Skip to main content

Table 7 Comparative behavior change patterns (at risk vs no/low risk), n = 167

From: Adolescents as agents of healthful change through scientific literacy development: A school-university partnership program in New Zealand

   Pre-intervention to 12 months post-intervention behavior change
   T0 behavior category = risk T0 behavior category = no/low risk  
Food item Self-reported consumption pattern defined as indicating risk n % Negative change towards greater risk % No change % Positive change towards no/low-risk category % Positive change into the no/low-risk category n % Retains position in no/low-risk category % Negative change into at risk category Odds ratio T0-risk/T0-no/low-risk T0–T4 change towards opposite category
Potato chips (crisps) > once/week 69 5.8 52.2 7.2 34.8 94 80.9 19.1 3.1 (95% CI 1.2 to 6.2), χ2(1) = 10.15, p < .001*
Fried food (e.g., hot chips, fried chicken, burgers) ≥ once per week 93 9.7 53.8 17.2 19.4 66 66.7 33.3 1.2 (95% CI 0.6 to 2.2), χ2(1) = 0.18, p = .671
Soft drinks (fizzy, cordials, sports drinks) ≥ 2–4 times per week 41 7.3 48.8 4.9 39.0 123 87.8 12.2 5.6 (95% CI 2.5 to 12.8), χ2(1) = 19.23, p < .001*
Sweet snacks (e.g., biscuit, muesli bar, sweet (candy)) > 2–4 times per week 28 N/A 50.0 N/A 50.0 131 82.4 17.6 4.696 (95% CI 1.974 to 11.173), χ2(1) = 13.60, p < .001*
Green vegetables (e.g., spinach, beans, lettuce) < Daily 58 10.3 50.0 5.2 34.5 101 84.2 15.8 3.5 (95% CI 1.7 to 7.4), χ2(1) = 11.29, p = .001*
Starchy vegetables (e.g., sweet potato, potato, pumpkin) ≤ once per week 26 3.8 23.1 7.7 65.4 133 91.0 9.0 27.3 (95% CI 9.6 to 78.2), χ2(1) = 56.85, p < .001*
Fruit (e.g., apples, pears, bananas) < Daily 57 14.0 40.4 17.5 28.1 109 84.9 15.1 5.7 (95% CI 2.7 to 12.0), χ2(1) = 56.85, p < .001*
Raw fruits and vegetables < Daily 88 19.3 28.4 29.5 22.7 74 73.0 27.0 3.0 (95% CI 1.5 to 5.7), χ2(1) = 10.61, p = .001*
  1. T0 pre-intervention, T2 6–12 weeks post-intervention, T4 12 months post-intervention, n number, p*significant (α = 0.05). The effect of T0 response group (“at risk” vs “low/no risk”) on change response was measured using ordinal logistic regression with proportional odds