Abstract
Objective Medically unexplained symptoms (MUS) are common and associated with high consumption of health care resources. Cross-sectional studies in selected and clinical populations show consistent linkages between history of childhood sexual abuse (CSA) and presentation with MUS and somatization. However, there are almost no well-controlled prospective studies in population samples. Methods Data were gathered in a longitudinal study of a New Zealand birth cohort born in 1977. Hospital e-record data for the period 2008-2015 (age, 30-38 years) were searched for a subsample of 408 study participants who were resident in one District Health Board region, and details of MUS contacts were recorded. Retrospective reports of CSA (<16 years) were obtained at ages 18 and 21 years. Associations between CSA and MUS were sequentially adjusted using logistic regression methods for both childhood confounders assessed before age 16 years and potential mediating mental health/family context up to age 30 years. Results Twenty (4.9%) participants were classified as having MUS, of whom 11 had a history of CSA. Severe CSA involving attempted/completed sexual penetration was strongly associated with risk of MUS (odds ratio = 11.6, 95% confidence interval = 4.3-31.7, p < .001). A substantial association remained after statistically adjusting for confounding and mediating mental health/family context (adjusted odds ratio = 5.1, 95% confidence interval = 1.2-21.3, p = .024). This strong association was specific to severe CSA (as opposed to childhood physical abuse) and to MUS rather than medically explained symptoms. Conclusions CSA involving attempted/completed penetration was strongly associated with attendance at secondary level care for MUS. Implications for prevention and treatment of MUS are discussed.