Background: Physical inactivity and increased sedentary activity are major lifestyle factors that contribute to increasing burden of disease globally. There is evidence that the rise in prevalence of non-communicable diseases (NCDs) was largely responsible for growth in private insurance spending between 1987 and 2002 in the USA. There is renewed interest by health plans to implement incentivized health promotion programs, in an effort to change health behavior and thereby reduce health care costs related to NCDs. However, the literature on the role of incentives and rewards in influencing physical activity behavior, specifically, is relatively sparse, and the field of behavioral economics applied to health behavior is relatively new.

Most studies that have been conducted typically report findings with limited, well-defined incentives that are offered for a short period. These studies generally have shown a moderation in health care costs compared to non-participants. In a cross-sectional study of approximately 940,000 members of a national private health insurance, Discovery Health, the investigators have previously demonstrated a significant and inverse relationship between levels of participation in fitness-related activities and medical claims associated with hospital admissions. More recently, in a retrospective, longitudinal analysis over 5 yrs, amongst more than 300,000 members, the investigators found that participation in fitness-related activities increased over a 5-yr period and was associated with a significantly lower probability of hospital admissions and inpatient claims costs.

Aim: The current study primarily aims to prospectively compare the effectiveness of diverse incentives on physical activity behavior, monthly over a 12 month period following registration, in a cohort of newly-enrolled, adult members of the Vitality health promotion program.

Study sample and random allocation to groups: Newly enrolled adult members of the Discovery Health medical plan and who have, themselves, prospectively registered for the Vitality health promotion program, will be eligible for the study, and will be invited to participate by Discovery Health. This recruitment, enrollment and random allocation to groups will be initiated via a web-based email communication, from Discovery Health, over a period of 5 months. A final sample of 9000 adult members will be randomly allocated by Discovery Vitality (with replacement) to one of six study (incentive) groups (n=1500 per group) and followed each month for a period of 12 months.

The intervention arms include a Control group, in which members will be offered no additional incentives, other than the existing incentives, to engage in all points-earning physical activity. There will also be a Communication Group, in which members will be offered the base program and, in addition, will be sent bi-weekly communications by alternating email and text messaging which included information on the importance of physical activity, tips on accumulating "fitness points" on the base program, a status update of points earned and information on benefits and rewards offered.

The incentive arms are comprised of a Direct Payment Group, a Charity Incentive Group, a Lottery Pay-out Group, and a Choice Group. In each case, members will be offered the base program and supplementary communication described above. In addition, individuals will be offered the equivalent of one of four possible incentive amounts depending on the level of participation in fitness-related activities. Enrollment will be over a period of 5 months, and the intervention will take place over a period of 12 months, following registration, with a monthly pay-out for the incentive arms of the trial.

The following outcomes of interest will be measured:

• Number of new participants (registration) in the fitness programs of Vitality
• Points earned, as a proxy for levels of participation, for engaging in physical activity (gym visits)
• Self-reported measures of physical activity levels, smoking and healthy eating behaviors
• Health care claims (de-personalized) for the 12 month period following the 12 months intervention
• Survey of attitudes to incentives and rewards in the post-study questionnaire

Ethical considerations: Ethics approval has been obtained from the University of Cape Town/ University of Pennsylvania Research Ethics committees. Consent will be obtained from each subject, who agree to participate or 'opt out' and agree that their medical claims and utilization data may be examined in a de-personalised manner according to their group, and level of participation. Further, should a member choose to drop out in the course of the study, or refuse to participate, there will be no adverse impact on their medical plan coverage and they are free to withdraw without prejudice at any time during the course of the study.

Inclusion Criteria:

• Newly enrolled adult members of the Discovery Health medical plan who voluntarily registered and paid a monthly subscription for the health promotion program (Vitality) were eligible for the study. The recruitment, enrollment and random allocation to intervention arms were initiated via a web-based email communication. Initially, an opt-in design was used but because of an exceedingly low enrollment rate in the first three months of the study (less than 10%), with permission from the governing IRB, an opt-out strategy was employed for the remaining 5 months of the recruitment window, resulting in an enrollment rate of 90% during this time period. Altogether, 12121 eligible persons were recruited electronically, 1046 had invalid email addresses, 163 opted out and there was no information concerning age for 5 subjects.

Exclusion Criteria:

• Members who opted out. All members were voluntarily screened for participation in physical activity, and this is also done on site by participating gyms.