Development of dietary tools to assess dietary patterns of New Zealand adolescents

Abstract

Despite the recognised importance of overall dietary patterns for adolescents’ health, dietary patterns have not been assessed by means of a diet index in New Zealand (NZ) adolescents. The lack of a validated adolescent-specific food frequency questionnaire has precluded the collection of dietary information in resource-limited research settings, while the absence of an NZ-specific diet index makes description of index-based dietary patterns impossible. This research aimed to address these gaps by developing valid and practical tools for assessment of dietary patterns among NZ adolescents aged 14 to 18 years. Four studies were conducted using data from the Otago School Students Lifestyle Survey Two (OSSLS2) and 2008/09 NZ Adult Nutrition Survey (2008/09 NZANS).

Study I involved adaptation, pretesting and assessing the relative validity of a non-quantitative, 72-item NZ Adolescent FFQ (NZAFFQ) in a pilot sample (n 41) of the OSSLS2. The NZAFFQ was found to be comparable to a four-day estimated food record in ranking participants according to 34 food groups (rs=0.40), and highly repeatable when test-retested within two weeks (rs=0.71). In Study II, this questionnaire provided dietary information required for the development of the food-based NZ Diet Quality Index for Adolescents (NZDQI-A). The NZDQI-A was calculated based on ‘variety’ and ‘adequacy’ for five equally weighted food group components. Achieving a NZDQI-A score in the highest one third, was significantly associated with a higher intake of dietary iron and lower intake of total and saturated fat (P-trend<0.05).

In Study III, the utility of NZDQI-A was demonstrated through its application in the OSSLS2 main study to examine the potential associations between diet quality and body composition among Year 11-13 students (n 681). Higher NZDQI-A scores were significantly associated with lower body fat percent (β=-0.19; 95% CI=-0.35 to -0.04, P<0.05), fat-to-lean mass ratio (β=-0.26; 95% CI=-0.46 to -0.05, P<0.05) and Fat Mass Index (β=-0.23; 95% CI=-0.45 to -0.004, P<0.05) after multivariate adjustment.

A 17-item Healthy Dietary Habits Score for Adolescents (HDHS-A) was developed in Study IV, using qualitative dietary habits data from the 2008/09 NZANS (n 694). A higher HDHS-A score was associated with being female, of not Māori or Pacific ethnicity and in the highest socioeconomic level. Compared to nutrient intakes derived from single 24-hour diet recalls, achieving a higher third of HDHS-A score was associated with a higher intake of protein, dietary fibre, polyunsaturated fat, lactose, and most micronutrients (P-trend<0.05). Significant associations between HDHS-A score and urinary sodium, whole blood, serum and red blood cell folate levels in the expected directions (P-trend<0.05) also established construct validity of the diet index.

Through rigorous development and validation, this research demonstrated the validity and application of the NZAFFQ, NZDQI-A and HDHS-A as dietary tools to assess dietary patterns of NZ adolescents aged 14-18 years. Further validation of these dietary tools (in particular NZAFFQ and NZDQI-A) in larger, more socioeconomically and ethnically diverse samples is recommended to advance dietary pattern methodology and promote better understanding of diet-health relationships among NZ adolescents.