With the exception of a few monogenic forms, Alzheimer disease (AD) has a complex aetiology that is likely to involve multiple susceptibility genes and environmental factors. The role of environmental factors is difficult to determine and, until a few years ago, the molecular mechanisms underlying gene–environment (G × E) interactions in AD were largely unknown. Here, we review evidence that has emerged over the past two decades to explain how environmental factors, such as diet, lifestyle, alcohol, smoking and pollutants, might interact with the human genome. In particular, we discuss how various environmental AD risk factors can induce epigenetic modifications of key AD-related genes and pathways and consider how epigenetic mechanisms could contribute to the effects of oxidative stress on AD onset. Studies on early-life exposures are helping to uncover critical time windows of sensitivity to epigenetic influences from environmental factors, thereby laying the foundations for future primary preventative approaches. We conclude that epigenetic modifications need to be considered when assessing G × E interactions in AD.

• Most forms of Alzheimer disease (AD) have a complex aetiology that is likely to involve multiple susceptibility genes and environmental factors.
• Studies in animal models and humans have shown that environmental AD risk factors, such as diet, lifestyle, alcohol, smoking and pollutants, can induce epigenetic modifications of key AD-related genes and pathways.
• Several naturally occurring antioxidants as well as caloric restriction and physical activity can reduce oxidation and prevent cognitive decline related to brain ageing and AD. Many antioxidants have been found to act throughout epigenetic mechanisms.
• The preconceptional, prenatal, perinatal and childhood phases of development are emerging as periods of sensitivity to epigenetic influences from environmental factors that can increase the risk of chronic conditions, including neurodegeneration, in adulthood.
• Epigenetic modifications must be considered when assessing gene–environment interactions in AD.