Abstract
BACKGROUND: Individuals who suffer from mental illness are more prone to obesity and related co-morbidities, including the metabolic syndrome. Autopsies provide an outstanding platform for the macroscopic, microscopic and molecular-biological investigation of diseases. Autopsy-based findings may assist in the investigation of the metabolic syndrome. To utilise the vast information that an autopsy encompasses to elucidate the pathophysiology behind the syndrome further, we aimed to both develop and evaluate a method for the post mortem definition of the metabolic syndrome. METHODS: Based on the nationwide Danish SURVIVE study of deceased mentally ill, we established a set of post mortem criteria for each of the harmonized criteria of the metabolic syndrome. We based the post mortem (PM) evaluation on information from the police reports and the data collected at autopsy, such as anthropometric measurements and biochemical and toxicological analyses (PM information). We compared our PM evaluation with the data from the Danish health registries [ante mortem (AM) information, considered the gold standard] from each individual. RESULTS: The study included 443 deceased individuals (272 male and 171 female) with a mean age of 50.4 (± 15.5) years and a median (interquartile range) post mortem interval of 114 (84-156) hours. We found no significant difference when defining the metabolic syndrome from the PM information in comparison to the AM information (P = 0.175). The PM evaluation yielded a high specificity (0.93) and a moderate sensitivity (0.63) with a moderate level of agreement compared to the AM evaluation (Cohen's κ = 0.51). Neither age nor post mortem interval affected the final results. CONCLUSIONS: Our model of a PM definition of the metabolic syndrome proved reliable when compared to the AM information. We believe that an appropriate estimate of the prevalence of the metabolic syndrome can be established post mortem. However, while neither the PM nor the AM information is exhaustive in terms of defining an individual's health status, a superlative estimate may be obtained by combining the PM and the AM information. With this model, we open up the possibility of utilising autopsy data for future studies of the metabolic syndrome.