Associations between childhood maltreatment and mid-life health risks and risky health behaviors in a nationally representative sample

Millions of U.S. children experience maltreatment each year, which is associated with subsequent health risks, including psychopathology. We use data from up to 9692 participants in the National Longitudinal Study of Adolescent to Adult Health (Add Health) to update and extend knowledge about such health risks into the start of midlife. These data include detailed retrospective measures of supervisory neglect, physical neglect, physical abuse, and sexual abuse, and also include data on health and health-relevant behaviors that were self-reported (general health, depression measured as CES-D negative affect subscale, smoking, alcohol use, and illegal drug use) and directly measured (BMI, hypertension, renal function, and cardiovascular disease risk) at ages 32-42. We use these data to (1) provide the strongest evidence to date about the persistence of health risks among formerly maltreated individuals approaching midlife and (2) examine the prevalence of such risks among two particularly vulnerable groups: those who experienced multiple types of maltreatment, and those who experienced frequent maltreatment. Our analyses reveal depression and smoking as salient midlife health risks among individuals who experienced physical neglect and sexual abuse. We also find evidence that multi-victimization and recurring maltreatment increases such risks but find no evidence of linear dose-response relationships between maltreatment frequency and health. Findings are robust to the inclusion of covariates that control for other individual-, family-, and neighborhood-level risk factors. • Childhood maltreatment was associated with multiple mid-life health risks • Associations with depression and smoking behavior were particularly strong • Children exposed to physical neglect and sexual abuse were most vulnerable • Multi-victimization and frequent maltreatment raised the likelihood of health risks • We did not find evidence of linear dose-response relationships