Nutrition transition among adolescents of a south-Mediterranean country: dietary patterns, association with socio-economic factors, overweight and blood pressure. A cross-sectional study in Tunisia

Abstract

Background: The increase in the burden of chronic diseases linked to the nutrition transition and associated dietary and lifestyle changes is of growing concern in south and east Mediterranean countries and adolescents are at the forefront of these changes. This study assessed dietary intake and association with socio-economic factors and health outcomes among adolescents in Tunisia.

Methods: Cross-sectional survey (year 2005); 1019 subjects 15-19 y. from a clustered random sample. Dietary intake was assessed by a validated semi-quantitative frequency questionnaire (134 items) as was physical activity; the Diet Quality Index International measured diet quality; dietary patterns were derived by multiple correspondence analysis from intakes of 43 food groups. Body Mass Index (BMI) ≥ 85th and 95th percentile defined overweight and obesity. Waist Circumference (WC) assessed abdominal fat. High blood pressure was systolic (SBP) or diastolic blood pressure (DBP) ≥ 90th of the international reference for 15-17 y., and SBP/DBP ≥ 120/80 mm Hg for 18-19 y.

Results: Energy intake levels were quite high, especially for females. The macro-nutrient structure was close to recommendations but only 38% had a satisfactory diet quality. A main traditional to modern dietary gradient, linked to urbanisation and increased economic level, featured an increasing consumption of white bread, dairy products, sugars, added fats and fruits and decreasing consumption of oils, grains, legumes and vegetables; regarding nutrients this modern diet score featured a decreasing relationship with total fat and an increase of calcium intake, but with an increase of energy, sugars and saturated fat, while vitamin C, potassium and fibre decreased. Adjusted for age, energy and physical activity, this modern pattern was associated with increased overweight in males (2nd vs. 1st tertile: Prevalence Odds-Ratio (POR) = 4.0[1.7-9.3], 3rd vs. 1st: POR = 3.3[1.3-8.7]) and a higher WC. Adjusting also for BMI and WC, among females, it was associated with decreased prevalence of high blood pressure (2nd vs. 1st tertile: POR = 0.5[0.3-0.8], 3rd vs. 1st tertile: POR = 0.4[0.2-0.8]).

Conclusion: The dietary intake contrasts among Tunisian adolescents, linked to socio-economic differentials are characteristic of a nutrition transition situation. The observed gradient of modernisation of dietary intake features associations with several nutrients involving a higher risk of chronic diseases but might have not only negative characteristics regarding health outcomes.

Figure 1

Conceptual framework for analysis of associations between dietary patterns, socio-economic factors and health outcomes. Conceptual framework for analysis of associations between dietary patterns, socio-economic factors and health outcomes. Arrows A, B and C (solid bold lines) represent the main associations studied, while the other associations represented (solid thin lines) were taken into account only to adjust for possible confounding; socio-cultural and psychological factors where not directly measured nor analysed in the study but represented so as to keep in mind that they may interact (dashed arrows) with the other factors at the different levels of the conceptual framework.

Figure 2

Histograms of distribution of modern and meat-fish dietary pattern scores (n = 1019). Distribution of modern and meat-fish dietary patterns scores derived by multiple correspondence analysis of dietary intakes of the 43 food groups (coded in quintiles). The modern and meat-fish pattern are respectively the first and second principal axes. For each subject the score for a given pattern is a weighted linear combination of the binary variables coding her/his values regarding the quintiles of the 43 food groups. Each axis was rescaled from 0 to 100 for readability purposes.

Figure 3

Intakes of the 43 food groups according to the modern and meat-fish patterns. By gender (male n = 432, female n = 587), according to the modern and meat-fish diet score patterns, intakes of the 43 food groups (in g/1000 kcal or g/4180 kj) from which the patterns were derived by multivariate analysis. Overall trends are estimated using local polynomial smoothing.

Figure 4

Selected macro and micronutrient intakes and DQI-I scores according to the modern and meat-fish patterns. By gender (male n = 432, female n = 587), intakes of selected macro and micronutrient and DQI-I (Diet Quality Index International) global and components scores according to the modern and meat-fish dietary patterns derived by multivariate analysis of the 43 food groups variables. Overall trends are estimated using local polynomial smoothing.