A PROMPT analysis of the study by Riviere et al (2001) that examined whether a nutritional education program could prevent weight loss in patients with Alzheimers disease found the information clearly presented in a standard format for empirical studies. The language is unbiased, specific, not overly technical and easily understood by a person with a scientific background. The authors had no obvious conflicts of interest and were sponsored by the European Health Promotion Program, a non-commercial organisation.
However, the scope of the study was small: only 151 subjects were studied for 12 months. This may be too small a cohort and too short a time to draw any significant conclusions. ...view middle of the document...
The study by Gardener et al (2012) examining the association between Alzheimer’s disease risk and the Mediterranean diet (MeDi) is relevant as the Australian cohort and UK population are closely matched. MeDi diet is widely recognised as a benchmark for a ‘healthy’ diet and as such this study is directly addressing the question posed.
It also used a cohort an order of magnitude bigger than the previous study and therefore we can be more confident in the significance of the results obtained.
It could be argued that the text is somewhat dense and technical but other than that it is difficult to find an issue with this study other than those addressed when evaluating the experimental design.
This study design has external validity as the cohort closely matches the UK population of interest: subjects 60 years old and over, living in the community without significant co-morbitities.
Subject selection was stringent and utilised clinical review panels made up of old age psychiatrists, neurologists, geriatricians and neuropsychologists to verify diagnosis against a number of recognised measures. The cohort were already selected as part of the Australian Imaging, Biomarkers, and Lifestyle study of Ageing study and a wide range of baseline measurements were recorded including genetic data.
It is a cross-sectional study therefore only association can be established, not causation. We can only say that Alzheimer’s patients adhered less to MeDi at the time of survey but we have no information on whether they had always been non-adherents or whether it is a disease-induced change in dietary pattern. Alzheimer’s patients usually suffer weight-loss, eating difficulties, resistance to eating, and change in dietary patterns which worsen as the disease progresses.
Other confounders include the ‘healthy person effect’ whereby the MeDi adherent is more likely to have a broader ‘healthy lifestyle’ and is likely to minimise other risk factors for Alzheimer’s. The authors do adjust for these biases in their statistical analyses however.
Based on these...