP were examined separately, which allows their relative contributions to be compared rather than collapsing them into a single index. This distinction matters because different dimensions may operate through different mechanisms. Education, typically completed well before midlife, may shape health through knowledge, cognitive resources, and employment opportunities. Occupational class reflects working conditions, job security, and workplace exposures. Housing tenure captures wealth accumulation and residential stability. Income determines material resources available for healthcare, nutrition, and living conditions. Disentangling these dimensions helps identify which levers might be most consequential for which health outcomes.

The inclusion of grandparental social class and ethnicity as baseline confounders acknowledges that socioeconomic position is transmitted across generations. Grandparental class captures intergenerational social advantage that precedes and potentially shapes both the cohort members’ childhood circumstances and their subsequent trajectories. Adjusting for these variables strengthens the interpretation of childhood SEP effects as being less confounded by deeper ancestral advantage, though residual confounding can never be fully excluded in observational research.

Placing these findings in the context of the broader literature, British birth cohorts have repeatedly documented strong socioeconomic gradients in premature mortality, with individuals from disadvantaged social class origins facing roughly 1.7 to 2.7 times the risk of death in adulthood and early old age compared with those from advantaged origins. Comparable gradients appear in the United States and Northern Europe, suggesting that the phenomenon is not unique to any single welfare regime. The persistence of childhood-origin health inequalities into early older age, a life stage with direct implications for work capacity, economic security, and the transition to retirement, underscores the stakes involved. Health inequalities at ages 60 to 66 influence how long people can remain economically active, the adequacy of their retirement resources, and the quality of the years that follow.

The life-course theoretical frameworks invoked by the study help organise these empirical patterns. The critical-period model holds that exposures during specific developmental windows have lasting effects that exposures at other times cannot replicate. Accumulation models emphasise the additive wear of disadvantage across the life course. Pathway models highlight how early circumstances channel individuals into later circumstances that themselves affect health. The persistence of direct effects alongside measurable indirect effects through adult SEP is consistent with all three models operating concurrently: childhood disadvantage leaves marks that adult circumstances do not fully erase, accumulates with ongoing disadvantage, and propagates forward through attainable socioeconomic positions.

It is worth noting the practical significance of the biomarker choices for understanding ageing biology. Systemic inflammation, indexed by C-reactive protein, is thought to be one conduit through which early adversity becomes biologically embedded, potentially accelerating age-related disease processes. Glycaemic regulation and lipid profiles track cardiometabolic risk that develops over decades. NT-proBNP, though less commonly used in social epidemiology, offers a window into subclinical cardiac strain that may precede symptomatic cardiovascular disease. The finding that these objective markers show weaker mediation through adult SEP than self-rated health does may indicate that physiological embedding of childhood disadvantage operates substantially outside the socioeconomic pathways that adult circumstances represent, or that the relevant adult mediators for biomarker outcomes were not fully captured by the four dimensions measured.

Finally, the study’s extension of mediation evidence from midlife into early older age fills an important evidentiary gap. Previous formal quantification using the same cohort had focused on frailty in midlife with a single latent measure of adult SEP. Whether the contribution of adult socioeconomic circumstances grows, shrinks, or remains stable as cohorts age has direct bearing on when interventions might yield the greatest benefit. The evidence that measured adult circumstances explain only a small part of childhood inequalities in early older age suggests that by this life stage, the health consequences of early disadvantage have already been substantially consolidated, reinforcing the case for policies that act early while also continuing to address socioeconomic disadvantage throughout adulthood.

Subject of Research: Life-course socioeconomic position and health in older adulthood

Article Title: Life-course socioeconomic position and health in older adulthood: a causal mediation analysis in the 1958 British Birth Cohort

Article References: Guo, Y., Pelikh, A., Ploubidis, G. B., & Goodman, A. (2026). Life-course socioeconomic position and health in older adulthood: a causal mediation analysis in the 1958 British Birth Cohort. The Lancet Regional Health – Europe, 71, Article 101890. https://doi.org/10.1016/j.lanepe.2026.101890

Image Credits: AI Generated

DOI: 10.1016/j.lanepe.2026.101890

Keywords: socioeconomic position, life-course epidemiology, 1958 British Birth Cohort, causal mediation, self-rated health, biomarkers, health inequalities, ageing, NCDS, public health, Life-course, socioeconomic