Skip to main content

Table 1 Mixed effects modelsa to evaluate intervention across 12 and 24 months post-baseline

From: Two-year outcomes of Faith in Action/Fe en Acción: a randomized controlled trial of physical activity promotion in Latinas

Outcomes Time X Condition Interaction P-value Condition
Intervention Control Difference (Inter – Control)
Adj Mean SE Adj Mean SE Diff in adj means P-value Effect Size
Primary
 Accelerometer MVPAb .81 4.71 0.05 4.59 0.05 0.12 .081 .18
 Self report leisure time MVPAb .37 3.29 0.20 2.44 0.22 0.85 .005 .30
Secondary
 Accelerometer light activity .69 2342.2 29.3 2324.1 25.0 18.1 .64 .05
 Accelerometer percent sedentary time .71 75.1 0.30 75.4 0.26 − 0.3 .61 .09
 BMI .10 30.3 0.13 30.5 0.12 − 0.2 .33 .12
 Waist circumference .45 95.7 0.4 96.3 0.4 − 0.6 .27 .10
   Odds Ratio   P-value
Meets MVPA recommendation from leisure & transport domainsc   Estimate 95% CI   
Intervention vs Control .75 1.80 1.21, 2.67   .004
  1. aMixed effects or generalized linear mixed models were used to adjust for the clustering effects of churches and to account for repeated measures over M2 and M3. If the time by condition interaction term was not significant, the term was dropped and the condition main effect was tested. All analyses were adjusted for the baseline measure of the outcome, age, marital status, employment and education
  2. bNegative binomial error distribution. Results are shown in logged units
  3. cBinomial error distribution (Logistic model)