TABLE 1.

Strategies and approaches with evidence of effectiveness for the general population and selected populations.^a

Category	Intervention Approaches	Evidence Type (Number of Studies)	Selected Effect Sizesb
Physical activity promotion interventions Individual Behavior change theories and models	Behavior change theories and techniques (general adult as well as more specific populations)	1 SR (12) 1 MA (11)	One meta-analysis reported a positive effect of providing lottery and escalating incentives on exercise session attendance when compared with no incentive for short-duration interventions lasting 4 to 26 wk; pooled results showed an increase in exercise attendance of 11.55%; 95% CI, 5.61%–17.50% (12).
Interventions specific to older adults	Effective strategies include identifying & problem-solving around age-specific PA barriers, social support, modeling of PA behaviors, rewards linked with PA behavior change	2 SR (158 and 18) 1 MA (24)	Interventions had a small effect on physical activity d = 0.14 (95% CI, 0.09–0.20, P < 0.001) (13).
Interventions specific to youth	Effective strategies include in-person education & experiential strategies (e.g., exercise classes) Above can be enhanced through family inclusion	2 MA (47 and 58)	1. Significant effect sizes for interventions targeting individuals only (g = 0.27; 95% Cl, 0.12–0.42), which were further enhanced when individual interventions also included families (g = 0.44; 95% Cl, 0.23–0.66) or school and print or digital media (e.g., newspaper, radio; g = 0.30; 95% Cl, 0.04–0.57) (14).
			2. Family-based physical activity interventions found a small but significant effect size favoring the intervention group (SMD, 0.41; 95% Cl, 0.15–0.67) (15).
Peer-Led	Peer-led behavioral self-management interventions (older adults and individuals with chronic disease)	1 MA (21)	Moderate effects for increases in physical activity overall among the 17 studies where effects sizes were available (SMD, 0.4; 95% Cl, 0.22–0.55, P< 0.001) (16).
Community Community-wide	Community-wide interventions that use intensive contact with the majority of the target population	2 SR (16 and 33) 1 MA (10) PAGMR	Effect sizes not mentioned in PAGAC chapter.
School interventions	Multiple-component programs occurring during school hours aimed at physical activity across the school day in primary school-age (typically ages 5 to 12 yr) and adolescent youth Revising the structure of physical education (PE) classes to increase in-PE physical activity in primary school-age and adolescent youth	5 SR (range, 8–129) 2 MA (13 and 15) AHA PAGMR	1. CATCH and SPARK trials: Results showed that vigorous physical activity was significantly higher among intervention students (mean (M) = 58.6 min) compared to controls (M = 46.5 min) (P< 0.003) (17). Students spent more minutes per week being physically active in teacher- and specialist-led classes compared to controls (33 min, 40 min, and 18 min, respectively, P< 0.001), although PA did not increase outside of school (18).
			2. Absolute difference of 10.37% (95% Cl, 6.33–14.41) of lesson time spent in moderate-to-vigorous physical activity in favor of the interventions over controls. This estimated difference of 10.37% of lesson time corresponds to 24% more active learning time in the intervention groups compared with the control condition (SMD, 0.62; 95% Cl, 0.39–0.84) (19).
Communication environment (information and communication technologies) Wearable activity monitors	Wearable activity monitors, including step counters (pedometers) and accelerometers, when used in conjunction with goal-setting and other behavioral strategies (general population of adults and those with type 2 diabetes or with overweight or obesity)	4 SR (range: 5–14) 3 MA (11 for each)	1. Accelerometry interventions across 12 trials resulted in a small but significant increase in physical activity levels (SMD, 0.26; 95% Cl, 0.04–0.49). The additional benefit of activity monitors when compared with an active comparison arm (e.g., a physical activity intervention without activity monitors) is less clear (SMD, 0.17; 97% Cl, −1.09 to 1.43) (20).
			3. Type 2 diabetes: step-counter use significantly increased physical activity by a mean of 1822 steps per day (7 studies, 861 participants; 95% Cl, 751–2894 steps per day), use of a step-counter in combination with setting a specific physical activity goal resulted in significantly more steps per day compared to control arms (weighted mean difference (WMD) of 3200 steps per day; 95% Cl, 2053–4347 steps per day), whereas step-counter use without a goal did not significantly increase physical activity relative to control arms (WMD of 598 steps per day; 95% Cl, −65 to 1260 steps per day). Use of a step diary or log also was related to a statistically significant increase in physical activity (WMD = 2816 steps per day), whereas when a step diary was not used, physical activity did not increase significantly (WMD = 115 steps per day) (21).
