Objective: To examine prospective changes in food habits and nutrient intakes in a representative New Zealand sample of community dwelling adults aged 70 y and over. Design: Longitudinal study with food intake data collected in 1988a89 and again in 1995a96. In an attempt to distinguish age, time and cohort effects, data were analysed longitudinally, cross-sectionally and timesequentially. Subjects: The sample for study consisted of all non-institutionalised people aged 70 years and over registered with the Mosgiel Health Centre in 1988. In 1988a89, 678 adults completed a dietary survey (85% of those eligible) and 248 adults participated again in 1995a96 (66% of those eligible). Results: Energy intakes declined longitudinally in men only; however, this decline appeared not to be an aging effect as energy intake was not found to decrease with age cross-sectionally. Percentage of energy from protein increased by 0.7% in women (95% con®dence interval 0.2 ± 1.2) in both the longitudinal and time-sequential analysis, suggesting a time effect. The percentage of energy from saturated fat decreased 0.7% (95% con®dence interval 7 1.4 to 7 0.1) and percentage of energy from polyunsaturated fat increased 0.4% (95% con®dence interval 0.0 ± 0.7) in women, and appears to be a time effect. However, the increase in saturated fat and decrease in polyunsaturated fat with advancing age seen cross-sectionally suggests a cohort effect also occurring. In 1995a96, more people were using margarine as a spread and vegetable oils to cook meat. Milk and milk product consumption increased (not signi®cantly), and meat intake decreased signi®cantly by 5 and 4 servings per month in men and women, respectively. There was an increase in the proportion of people who ate breakfast cereal at least once a week, and more women ate brown or wholemeal bread in 1995a96. Conclusion: Over the 6 y follow-up period studied, there was no indication of an age effect on nutrient intakes in adults aged 70 y and older; however, the changes occurring over time demonstrate that older adults, particularly women, are making changes towards healthier food choices. Sponsorship: Health Research Council of New Zealand and the University of Otago Medical School.

No abstract