Service de médecine aiguë gériatrique, CHU de Grenoble, France
Age-related immune impairment may be one of the factors influencing successful and pathological aging, being strongly tied to nutritional status. Several long term cohort studies suggest that a lower total lymphocyte count is associated with higher mortality. Nevertheless, prevalence, incidence and impact of lymphopenia on frailty and prediction of pathological events have not been described extensively. The principal aim of this study was to examine the relation of lymphopenia and intra-hospital mortality in the elderly. Materials and method: This cohort study has been carried out in a geriatric acute care unit of the Grenoble University hospital in France. Clinical and biological data have been retrospectively retrieved from the electronic medical record of each patient. All patients aged 75 or older admitted in the unit from May to October 2011 were eligible for inclusion. A lymphocyte count was obtained within 48h hours before or after admission. Results: 239 patients were included. Mean age (SD) was 87.04(5.50) years, 82(34.3%) patients were men, median ADL (activities of daily living) score prior to hospitalization was 4 (Q1:2; Q3:6). 31(13%) patients died during their admission. A lympocyte's threshold of 1,100cells/μL establishes a sensitivity of 79.3%, a specificity of 57.2%, a positive predictive value (PV+) of 21.7%, as well as a negative predictive value (PV-) of 94.9%. The OR was 3.44(IC95%, [(1.54-8.10]). The AL's threshold was found to be 3g/dL, establishing a sensitivity of 79.3%, a specificity of 62.0%, a PV+ of 22.8%, PV- of 95.5%. The OR was 4.90(IC 95%, [(2.17-11.87]). In multivariate analysis, LC and AL were significantly predicting in hospital death (OR=2.80 IC95% [(1.18–7.02] p=0.02, OR=3.34 (IC95%, [(1.41 –8.36]) p=0.007 respectively). Conclusion: This observational study carried out with malnourished, functionally impaired older inpatients with multiple comorbidities shows that lymphopenia independentely predicts intra-hospital mortality. In multivariate analysis lymphocyte count and albumin level independently predict intra-hospital mortality in with a similar predictive performance. Low albumin levels have previously been shown to be an independent risk factor for all-cause mortality in community-dwelling older persons as well as to predict intra-hospital mortality. However in our study we included patients with an acute condition and multiple comorbidities, potentially confounding the relation between lymphopenia and mortality. Lymphopenia may be an interesting marker of frailty and prognosis in very elderly people presenting an acute condition.