Abstract
This study sought to determine the association between levels of fruit and vegetable consumption and time to death, and to explore potential moderators. We analyzed a nationally-representative sample of 1699 older adults aged 65+ who were followed up for a period of 6 years. Participants were classified into low (≤3 servings day), medium (4), or high (≥5) consumption using tertiles. Unadjusted and adjusted cox proportional hazard regression models (by age, gender, cohabiting, education, multimorbidity, smoking, physical activity, alcohol consumption, and obesity) were calculated. The majority of participants (65.7%) did not meet the recommendation of five servings per day. High fruit and vegetable intake increased by 27% the probability of surviving among older adults with two chronic conditions, compared to those who consumed ≤3 servings per day (HR = 0.38, 95%CI = 0.21-0.69). However, this beneficial effect was not found for people with none, one chronic condition or three or more, indicating that this protective effect might not be sufficient for more severe cases of multimorbidity. Given a common co-occurrence of two non-communicable diseases in the elderly and the low frequency of fruit and vegetable consumption in this population, interventions to promote consuming five or more servings per day could have a significant positive impact on reducing mortality.