Skip to main content
. 2021 Mar 12;23(3):e24135. doi: 10.2196/24135

Table 2.

Prevalence of individual risk behaviors over time in intervention and control adolescents.

Behavior Intervention Control Logistic regression, P valuea

Baseline (n=145), n (%) 3 months (n=138), n (%) 6 months (n=139), n (%) Baseline (n=155), n (%) 3 months (n=145), n (%) 6 months (n=145), n (%)
Low fruit or vegetable intake 115 (79.3) 106 (76.8) 98 (70.5) 132 (85.1) 118 (81.4) 111 (76.6) .93
Low sleep time 46 (31.7) 52 (37.7) 65 (46.8) 54 (34.8) 55 (37.9) 72 (49.6) .89
Low physical activity 39 (26.9) 44 (31.9) 36 (25.9) 50 (32.3) 52 (35.9) 51 (35.2) .77
Inconsistent helmet use 37 (25.5) 24 (17.4) 22 (15.8) 39 (25.2) 25 (17.2) 20 (13.8) .54
High sugary beverage intake 28 (19.3) 39 (28.2) 36 (25.9) 36 (23.2) 37 (25.5) 35 (24.1) .47
Depression 13 (9.0) 15 (10.9) 14 (10.1) 23 (14.8) 15 (10.3) 18 (12.4) .24
Inconsistent seatbelt use 16 (11.0) 7 (5.1) 11 (7.9) 10 (6.5) 7 (4.8) 10 (6.9) .11
Texting while driving 9 (5.8) 10 (7.2) 8 (5.8) 13 (8.4) 8 (5.5) 10 (6.9) —b
Marijuana use 10 (6.9) 4 (2.9) 3 (2.2) 7 (4.5) 4 (2.8) 5 (3.4)
Alcohol use 8 (5.5) 3 (2.2) 4 (2.9) 6 (3.9) 4 (2.8) 4 (2.8)
Tobacco use 4 (2.8) 0 (0.0) 1 (0.7) 4 (2.6) 2 (1.4) 3 (2.1)
Sexual risk 1 (0.7) 3 (2.2) 3 (2.2) 4 (2.6) 2 (1.4) 3 (2.1)
Driving under the influence 0 (0.0) 0 (0.0) 0 (0.0) 2 (1.3) 1 (0.7) 0 (0.0)

aP values were based on the likelihood ratio test comparing mixed effects logistic regression with and without period-by-group interaction. Both models controlled for random effects corresponding to within-individual clustering.

bOwing to concerns about estimate instability, we did not conduct analyses for individual behaviors in which fewer than 10 adolescents per study arm endorsed the behavior.