To the Editor: Based on a subpopulation of the Framingham cohort, Dr Lieb and colleagues1 suggested that higher levels of leptin are associated with reduced incidence of dementia and Alzheimer disease (AD), correlated with positive associations between leptin levels and total brain and hippocampal volumes. We believe that caution is needed in the interpretation of those results.
Leptin was measured in nonfasting plasma. However, leptin levels have ultradian and circadian rhythmicity,2 and measuring leptin at different times of the day creates a possible confounding factor. In addition, statistical significance for the association between leptin levels and AD was present only in nonobese individuals. Obese individuals, in spite of their hyperleptinemia, might not be protected from developing AD, possibly because of high leptin resistance in the brain. Two interesting questions emerge from these observations: if high leptin levels do not promote satiety in obese individuals, can they promote neuroprotection? If neuroprotection is not achieved by hyperleptinemia in obese individuals, are the underlying molecular mechanisms the same as those that cause neuronal resistance to leptin's satiety effects?