Measuring mortality and the burden of adult disease associated with adverse childhood experiences in England: a national survey

Abstract

Background: ACE (adverse childhood experience) studies typically examine the links between childhood stressors and adult health harming behaviours. Using an enhanced ACE survey methodology, we examine impacts of ACEs on non-communicable diseases and incorporate a proxy measure of premature mortality in England.

Methods: A nationally representative survey was undertaken (n = 3885, aged 18-69, April-July 2013). Socio-demographically controlled proportional hazards analyses examined the associations between the number of ACE categories (<18 years; e.g. child abuse and family dysfunction such as domestic violence) and cancer, diabetes, stroke, respiratory, liver/digestive and cardiovascular disease. Sibling (n = 6983) mortality was similarly analysed as a measure of premature mortality.

Results: Of the total, 46.4% of respondents reported ≥1 and 8.3% ≥4 ACEs. Disease development was strongly associated with increased ACEs (e.g. hazard ratios, HR, 0 versus ≥4 ACEs; cancer, 2.38 (1.48-3.83); diabetes, 2.99 (1.90-4.72); stroke, 5.79 (2.43-13.80, all P < 0.001). Individuals with ≥4 ACEs (versus no ACEs) had a 2.76 times higher rate of developing any disease before age 70 years. Adjusted HR for mortality was strongly linked to ACEs (≥4 versus 0 ACEs; HR, 1.97 (1.39-2.79), P < 0.001).

Conclusions: Radically different life-course trajectories are associated with exposure to increased ACEs. Interventions to prevent ACEs are available but rarely implemented at scale. Treating the resulting health costs across the life course is unsustainable.

Keywords: children; chronic disease; morbidity and mortality.

Fig. 1

Cumulative proportion of individuals not diagnosed with a major disease with age: unadjusted survival at period end. Respondents reported year of first diagnosis by a doctor or nurse with cancer, CVD, diabetes type 2, stroke, respiratory disease and liver/digestive disease. Details of what constitutes an ACE are given in Supplementary data, Box. See Methods for more details. Kaplan–Meier analysis, variation between ACE categories (Mantel–Cox, χ² = 71.671, P < 0.001).

Fig. 2

Differences in modelled cumulative morbidity and mortality between deprived males with no or ≥4 adverse childhood experiences. Figures are based on Cox regression models with survival estimates generated at 10-year intervals before age 70 years. Analyses include n = 340 (male respondents) and n = 849 (siblings) in the most deprived quintile. Morbidity estimates are adjusted to account for proportions of population who have died (see Methods). SE = standard errors, which are shown for morbidity and mortality.