Determinants of dietary patterns in school going adolescents in Urban Zambia

BACKGROUND: Understanding dietary patterns in a population is critical for decision making. This study aimed to identify the prevailing dietary patterns and their associated individual and school environment factors among school going adolescents in Lusaka, Zambia. METHOD: A cross-sectional study involving 404 Grade 10 pupils from 10 secondary schools in Lusaka district was conducted. A 108-item unquantified Food Frequency Questionnaire (FFQ) was used to assess the learner's food intake practices. Principal component analysis (PCA) was used to derive dietary patterns from the 108 food items. In addition, a mapping of food vendors and types of food sold was conducted in the same 10 schools using a semi-structured observation checklist. Bivariate and multivariate multilevel regression was used to analyse the individual and school level determinants of the adolescent dietary patterns. RESULTS: The average age of learners was 16.1 years (SD 1.4 years); 234 (58%) were female while 170 (42%) male. “Snacking,” “vegetarian,” “health conscious,” and “traditional” dietary patterns accounting for 54.5% of variability in learner's diets were identified using PCA. At individual level, having weekly pocket money was significantly associated with snacking (p ≤ 0.0001). Self-identified poverty was associated with snacking (p ≤ 0.0001), vegetarian (p = 0.009) and traditional (p = 0.009) dietary patterns. School level factors like a school tuckshop (similar to canteen) that sells fast foods or a kantemba (semi-permanent makeshift store) within the school vicinity (p = 0.023) were significantly associated with a snacking dietary pattern. School tuckshop selling nshima (a thick maize based porridge) was significantly associated with vegetarian (p = 0.007), health conscious (p = 0.02) and traditional dietary patterns (p=0.01) while a tuckshop with fruit significantly predicted traditional (p ≤ 0.0001), vegetarian (p = 0.041), and snacking (p = 0.002), dietary patterns. Having a supermarket or fast food restaurants in the school vicinity did not significantly influence any dietary pattern. CONCLUSION: Both individual behavioral and school environment level factors were found to be significant determinants of the four dietary patterns identified in this study.