Predictors of food variety and dietary diversity among older persons in Botswana

Objective We investigated whether food variety and diversity are associated withphysical and cognitive functioning in older adults in Botswana anddesigned a simple set of screening questions that predict food varietyin this population. Methods Datawere collected (1998) as a national household survey of 1085 subjects60 y and older. A food variety score, based on a foodfrequency checklist, was calculated by summing the frequency of weeklyintakes of 16 food items (0 to 66). A dietary diversity score wascalculated as the number of food groups consumed weekly (0 to 5). Arepresentative subsample (n = 393) wasrandomly selected for the clinical component of the survey, andmeasurements on dependency and cognitive function wereconducted. Results Low food variety was found: 35.2%, 59.3%, and 22.4% ofsubjects consumed no dairy products, fruits, and vegetables,respectively. A higher food variety score was associated with urbanresidence, ownership of cattle, higher education, and more frequentmeals, and these indicators were used to construct a nutritional riskindicator. Higher food variety score was associated with betterself-reported health and better cognitive function. Similarly, a higherscore on the nutritional risk indicator screening tool was associatedwith desirable healthoutcomes. Conclusion A limited number of foods is consumed, leading to an overall pattern ofpoor food variety. Higher food variety was associated with improvedphysical and cognitive functions. A screening tool that predicts foodvariety in this population has been developed and is recommended to beincorporated at a primary care level to identify older adults most atrisk of a poor qualitydiet.