Population-averaged and subjectspecific approaches for clustered categorical data

Framing Social Comparison Feedback With Financial Incentives for Physical Activity Promotion: A Randomized Trial

Background: Social comparison feedback is often used in physical activity interventions but the optimal design of feedback is unknown. Methods: This 4-arm, randomized trial consisted of a 13-week intervention period and 13-week follow-up period. During the intervention, 4-person teams were entered into a weekly lottery valued at about $1.40/day and contingent on the team averaging ≥7000 steps per day. Social comparison feedback on performance was delivered weekly for 26 weeks, and varied by reference point (50th vs 75th percentile) and forgiveness in use of activity data (all 7 d or best 5 of 7 d). The primary outcome was the mean proportion of participant-days achieving the 7000-step goal. Results: During the intervention period, the unadjusted mean proportion of participant-days that the goal was achieved was 0.47 (95% confidence interval [CI]: 0.38 to 0.56) in the 50th percentile arm, 0.38 (95% CI: 0.30 to 0.37) in the 75th percentile arm, 0.40 (95% CI: 0.31 to 0.49) in the 50th percentile with forgiveness arm, and 0.47 (95% CI: 0.38 to 0.55) in the 75th percentile with forgiveness arm. In adjusted models during the intervention and follow-up periods, there were no significant differences between arms. Conclusions: Changing social comparison feedback did not impact physical activity.

A Randomized, Controlled Trial of Lottery-Based Financial Incentives to Increase Physical Activity Among Overweight and Obese Adults

PURPOSE

To evaluate the effect of lottery-based financial incentives in increasing physical activity.

DESIGN

Randomized, controlled trial.

SETTING

University of Pennsylvania Employees.

PARTICIPANTS

A total of 209 adults with body mass index ≥27.

INTERVENTIONS

All participants used smartphones to track activity, were given a goal of 7000 steps per day, and received daily feedback on performance for 26 weeks. Participants randomly assigned to 1 of the 3 intervention arms received a financial incentive for 13 weeks and then were followed for 13 weeks without incentives. Daily lottery incentives were designed as a "higher frequency, smaller reward" (1 in 4 chance of winning $5), "jackpot" (1 in 400 chance of winning $500), or "combined lottery" (18% chance of $5 and 1% chance of $50).

MEASURES

Mean proportion of participant days step goals were achieved.

ANALYSIS

Multivariate regression.

RESULTS

During the intervention, the unadjusted mean proportion of participant days that goal was achieved was 0.26 in the control arm, 0.32 in the higher frequency, smaller reward lottery arm, 0.29 in the jackpot arm, and 0.38 in the combined lottery arm. In adjusted models, only the combined lottery arm was significantly greater than control ( P = .01). The jackpot arm had a significant decline of 0.13 ( P < .001) compared to control. There were no significant differences during follow-up.

CONCLUSIONS

Combined lottery incentives were most effective in increasing physical activity.

A Randomized Trial of Social Comparison Feedback and Financial Incentives to Increase Physical Activity

PURPOSE

To compare the effectiveness of different combinations of social comparison feedback and financial incentives to increase physical activity.

DESIGN

Randomized trial (Clinicaltrials.gov number, NCT02030080).

SETTING

Philadelphia, Pennsylvania.

PARTICIPANTS

Two hundred eighty-six adults.

INTERVENTIONS

Twenty-six weeks of weekly feedback on team performance compared to the 50th percentile (n = 100) or the 75th percentile (n = 64) and 13 weeks of weekly lottery-based financial incentive plus feedback on team performance compared to the 50th percentile (n = 80) or the 75th percentile (n = 44) followed by 13 weeks of only performance feedback.

MEASURES

Mean proportion of participant-days achieving the 7000-step goal during the 13-week intervention.

ANALYSIS

Generalized linear mixed models adjusting for repeated measures and clustering by team.

RESULTS

Compared to the 75th percentile without incentives during the intervention period, the mean proportion achieving the 7000-step goal was significantly greater for the 50th percentile with incentives group (0.45 vs 0.27, difference: 0.18, 95% confidence interval [CI]: 0.04 to 0.32; P = .012) but not for the 75th percentile with incentives group (0.38 vs 0.27, difference: 0.11, 95% CI: -0.05 to 0.27; P = .19) or the 50th percentile without incentives group (0.30 vs 0.27, difference: 0.03, 95% CI: -0.10 to 0.16; P = .67).

CONCLUSION

Social comparison to the 50th percentile with financial incentives was most effective for increasing physical activity.

