Table 2.
The estimates and 95% confidence intervals of hazard ratio (HR) of variables in the Cox models for assessing the impact of multiple variables of interest on the participant retention time in the first 43 weeks of the study for Phone-Active, Phone-Passive, and Fitbit-Passive data streams, respectively.
| Variables | Phone-Active | Phone-Passive | Fitbit-Passive | ||||
|---|---|---|---|---|---|---|---|
| HR (95% CI) | p | HR (95% CI) | p | HR (95% CI) | p | ||
| Age | <30 | Reference | – | Reference | – | Reference | – |
| 30–39 | 0.8 (0.53–1.2) | 0.28 | 0.81 (0.54–1.21) | 0.30 | 0.7 (0.4–1.22) | 0.21 | |
| 40–49 | 0.75 (0.48–1.17) | 0.20 | 0.76 (0.49–1.19) | 0.23 | 0.5 (0.27–0.93) | 0.03 | |
| 50–59 | 0.73 (0.46–1.15) | 0.18 | 0.74 (0.47–1.17) | 0.20 | 0.59 (0.33–1.08) | 0.09 | |
| >60 | 0.56 (0.34–0.93) | 0.02 | 0.56 (0.34–0.93) | 0.02 | 0.42 (0.22–0.81) | 0.01 | |
| Gender | Female | Reference | – | – | – | Reference | – |
| Male | 0.85 (0.63–1.15) | 0.29 | – | – | 0.79 (0.54–1.18) | 0.25 | |
| Marital status | Single | Reference | – | Reference | – | Reference | – |
| Married | 0.97 (0.72–1.31) | 0.85 | 0.96 (0.71–1.29) | 0.78 | 0.84 (0.58–1.23) | 0.37 | |
| Employment | No | Reference | – | Reference | – | Reference | – |
| Yes | 0.86 (0.63–1.16) | 0.32 | 0.85 (0.63–1.15) | 0.28 | 0.75 (0.5–1.11) | 0.14 | |
| Having children | No | Reference | – | Reference | – | Reference | – |
| Yes | 0.93 (0.66–1.31) | 0.68 | 0.94 (0.67–1.32) | 0.71 | 1.45 (0.93–2.26) | 0.10 | |
| Years in education | 1.01 (0.99–1.03) | 0.19 | 1.01 (0.99–1.03) | 0.21 | 0.99 (0.96–1.02) | 0.49 | |
| Annual income (£/€) | <15,000 | Reference | – | Reference | – | Reference | – |
| 15,000–55,000 | 1.13 (0.82–1.57) | 0.46 | 1.14 (0.82–1.59) | 0.43 | 0.95 (0.64–1.42) | 0.81 | |
| >55,000 | 0.98 (0.6–1.61) | 0.95 | 1 (0.61–1.65) | 0.99 | 0.75 (0.38–1.47) | 0.40 | |
| Accommodation | Own outright | Reference | – | Reference | – | Reference | – |
| Renting | 1.14 (0.84–1.54) | 0.40 | 1.14 (0.84–1.54) | 0.41 | 0.84 (0.57–1.24) | 0.38 | |
| Living rent-free | 0.8 (0.46–1.39) | 0.42 | 0.79 (0.45–1.38) | 0.41 | 0.59 (0.29–1.21) | 0.15 | |
| PHQ8 score | 1.01 (0.98–1.03) | 0.59 | 1.01 (0.99–1.03) | 0.50 | 1 (0.97–1.03) | 0.81 | |
| Having comorbidities | No | Reference | – | Reference | – | Reference | – |
| Yes | 0.88 (0.68–1.16) | 0.37 | 0.88 (0.67–1.15) | 0.33 | 1.17 (0.82–1.66) | 0.39 | |
| Medication | No | Reference | – | Reference | – | Reference | – |
| Yes | 1.14 (0.86–1.51) | 0.36 | 1.13 (0.86–1.5) | 0.38 | 1.06 (0.72–1.55) | 0.77 | |
| Study site | CIBER | Reference | – | Reference | – | Reference | – |
| KCL | 0.86 (0.59–1.26) | 0.43 | 0.87 (0.59–1.27) | 0.47 | 0.59 (0.38–0.94) | 0.03 | |
| VUMC | 1.49 (0.95–2.34) | 0.09 | 1.54 (0.98–2.43) | 0.06 | 0.4 (0.21–0.76) | <0.001 | |
| Brand of smartphone | Other | Reference | – | Reference | – | Reference | – |
| Motorola | 0.26 (0.17–0.4) | <0.001 | 0.26 (0.17–0.4) | <0.001 | 0.56 (0.34–0.92) | 0.02 | |
| Samsung | 0.58 (0.43–0.79) | <0.001 | 0.57 (0.42–0.77) | <0.001 | 0.78 (0.53–1.16) | 0.22 | |
| Phone status | Own phone | Reference | – | Reference | – | Reference | – |
| Provided phone | 1.67 (1.06–2.64) | 0.03 | 1.65 (1.05–2.6) | 0.03 | 1.61 (0.93–2.77) | 0.09 | |
Note, in the Phone-Passive model, an interaction term of the “gender” variable with a split time variable (cut points are 130 and 240) was used to make the variable meet the model assumption. HR of Male (1–129) = 0.67 (0.39–1.14) (p = 0.14), HR of Male (130–239) = 0.58 (0.34–0.99) (p = 0.05), and HR of Male (240–301) = 1.52 (0.95–2.42) (p = 0.08).