			4. Overweight/Obesity: a significant positive intervention effect for steps per day was found for behavioral physical activity interventions that included an activity monitor when compared with waitlist or usual care interventions (n = 4) (SMD, 0.90; 95% Cl, 0.61–1.19, P< 0.0001). A similar intervention comparison also found a significant positive effect for total moderate-to-vigorous physical activity minutes per time unit (n = 3) (SMD, 0.50; 95% Cl, 0.11–0.88, P= 0.01). In a meta-analysis of a similar intervention comparison (i.e., the addition of an activity monitor to an existing intervention versus when it was not added) using the mean difference for walking MET-minutes per week as the outcome and involving only two studies (both of which included women only), a statistically significant positive effect was found (mean difference for walking MET-minutes per week = 282; 95% Cl, 103.82–460.18, P= 0.002) (22).
Telephone-assisted interventions	Telephone-assisted interventions (general adult population, including older adults)	2 SR (11 and 27)	Effect sizes not mentioned in the PAGAC chapter.
			“The majority of high-quality studies in this area produced effect sizes indicating a moderate or better intervention effect (i.e., d> 0.5). The evidence indicates that longer-duration interventions (i.e., 12 months or more) are associated with greater effectiveness.”
Web-based or Internet-delivered interventions	Internet-delivered interventions that include educational components (general adult population)	3 SR (range: 7–15) 1 MA (34)	1. Overall effect size estimates indicate a small but positive intervention effect on physical activity in the general adult population (d = 0.14). Studies that initially screened participants and enrolled only those classified as sedentary or insufficiently active produced larger effects (d = 0.37) relative to studies that did not screen participants for physical activity level (d = 0.12). The meta-analysis, which targeted either physical activity only (n = 21) or physical activity and additional health-related behaviors, such as nutrition or weight management behaviors (n = 13), found that the two different types of interventions produced similar effect sizes (23).
			2. In a systematic review of nine web-based physical activity interventions in individuals with type 2 diabetes, six studies reported significant short-term increases (less than 6 months, typically) in physical activity when compared with a control arm. The overall magnitude of the physical activity increases reported in this review ranged from 3% to 125% (24).
			3. In a systematic review of seven self-guided web-based physical activity intervention trials among patients with a range of chronic disease conditions (e.g., multiple sclerosis, heart failure, type 2 diabetes mellitus, physical disabilities, metabolic syndrome), three studies reported significant physical activity improvements relative to controls, while four studies reported nonsignificant differences between groups. Effect sizes ranged from 0.13 to 0.56, with wide variability in physical activity change across studies (25).
Computer-tailored print interventions	Computer-tailored print interventions that collect user information through mailed surveys that is then used to generate computer-tailored mailings containing personalized physical activity advice and support (general adult population)	2 SR (11 and 26)	The majority of studies in this area produced effect sizes that were small (i.e., Cohen’s d ranging from 0.12 to 0.35) when compared to minimal or no-intervention control arms. (F11–57)
Mobile phone programs	Mobile phone programs consisting of or including text-messaging (general adult population) Use of smartphone applications (children and adolescents)	5 SR (range: 9–30) 3 MA (range: 11–74)	1. General adult population: text messaging interventions aimed at general adult populations found significant positive effect sizes, relative to controls, that were on average 0.40 or greater, with a median effect size in one systematic review of 0.50 (26).
			2. Smartphone apps - youth: Interventions have been reported to have small to moderate effects in both girls and boys, with one systematic review reporting Cohen’s d coefficients ranging from −0.36 to 0.86. As part of these meta-regression analyses, investigators were able to explain 45 percent of the variability in physical activity effect size among children and 62 percent of the variability in physical activity effect size among adolescents (27).
Physical environment and policy Point-of-decision prompts to promote stair use	Point-of-decision prompts to use stairs versus escalators or elevators (adults)	2 SR (6 and 67) AHA	In one systematic review of 67 studies, 77 percent reported increases in stair use. For those studies with significant effects (n = 55 studies), the percent stair use increase ranged from 0.3 percent to 34.7 percent. When odds ratios were reported, they ranged from 1.05 (95% Cl, 1.01–1.10) to 2.90 (95% Cl, 2.55–3.29) (28).
