Association of birth weight with incident dementia: The Atherosclerosis Risk in Communities (ARIC) Neurocognitive Study

Abstract

AbstractBackgroundEarly life factors, such as birth weight (BW), may have long‐term health consequences, but the importance of these factors in dementia risk is poorly understood. Our objective was to prospectively examine the association of BW with dementia incidence over mid‐ to late‐adulthood in the ARIC study. We hypothesized that participants who report lower BW experience more incident dementia than participants who report normal BW, and this association may differ by race.Method10,789 participants who were dementia‐free at ARIC baseline (Visit 1: 1987‐1989; ages 45‐64) and followed through 2019 were included. Participants were asked their BW through standard interviews at Visit 4 (1996‐1998), but if unable to recall, they selected their BW category: low (<5.5 lbs.), medium (normal) (5.5‐9.0 lbs.), or high (>9.0 lbs.), and to identify if they were premature (yes/no). Adjudicated dementia cases from baseline through 2019 were classified using in‐person cognitive evaluations, informant interviews, and phone assessments as well as hospitalization codes and death certificate codes. The associations between having high or low (vs medium) BW and dementia, or premature birth status and dementia were investigated using Cox proportional hazards models, with testing for interaction by race.ResultOur overall population was 54 yo at baseline, 56% female, and 20% Black; 2,255 dementia cases were identified over a median of 28.2 years. Participants who reported a low BW had 1.2 times (95% CI 1.02, 1.41) increased risk of incident dementia compared to those with medium BW in demographic‐ and APOE‐adjusted models (Table 1). This association was no longer significant when the model additionally adjusted for vascular risk factors. The risk for dementia was not increased among participants who reported having been born premature (vs not premature). When stratified by race, the risk of incident dementia was elevated in Black participants reporting high (vs medium) BW, but not in White participants reporting high BW (p‐interaction = 0.006) (Table 2).ConclusionIn this cohort, low BW was associated with risk of dementia in minimally adjusted models. The association was attenuated after accounting for vascular risk factors. High BW was associated with risk of dementia in Black participants only.