Table 3.
Results of individual studies—randomized controlled trials.
| Serial # | Paper reference | Dataset size | Variables evaluated | Results |
| 1 | Soetens et al [65] | 803 | Effect of time over increase in PAa | PA increases in all groups, time has no significant effect on all completers though has significant effect on those who had low baseline scores for total PA minutes (P<.001) |
| 5 | Storm et al [7] | 790 | Strength of habit for PA measured with abbreviated version of Self-Reported Habit Index, self-efficacy, and planning | Self-efficacy (P=.1), planning (P=.2), and habit strength (P=.006) improved in the intervention group |
| 6 | Schulz et al [47] | 5055 | Minutes of PA per day in control, sequential intervention module delivery, and simultaneous module delivery | No statistical difference in sequential and simultaneous delivery for PA or with respect to control group. Sequential delivery could be more effective than simultaneous module delivery after 12 months (P=.7) |
| 13 | Yom-Tov et al [9] | 27 | PA minutes per week, change in activity with message policy, change from initial to RLb-based learned policy | No statistical difference in treatment and control arm (P=.30) for PA minutes per week. Difference in change of activity between initial and learned message policy statistically significant (P=.004) |
| 15 | Cook et al [30] | 555 | PA (minutes per week) behavior difference at baseline and postmeasurement for 3 parameters: commuting, leisure time PA, and PA in school | Improvement found in leisure time MVPAc (P<.05), for increase in commute by bicycle (around 30 min) (P<.01) and total MVPA (P<.05) |
| 17 | Short et al [84] | 724 | Minutes per week of MVPA and resistance training score for all 3 arms–3 module interventions delivered monthly, weekly, or single-module | Significant improvement of MVPA across all groups (P<.05). Significant improvement in resistance score from monthly 3-module intervention to single module (P=.01) |
| 21 | Irvine et al [83] | 368 | Cardiovascular exercises, stretching exercises, strength exercises, balance exercises (all measured in minutes per week), and number of activities | Improvement in intervention group as compared with control in all (P<.001) |
| 22 | Friederichs et al [38] | 4302 | Minutes of MVPA per week and number of days ≥30 min activity in I-Move intervention, Active Plus intervention, and control group | I-Move had small but more significant effect than Active Plus in minutes of MVPA per week (P=.03 and P=.07). I-Move had medium sized effect and Active Plus had large size effect for number of days ≥30 min |
| 23 | Blake et al [108] | 296 | Active travel, moderate activity at work and recreation and vigorous activity at work and recreation in 2 arms for different delivery modes, both with tailored content, one with SMSd and another with email | No significant difference between email and SMS, but significant difference in moderate activity at work (hours per day), with email more effective than SMS (P=.24). |
| 25 | Hargreaves et al [93] | 97 | Step count | No difference at baseline and 12 weeks. Significant increase in step count of intervention group between week 12 and week 24 (P=.055) but not so significant in comparison group (P=.15) |
| 30 | Mistry et al [90] | 337 | PA between the 3 groups–standard care, generic message, and intervention group after 4 weeks | No significant difference between groups for change in PA (P>.05) |
| 31 | Peels et al [27] | 1729 | Number of METe hours in 4 kinds of tailoring: printed, and Web-based (basic and environment-based in each) and control group | Printed (both basic and environmental) had statistically significant increase in MET hours (P=.025 and P=.31, respectively). No significant increase in both Web-based interventions (P=.59 and P=.887, respectively) |
| 34 | Pyky et al [41] | 496 | Self-rated health and fitness and leisure time PA | Changes in self-rated fitness and leisure time PA are associated with improved self-rated health (P<.026 and P<.04, respectively). No significant difference between intervention and control for self-reported daily sitting (P=.32) and light housework (but no other leisure time) PA (P=.43) |
| 37 | Marsaux et al [35] | 1607 | Objective PA in control group, group with personalized advice on diet and PA (L1 group), L1+phenotype (L2 group) and L2+genotype (L3 group) | No significant difference between control and any of the 3 groups in objective PA level measured (P=.73) |
| 38 | Alley et al [43] | 154 | PA (min per week) for 3 groups: control, tailoring only, and tailoring+video coaching group | Significant difference in PA between tailoring+video coaching versus control group (P=.01) but no significant difference in PA between the 2 intervention groups (P=.54) |
| 39 | Mitchell et al [60] | 171 | Sedentary time, LPAf, and MVPA for intervention group with personalized step goals versus control group with generic advice | Decrease in sedentary time, Improvement in LPA and MVPA for both groups (P<.005). |
| 41 | De Cocker et al [64] | 312 | Sitting time in 3 groups: control, generic intervention, and tailored intervention | Self-reported total sitting time decreased more in tailored group compared with both generic group (P=.002) and control group (P=.002). But no significant difference in objectively measured data |
| 47 | Kattelmann et al [72] | 1639 | Total MET-minutes per week estimated from self-reported data | No difference between control and intervention for total MET-minutes per week (P=.90). Significant time effect for moderate MET-minutes per week (P=.002) and significant time × group × gender effect for vigorous MET-min per week (P=.05) |
| 48 | Partridge et al [68] | 214 | Self-reported PA data analyzed as MET-minutes per week | Significant effect of intervention on average MET minutes per week at 12 weeks (P=.05). Total PA days (P=.003) and number of walking days (P=.02) increased in intervention group |
| 49 | Walthouwer et al [71] | 1419 | PA duration in text-tailored, video-tailored, and control arm. In the tailoring group, 2 groups were compared, 1 where preference of user to video/text was matched and another without the matching | No significant difference in condition match/mismatch for PA (P=.33). Also, no significant difference for video-tailoring × intervention used (P=.83) and text-tailoring × intervention used (P=.81) |
aPA: physical activity.
bRL: reinforcement learning
cMVPA: moderately vigorous physical activity.
dSMS: short messaging service.
eMET: metabolic equivalent.
fLPA: light physical activity.