Built environment characteristics that support active transport	Built environment characteristics and infrastructure that support active transport to destinations (e.g., Safe Routes to School programs, street connectivity, a mix of residential, commercial, and public land uses) (children, adults, and older adults)	3 SR (range: 12–42) 1 MA (42) AHA The Guide to Community Preventive Services	1. RESIDE study: each unit increase in perceived safety from crime was associated with 3.2 min-wk−1 more of transport physical activity. In addition, the association remained similar (3.6 min-wk−1 increases with unit increases in perceived safety from crime) when also controlling for built environmental characteristics such as residential density, streets connectivity, and number of local destinations (29).
			2. Youth: positive associations found between land-use mix and children’s physical activity (OR ranged from 1.8 (95% Cl, 1.05–3.42) to 3.46 (95% Cl, 1.6–7.47)). (AHA Scientific Statement)
			3. Older adults: A meta-analysis of 42 quantitative studies found significant positive associations among a number of environmental variables and active transport behaviors, including residential density and urbanization, walkability, easy access to building entrances, and access to and availability of services and destinations. A weak, negative association was found between neighborhood disorder (e.g., litter, vandalism and decay) and total walking for transport (30).
Community design and characteristics that support recreational physical activity	Community design and characteristics that support physical activity, such as having safe and readily usable walking and cycling infrastructure and other favorable built environment elements (children and general adult populations)	1 SR (600) AHA The Guide to Community Preventive Services	1. RESIDE study: each unit increase in perceived safety from crime was associated with 13.5 min-wk−1 more of recreational physical activity over a 7-yr follow-up period. This amount of increase remained similar (13.7 min-wk−1) when also controlling for additional built environmental characteristics (i.e., residential density, streets connectivity, and mix of local destinations) (29).
			2. Adults: the largely cross-sectional studies reviewed by the AHA Scientific Statement generally indicated a significant relationship between neighborhood aesthetics and leisure-time physical activity, walking, or meeting physical activity recommendations (ORs ranged from 1.13 to 2.6). Absence of heavy traffic was associated with more walking and leisure-time physical activity (OR = 1.22; 95% Cl, 1.08–1.37). (AHA Scientific Statement)
Access to indoor and/or outdoor recreation facilities or outlets	Having access to indoor (e.g., gyms) and/or outdoor recreation facilities or outlets, including parks, trails, and natural or green spaces (children and general adult populations)	3 SR (range: 12–90) AHA	Greater access generally was shown to be related to more physical activity among adults (OR 1.20; 95% Cl, 1.06–1.34). (AHA Scientific Statement)
Sedentary behavior reduction interventions Youth	Reductions in television viewing and other screen-time behaviors in primarily school-based settings (youth) through school-based counseling, parental involvement, tailored feedback, and use of screen allowance devices at home	5 SR (range: 10–22) 4 MA (range: 13–34)	1. As a whole, the studies reviewed showed small but consistent effects on sedentary behavior reduction (e.g., mean difference was −20.44 min-d−1; 95% Cl, −30.69 to −10.20) (31).
			2. School-based interventions focused primarily on reducing screen time in children through in-class or after-school curricula, and typically included messages targeting screen time as well as other health behaviors (e.g., exercise, diet). Such interventions had small but consistent effects in reducing sedentary time, particularly for those lasting longer than 6 months (e.g., mean difference was −0.25 h-d−1; 95% Cl, −0.37 to −0.13) (32).
Worksite	Interventions targeting sedentary behavior in worksites (adults), including physical changes to workstations	2 SR (15 and 40) 2 MA (8 and 21)	Interventions that focused on providing educational or motivational support showed only small and inconsistent effects on sedentary behavior (e.g., mean difference was −15.52 min per 8-h workday [95% Cl, −22.88 to −8.16]). Interventions that targeted physical changes to work stations (i.e., predominantly the addition of sit-stand workstations, with a few that used treadmill desks or portable pedal machines) had consistently medium to large effects (e.g., mean difference was −72.78 min per 8-h workday [95% Cl, −104.92 to −40.64]). Additionally, these effects were stronger when these types of work station changes were combined with educational and behavioral support (e.g., mean difference was −88.80 min per 8-h workday [95% Cl, −132.69 to −44.61]) (33).

^aIntervention categories derived from a comprehensive search of eligible systematic reviews and meta-analyses along with selected governmental reports identified during a 2011–2016 evidence search.

^bEffect sizes are from articles mentioned in manuscript. Due to space limitations, not all articles included in the Committee Report were captured in the article; however, a full list of all systematic reviews, meta analyses, and effect sizes can be found in 2018 Physical Activity Guidelines Advisory Committee Report.

SR, systematic review; MA, meta-analysis; PAGMR, Physical Activity Guidelines Midcourse Report; AHA, American Heart Association Scientific Statement.