Is thiamine deficiency in elderly people related to age or co-morbidity?

Objectives: to compare erythrocyte thiamine pyrophosphate concentrations in elderly people with those in healthy younger people; to determine if any d ifferences can be attributed to age or to co-morbidities. Design: cross-sectional and 3-year longitudinal surveys. Setting: primary care. Patients: 100 volunteer blood donors and 222 elderly people from a general practice register. Measurements: thiamine pyrophosphate concentrations using high performance liquid chromatography; physical examination, medical and medication history ; grip strength, body mass index and plasma albumin. Results: the mean [95% confidence interval (CI)] thiamine pyrophosphate con centration was 152 nmol/l (147-158) in the elderly group and 224 (213-235) nmol/l in the younger group (P < 0.001). Ninety-six (43.4%) of the elderly subjects had thiamine pyrophosphate concentrations below the fifth percenti le of the younger subjects (140 nmol/l). Over 3 years thiamine pyrophosphat e concentrations fell in the elderly cohort by 20% (35% CI: 14.5-24.5%; P < 0.01). Thiamine pyrophosphate concentrations in 39 healthy older people we re no different from those in elderly people with co-morbidity but were sig nificantly lower than those in the younger people. Elderly people with abse nt vibration sense in their feet had a lower thiamine pyrophosphate concent ration than the rest of the group [129 (117-142) nmol/l compared with 156 ( 150-162) nmol/l; P < 0.01)]. Thiamine pyrophosphate concentrations were not related to prevalent diseases, common medications, body mass index, grip s trength or plasma albumin. Conclusion: lower thiamine pyrophosphate concentrations in elderly people a ppear to be related more to age itself than to co-existent illnesses.