The psychometric correlates of mortality can be underestimated in a world that values physiologic explanations. Nonetheless, both cognitive impairment and depression are significant predictors of mortality in elderly populations. It is important to note that these risks are often statistically independent of, and additive to, those associated with comorbid physical conditions. This suggests that psychiatric illness itself may increase mortality via direct effects on the CNS.
In this issue of CHEST (see page 1687), Antonelli-Incalzi et al1 report a significant association between performance on a single cognitive measure, the Hécaen copy to landmarks test, and mortality in patients with advanced COPD. This effect was independent of physiologic risks, such as hypoxemia, comparable in size (actually stronger than FEV1!) and additive to the risk of hypoxemia. More than 30% of nonhypoxemic patients who failed this task died over 5 years. Another striking feature is that this effect was limited to the Hécaen copy to landmarks test, and was not seen with the other cognitive measures in their battery.