RESUMEN
Low taste sensitivity may be one factor related to undernutrition, which is a major problem in developing countries. The purpose of this cross-sectional study was to examine the association between underweight, one indicator of undernutrition, and taste sensitivity in middle- to old-aged Sri Lankan nursing home residents. Participants were 946 residents with BMI of <25·0 from 25 nursing homes. Data were obtained on height, weight, taste sensitivity, subjective taste ability, sex, age, ethnicity, number of years in nursing homes, activities of daily living (ADL), frequency of exercise, bowel movements, smoking status, drinking status, current number of chronic diseases, number and kinds of medications used, self-reporting questionnaire 20 (SRQ20), subjective smell ability, number of teeth present, Eichner index and flow rate of saliva. Low sensitivity to bitter taste, being male, old age, low ADL, smoking experience, drinking experience, fewer medications used and no use of medication for hypertension and diabetes were each associated with underweight (P < 0·05). In a multilevel Poisson regression model adjusted for sex, age, ADL, smoking status, drinking status, number of medications used, use of medication for hypertension and diabetes and flow rate of saliva, subjects with low sensitivity (>0·003% quinine hydrochloride dihydrate) to bitter taste had a significant 1·70 times higher prevalence ratio (95% confident interval 1·04-2·80) for underweight compared with those with high sensitivity (0·0001% quinine hydrochloride dihydrate). These results suggest that low taste sensitivity to bitter taste may be one factor related to underweight.