Individual Versus Team-Based Financial Incentives to Increase Physical Activity: A Randomized, Controlled Trial

BACKGROUND

More than half of adults in the United States do not attain the minimum recommended level of physical activity to achieve health benefits. The optimal design of financial incentives to promote physical activity is unknown.

OBJECTIVE

To compare the effectiveness of individual versus team-based financial incentives to increase physical activity.

DESIGN

Randomized, controlled trial comparing three interventions to control.

PARTICIPANTS

Three hundred and four adult employees from an organization in Philadelphia formed 76 four-member teams.

INTERVENTIONS

All participants received daily feedback on performance towards achieving a daily 7000 step goal during the intervention (weeks 1- 13) and follow-up (weeks 14- 26) periods. The control arm received no other intervention. In the three financial incentive arms, drawings were held in which one team was selected as the winner every other day during the 13-week intervention. A participant on a winning team was eligible as follows: $50 if he or she met the goal (individual incentive), $50 only if all four team members met the goal (team incentive), or $20 if he or she met the goal individually and $10 more for each of three teammates that also met the goal (combined incentive).

MAIN MEASURES

Mean proportion of participant-days achieving the 7000 step goal during the intervention.

KEY RESULTS

Compared to the control group during the intervention period, the mean proportion achieving the 7000 step goal was significantly greater for the combined incentive (0.35 vs. 0.18, difference: 0.17, 95 % confidence interval [CI]: 0.07-0.28, p <0.001) but not for the individual incentive (0.25 vs 0.18, difference: 0.08, 95 % CI: -0.02-0.18, p = 0.13) or the team incentive (0.17 vs 0.18, difference: -0.003, 95 % CI: -0.11-0.10, p = 0.96). The combined incentive arm participants also achieved the goal at significantly greater rates than the team incentive (0.35 vs. 0.17, difference: 0.18, 95 % CI: 0.08-0.28, p < 0.001), but not the individual incentive (0.35 vs. 0.25, difference: 0.10, 95 % CI: -0.001-0.19, p = 0.05). Only the combined incentive had greater mean daily steps than control (difference: 1446, 95 % CI: 448-2444, p ≤ 0.005). There were no significant differences between arms during the follow-up period (weeks 14- 26).

CONCLUSIONS

Financial incentives rewarded for a combination of individual and team performance were most effective for increasing physical activity.

TRIAL REGISTRATION

Clinicaltrials.gov identifier: NCT02001194.

On fitting generalized linear mixed-effects models for binary responses using different statistical packages

The generalized linear mixed-effects model (GLMM) is a popular paradigm to extend models for cross-sectional data to a longitudinal setting. When applied to modeling binary responses, different software packages and even different procedures within a package may give quite different results. In this report, we describe the statistical approaches that underlie these different procedures and discuss their strengths and weaknesses when applied to fit correlated binary responses. We then illustrate these considerations by applying these procedures implemented in some popular software packages to simulated and real study data. Our simulation results indicate a lack of reliability for most of the procedures considered, which carries significant implications for applying such popular software packages in practice.

Deer browse resistant exotic-invasive understory: an indicator of elevated human risk of exposure to Ixodes scapularis (Acari: Ixodidae) in southern coastal Maine woodlands

We evaluated the relationships between forest understory structure and the abundance of questing adult and nymphal blacklegged ticks, Ixodes scapularis Say (Acari: Ixodidae), in three Maine towns endemic for Lyme disease, 2001-2003. In fragmented New England woodlands, over-abundant white-tailed deer, Odocoileus virginianus Zimmerman, overbrowse palatable species, allowing browse-resistant exotic-invasive species to replace native forest understory structures. We predicted there would be more ticks in plots dominated by exotic-invasive shrubs (such as Japanese barberry, Berberis thunbergii DC) than in plots dominated by native shrubs, ferns, or open understory. We assessed canopy composition and closure, tree basal area, litter composition, percentage of coverage and stem density of understory species, litter depth, soil moisture, and abundance of small mammals and white-tailed deer pellet groups. We used generalized linear mixed model analysis of covariance to determine the effect of understory structure on tick counts, controlling for continuous habitat and host covariates and adjusting for random spatial effects. There were twice as many adults and nearly twice as many nymphs in plots dominated by exotic-invasives than in plots dominated by native shrubs. Both adult and nymphal counts were lowest in open understory with coniferous litter. Adults were positively associated with increasing litter depth, medium soil moisture, and increasing abundance of white-footed deer mice, Peromyscus leucopus Rafinesque, and deer pellet group counts. Nymphs were positively associated with increasing litter depth, moderately wet soil, and mice. We concluded that deer browse-resistant exotic-invasive understory vegetation presented an elevated risk of human exposure to the vector tick of Lyme disease.