AbstractBackgroundThe objective of this research was to examine associations between diabetes mellitus (DM) and neurocognitive impairment in survivors of childhood cancer while exploring mechanistic associations with treatment exposures and moderating associations with cardiovascular disease (CVD) and risky lifestyle factors.MethodsSurvivors of the Childhood Cancer Survivor Study (N = 16,196; mean ± standard deviation age, 32.9 ± 7.9 years; 50.2% female; N = 615 with DM) self‐reported neurocognitive functioning, risky drinking, physical inactivity, and tobacco use. Medical data were collected from chart review. Multivariable regression determined the association of DM with neurocognitive impairment while adjusting for relevant treatment exposures. Interactions between DM and treatment, risky lifestyle, and cardiovascular predictors on impairment were explored. Path analysis was used to examine the effects of treatment exposures through DM and CVD on impairment. Among survivors with DM, multivariable regressions examined predictors of neurocognitive change over time (mean, 11.21 years).ResultsSurvivors with DM had an increased risk of impairment relative to survivors without DM (task completion: odds ratio [OR], 1.5; 95% confidence limits [CI], 1.2–1.9; emotion regulation: OR, 1.4; 95% CI, 1.1–2.0; and organization: OR, 1.5; 95% CI, 1.1–2.0). The effects of cranial radiation on neurocognition were mediated by DM, including task completion (β = 0.02), emotion regulation (β = 0.02), memory (β = 0.01), and organization (β = 0.02; all p < .01). Among survivors with DM, CVD was associated with declines in task completion (estimate = 0.44; p < .01) and organization (estimate = 0.27; p = .03).ConclusionsSurvivors with DM are at increased risk of neurocognitive impairment. Although CVD did not exacerbate concurrent risk for impairment, it was associated with a decline in neurocognitive functioning over time in survivors with DM. Preventing/managing CVD in survivors with DM could mitigate additional neurocognitive decline.