Beck Depression Inventory and survival in elderly hemodialysis patients

Sir, We read with great interest the article by Chilcot et al. [1]. They showed that depression, as assessed by the Beck Depression Inventory-II (BDI), is an independent predictor of mortality in patients with end-stage renal disease (ESRD) on dialysis. We wish to state that this conclusion is not necessarily applicable to the fastest growing cohort of ESRD patients—the elderly (≥65 years). We recently published our findings that show that both the BDI and the 15-item Geriatric Depression Scale are useful tools to screen for depression in the elderly hemodialysis (HD) population (compared to the gold standard psychiatric interview) [2]. Eighty-nine patients were included in this study, aged 73.5 ± 6.2 years, with 39 males, 50 Caucasians and 39 African Americans. In this subsequent study, we followed this same cohort for 5.5 years or until death, transferring out of our HD unit or the end of the study period. Data was available on 76 patients. Forty-five patients died (59.2%). Depression (D), as assessed by BDI score ≥10, was noted in 31 subjects, while 45 subjects had BDI scores <10 no depression (ND). Seventeen patients from the (D) group died within the study period (54.8%) as compared to 28 from the (ND) group (62.2%). Using Kaplan–Meier survival method, the mean survival was 32.9 ± 3.3 months and 34.5 ± 3.4 months for (D) and (ND) groups, respectively. The median survival time was 41.9 months (95% confidence interval: 22.1–66 months) and 30.2 months (95% confidence interval: 18.9–59.2 months) for the two groups, respectively. None of these parameters showed any difference which was statistically significant. Several studies have shown that depression, as assessed by BDI, was associated with mortality in patients with ESRD [1, 3] and non-ESRD patients [4]. Other studies such as Pinquart and Duberstein's [5] contradicted these findings and were similar to our findings of poor association between depressive symptoms, as assessed by BDI, and mortality. A potential explanation could be related to the ages of the patients tested. Our study, as well as the meta-analysis by Pinquart and Duberstein, were done (mainly) in the elderly (73.5 ± 6.2 and 64.5 ± 11.6 years, respectively). The preponderance of somatic symptoms in elderly patients with ESRD in our population and with cancer in the other study, as picked up by the multiple somatic items in the BDI, may not be necessarily related to depression [5]. We recommend caution with interpreting data related to the association of depression, as assessed by the BDI, and mortality, especially in the elderly. Funding. National Kidney Foundation serving Virginia (formerly National Kidney Foundation of the Virginias) grant to RAB, SAB and EMR. Conflict of interest statement. None declared. Editorial Note: Dr Chilcot et al. had no further comments on this letter.