Conflicting results have been reported regarding the association between white matter lesions (WML) and cognitive impairment. We hypothesized that education, a marker of cognitive reserve (CR), could modulate the effects of WML on the risk of mild cognitive impairment (MCI) or dementia.

We followed 500 healthy subjects from a cohort of community-dwelling persons aged 65 years and over (ESPRIT Project). At baseline, WML volume was measured using a semi-automatic method on T2-weighted MRI. Standardized cognitive and neurological evaluations were repeated after 2, 4, and 7 years. The sample was dichotomized according to education level into low (≤8 years) and high (>8 years) education groups. Cox proportional hazard models were constructed to study the association between WML and risk of MCI/dementia.

The interaction between education level and WML volume reached significance (p = 0.017). After adjustment for potential confounders, the association between severe WML and increased MCI/dementia risk was significant in the low education group (≤8 years) (p = 0.02, hazard ratio [HR]: 3.77 [1.29-10.99]), but not in the high education group (>8 years) (p = 0.82, HR: 1.07 [0.61-1.87]).

Severe WML significantly increases the risk of developing MCI/dementia over a 7-year period in low educated participants. Subjects with higher education levels were seen to be more likely to be resilient to the deleterious effects of severe WML. The CR hypothesis suggests several avenues for dementia prevention.