Food purchase patterns indicative of household food access insecurity,children’s dietary diversity and intake,and nutritional status using a newly developed and validated tool in the Peruvian Amazon

Food security, defined as the capacity to acquire preferred food at all times, can manifest in many dimensions. Following a mixed methods approach used in India and Burkina Faso, we developed a 58-item experience-based measure in the Peruvian Amazon, based on investigator observations, relevant literature, and pre-testing with community field workers. The tool encompasses seven dimensions of food security and included measures of (1) food purchases, frequency of purchase, and location of acquisition, (2) food expenses, (3) coping mechanisms, (4) preparation of leftover food, (5) food safety (refrigerator access), (6) fishing intensity and (7) selling food. The survey was piloted among 35 randomly selected families from the Malnutrition Enteric Disease (MAL-ED) birth cohort in Santa Clara, Peru and the surrounding communities. Subsequently, based on a focus group discussion, a pile-sorting exercise, and pilot results, we reduced the survey to 36 items to be collected monthly among 203 MAL-ED households from November 2013 to January 2015. Validity and reliability were then assessed using principal component analysis and exploratory factor analysis, revealing four groups of purchase and coping strategy behaviors: (1) Sweets and sugary items, (2) Less preferred, (3) More preferred, and (4) Minimum meal. Internal consistency of the final 22-item scale had an acceptable cutoff of Cronbach’s α of 0.73. Criterion and construct validity of the factor groups revealed there were: (1) food purchase patterns that were distinctive to quality and quantity aspects of the Household Food Insecurity Access scale, (2) unique correlations of child’s intake of fats, animal source protein, fiber and other micronutrients, (3) household purchase patterns from the “more preferred” group (fish, red meat) associated with child’s weight-for-age. Food purchase and frequency, and context-specific behaviors at the household level can be used as surrogates for dietary intake patterns and nutritional status among children. Food purchase and frequency measurement is a quick, objective, non-intrusive survey method that could be used as an indicator for acute changes in household food security status with appropriate pilot testing and validation.