Inverse relationship between cancer and Alzheimer's disease: a systemic review meta-analysis

Alzheimer's disease (AD) and cancer are both prevalent in the elderly. Some epidemiological researches have reported the negative association between AD and cancer, but the results are controversial. The present meta-analysis is aimed to clarify the association between cancer and AD. PubMed, Web of knowledge and the Cochrane library databases were searched for eligible publications. The analysis indicated that history of cancer was associated with a reduced risk of AD (ES 0.62, 95 % CIs 0.53-0.74; p < 0.001), with no significance between-study heterogeneity and publication bias. Similar results were found in subgroup analysis by stratifying variables with education and APOE epsilon 4 carriers, years of follow-up and sample size of cases. The negative association was also found in analysis of risk of cancer among patients with AD (ES 0.59, 95 % CIs 0.42-0.82; p = 0.002), but with evidence of between-study heterogeneity and publication bias. In order to identify sources of the heterogeneity, subgroup analysis was performed by stratifying variable with or without education adjusted, sample size of cases and years of follow-up. Negative association was found in all subgroup analysis except in studies with less than 5-year follow-up and with heterogeneity disappeared only in the subgroup analysis stratified with sample size of cases. Our results in the present meta-analysis support the negative association between AD and cancer. But further well-designed perspective studies with strict control of confounding factors are needed to clarify the association between AD and cancer.