Household diet quality in relation to mortality in Portuguese regions: an ecological study

Abstract

The aim of this study was to investigate the association between regional household diet quality and corresponding mortality data. Two household diet quality scores derived from household budget survey food and nutrient data were used: the Mediterranean Adequacy Index (MAI), assessing adherence to a Mediterranean food pattern, and a revised Healthy Diet Indicator (HDIr), based on compliance with WHO population dietary goals. Seven Portuguese regions were considered, in accordance with the officially adopted national territorial division NUT II. The corresponding regional age-standardised death rates were calculated for selected diet-related mortality causes. In order to allow for a minimum of 10 years latency period between exposure and outcome, dietary scores from 1989/1990 and its components were correlated to the 3-year average 1999/ 2001 death rates. Tobacco expenses were included for adjustment. Significant negative correlations with the proportion of families with highest MAI (≥4) or HDIr (≥7) scores were observed for mortality from diabetes mellitus (−0.87, P<0.05; −0.89, P<0.5), cardiovascular diseases (−0.89, P<0.05; −0.96, P<0.01) and ischaemic heart disease (−0.89, P<0.05; −0.90, P<0.05). Malignant neoplasms were positively correlated with protein, cholesterol and soft drinks, but inversely with polyunsaturated fatty acids (PUFA). Diabetes showed significant positive correlations with simple sugars and animal fat intake, but inversely with fruits and vegetables. Cardiovascular diseases presented positive correlations with simple sugars, milk and dairy products and animal fat, but inverse with fruits and vegetables and vegetable oils. Despite the limitations of ecological approaches in general and of the statistical power in this particular study, due to the use of broad geographical areas, significant correlations in the expected direction of reduced mortality risk with improved household diet quality was reported. The use of diet quality scores in data from household budget surveys could depict variations in Portuguese regional food and nutrient patterns, which were revealed to be related and partially explain differences in regional cause-specific mortality rates from diet-related diseases.