Age-related changes in headache days across the cognitive spectrum

Introduction. Prior studies demonstrating age-related declines in headache prevalence have not accounted for the potentially confounding effects of cognitive impairment. Our primary goal was to assess the relationship between aging and self-reported monthly headache days across the cognitive spectrum. Methods. A detailed headache questionnaire was included prospectively as part of an annual mailing to participants in the Sanders-Brown Center on Aging longitudinal cohort. Results. The overall survey response rate was 58.6%, yielding a cohort of 332 cognitively normal, 71 mild cognitive impairment (median Mini-Mental Status Examination [MMSE] score 5 27, interquartile range [IQR] 5 25-28), and 51 demented (median MMSE score 5 24, IQR 5 19-26) individuals. Current headaches were reported by 22.8%, 25%, and 27.1%, respectively, across normal, mild cognitive impairment, and dementia subgroups. A negative correlation was observed between age and average headache days in cognitively normal (q 5 -0.163, 95% confidence interval [CI] 5 -0.246 to -0.022, P 5 0.004), mild cognitive impairment (q 5 -0.255, 95% CI 5 -0.274 to 0.229, P 5 0.0475), and dementia groups (q 5 -0.295, 95% CI 5 -0.457 to 0.159, P 5 0.068). Ordinary least-squares regression with backward selection identified age alone, but not gender or MMSE, as predicting headache days in the overall cohort. Conclusions. Aging is associated with a decline in headache days in the absence of any confounding cognitive pathology and is weakly predictive of headache days across the cognitive spectrum. Whether this represents a reporting bias due to dementia or has neurobiological significance warrants further investigation.