Skip to main content
. 2016 May 10;4(2):e42. doi: 10.2196/mhealth.5280

Table 3.

Mean differences in changes between the ‘Get Healthy, Stay Healthy’ (GHSH) intervention (I; n=104) and control groups (C; n=114).


Mean difference
(GHSH – control)a
Between-group interpretation

Mean (95% CI) P
Anthropometry

Weight (kg) 1.35 (−2.25, −0.46) b .003 significantly better

Waist circumferencec(cm) −1.18 (−2.65, 0.29) .116 inconclusive
Physical activity (PA)

Vigorous PA (sessions/week)d 0.15 (−0.34, 0.63) .547 inconclusive

Moderate PA (sessions/week)d 0.55 (0.14, 0.96) .008 significantly better

Walking PA (sessions/week) −0.69 (−1.46, 0.08) .077 inconclusive

Accelerometer PA (mins/week)e 24.16 (5.07, 43.25) .007 significantly better
Dietary behaviors

Vegetables (servings/day) 0.15 (−0.21, 0.50) .408 inconclusive

Fruit (servings/day)d 0.16 (−0.05, 0.37) .133 similar

Sweetened drinks (cups/day) −0.05 (−0.19, 0.10) .537 similar

Takeaways (meals/week) 0.01 (−0.15, 0.18) .864 similar

FFBQ Total Index Score (1–5) 0.05 (−0.03,0.13) .195 similar

FFBQ Fat Index Score (1–5) 0.02 (−0.07, 0.12) .615 similar

FFBQ Fiber Index Score (1–5) 0.08 (−0.03, 0.19) .147 similar

aMean difference (β) with 95% CI, and Pvalue from linear regression models, adjusted for baseline values of the outcome and confounders (listed in Multimedia Appendix 1).

bItalic values indicate statistical significance at P˂.05.

cn=102 GHSH group; n=112 control group (item missing data for waist circumference).

dn=103 GHSH group; n=114 control group (item missing data for vigorous PA, moderate PA and Fruit).

eEstimated using marginal means from GEE models of log-transformed daily MVPA (repeated term for “day”), adjusting for confounders, and correcting for regression to the mean using the method [37] of including the term for assessment (pre/post) and the assessment x group interaction, but not the conditional term for group. Estimates were back-transformed to the original scale, then multiplied by 7 to yield minutes per week. n=112 GHSH group; n=114 control group (all participants with data at either time point examined).