Plasma ceramides independently predict all-cause mortality in men aged 85

BACKGROUND: assessing cardiovascular and mortality risk with conventional biomarkers is challenging in oldest-old due to multimorbidity and polypharmacy. Ceramides are bioactive lipids shown to predict mortality in late middle-aged cohorts. OBJECTIVE: to assess whether plasma ceramides have independent prognostic value for mortality among oldest-old (85+). DESIGN: longitudinal cohort study (Helsinki Businessmen Study, HBS) with a 3.5-year follow-up. SETTING AND SUBJECTS: survivors of HBS (125 men born in 1919-1934) visited the clinic for laboratory and clinical examination. METHODS: functional status including physical (short physical performance battery) and Montreal Cognitive Assessment (MoCA) cognitive performance was assessed and laboratory examinations included a large set of biomarkers. Plasma ceramide concentration (Cer(d18:1/16:0)) was measured using a targeted liquid chromatography-tandem mass spectrometry assay. Mortality was retrieved from national registers. RESULTS: median age was 88 years, two-thirds had multimorbidity and 59% were on statin treatment. During the follow-up, 22 (18%) men died. In a model adjusted for variables associated with mortality in the whole cohort at P < 0.20 (log glucose, SPPB, MoCA and statin use), Cer(d18:1/16:0) as a continuous trait was associated with increased mortality: hazard ratio (HR) per 1 SD 1.64 (95% confidence interval [CI] 1.23-2.18). Compared with the bottom tertile of Cer(d18:1/16:0), HR of mortality was 5.44-fold (95% CI 1.17-25.3) in the top tertile. CONCLUSIONS: these data raise the hypothesis that plasma ceramide concentrations and especially Cer(d18:1/1:60) may offer a clinically useful biomarker to evaluate prognosis in very old age. Such biomarkers are needed for geriatrics, where multimorbidity and pharmacotherapies, such as statins are prevalent hampering assessment of prognosis using conventional methods.