To estimate the within-households association between change in income over time and food purchases in a national panel of households. The need to shift towards healthy and sustainable diets is widely recognized, thus the importance of identifying the factors that influence food purchase decisions.

Longitudinal observational study; for each of the 33 food items queried, we ran a conditional logistic fixed-effect regression model to evaluate the association between change in income per-capita and food purchases (yes/no) during the past week, adjusted by covariates.

Mexican Family Life Survey

6,008 households that participated in the survey for at least two of the three available waves of study (2002, 2005, and 2009).

Within-households, the odds ratio (95% CI) of purchasing the food in the past week for an increase in 1 SD of income was 1.09 (1.02, 1.16) for rarer fruits (other than bananas, apples, and oranges); 1.11 (1.04, 1.18) for beef; 1.06 (1.00, 1.13) for canned tuna/sardines; 1.09 (1.02, 1.18) for fish/shellfish; 1.08 (1.02, 1.16) for discretionary packaged products, and 1.15 (1.08, 1.23) for soft drinks. There were some differences by urban/rural area or SES; mainly, those with lower SES had increased odds of purchasing the food item in more cases (10 out of 33 food items).

households’ income growth can have mixed effects on the healthiness and sustainability of food purchases. Public policies to improve the food environment and nutrition education are necessary to enhance the positive and counteract the negative effect of income.