Fig. 1Distribution of the initial body mass index per town. Darker colors define increasing levels of weight (“normal,” “overweight,” “obesity”). Pie charts and map were realized with SAS software.
Fig. 2Distribution of the initial body mass index per area in United States (northeast, west, and midwest). The cities were grouped per area for the small size. Pie charts and map were realized with SAS software.
Fig. 3Scatter plot of the weight loss and intensity of air pollution. aDiff. BMI=final BMI–initial body mass index per user, bPollution=mean of the air quality index computed on the entire observation period per user.
Table 1Demographic characteristics (first part)
Characteristic |
Male (n=642) |
Female (n=1,966) |
P value |
Total (n=2,608) |
Age, yr |
42.575 (41.699 to 43.451) |
36.939 (36.426 to 37.451) |
<0.001 |
38.326 (37.874 to 38.778) |
Height, cm |
177.80 (177.30 to 178.40) |
164.60 (164.30 to 164.90) |
<0.001 |
167.881 (167.542 to 168.221) |
Weight, kg |
91.718 (90.286 to 93.150) |
75.556 (74.669 to 76.443) |
<0.001 |
79.535 (78.734 to 80.336) |
BMIa, kg/m2
|
30.948 (30.515 to 31.381) |
29.476 (29.144 to 29.807) |
<0.001 |
29.838 (29.565 to 30.111) |
Underweight (BMI <18.5) |
0 |
10 (0.51) |
NS |
10 (0.38) |
Normal (18.5<BMI<25) |
71 (11.06) |
614 (31.23) |
<0.001 |
685 (26.27) |
Overweight (25<BMI<30) |
258 (40.19) |
583 (29.65) |
<0.001 |
841 (32.24) |
Obesity class I (30<BMI<35) |
186 (28.97) |
365 (18.57) |
<0.001 |
551 (21.13) |
Obesity class II (35<BMI<40) |
77 (11.99) |
211 (10.73) |
NS |
288 (11.04) |
Obesity class III (BMI >40) |
50 (7.79) |
183 (9.31) |
NS |
233 (8.93) |
Daily calories, kcal/person insertion days |
1,382.40 (1,357.10 to 1,407.70) |
1,096.70 (1,084.30 to 1,109.10) |
<0.001 |
1,166.76 (1,154.57 to 1,178.96) |
Paired t-test (difference between final and initial BMI) |
|
|
|
|
Diff. BMI |
−2.412 (−2.652 to −2.172) (P<0.001) |
−2.051 (−2.181 to −1.919) (P<0.001) |
|
−2.139 (−2.255 to −2.024) (P<0.001) |
Users who lost weight (diff. BMI <0) |
542 (84.42) |
1,545 (78.59) |
|
2,087 (80.02) |
Users with stable weight (diff. BMI=0) |
8 (1.25) |
34 (1.73) |
|
42 (1.61) |
Users who gained weight (diff. BMI >0) |
92 (14.33) |
387 (19.68) |
|
479 (18.37) |
Table 2Distribution of pollutants and mean BMI variation by area (first part)
Area (n=2,608) |
Mean annual PM10a, µg/m3
|
Mean annual PM2.5a, µg/m3
|
Mean BMI variation, kg/m2 (95% Cl) |
Chicago (n=399) |
22 |
12 |
−2.474 (−2.811 to −2.137) |
Detroit (n=285) |
13 |
7 |
−2.506 (−2.897 to −2.114) |
Los Angeles (n=225) |
20 |
11 |
−2.313 (−2.737 to −1.888) |
New York (n=355) |
16 |
9 |
−2.502 (−2.887 to −2.117) |
Seoul (n=393) |
46 |
24 |
−1.261 (−1.452 to −1.071) |
Amsterdam (n=135) |
23 |
16 |
−1.853 (−2.259 to −1.447) |
Tokyo (n=67) |
28 |
15 |
−1.392 (−1.809 to −0.975) |
Berlin (n=512) |
24 |
16 |
−2.046 (−2.259 to −1.833) |
Sydney (n=57) |
17 |
8 |
−2.775 (−3.636 to −1.913) |
London (n=180) |
22 |
15 |
−2.358 (−2.824 to −1.893) |
Table 3Comparison of mean BMI variation (final BMI–initial BMI) between areasa
Area |
Difference in mean BMI variation |
95% Cl |
P value |
Seoul vs. Amsterdam |
0.592 |
−0.346 to 1.529 |
NS |
Seoul vs. Tokyo |
0.131 |
−1.112 to 1.373 |
NS |
Seoul vs. Berlin |
0.785 |
0.155 to 1.415 |
<0.05 |
Seoul vs. Los Angeles |
1.052 |
0.266 to 1.838 |
<0.05 |
Seoul vs. London |
1.097 |
0.251 to 1.943 |
<0.05 |
Seoul vs. Chicago |
1.213 |
0.545 to 1.881 |
<0.05 |
Seoul vs. New York |
1.241 |
0.553 to 1.929 |
<0.05 |
Seoul vs. Sydney |
1.513 |
0.181 to 2.845 |
<0.05 |
Seoul vs. Detroit |
1.245 |
0.513 to 1.976 |
<0.05 |
Table 4Effects of PM2.5 and PM10 on weight loss (first and second part)
Variable |
Coefficients estimate |
95% CI |
P value |
Multilevel model with users nested in cities (first part) |
PM2.5 |
0.085 |
0.054 to 0.115 |
<0.001 |
PM10 |
0.043 |
0.021 to 0.064 |
0.002 |
Sex (ref=male) |
0.735 |
0.400 to 1.070 |
0.001 |
Input frequency of dinner information |
−1.932 |
−2.340 to −1.523 |
<0.001 |
Age |
0.025 |
0.015 to 0.035 |
<0.001 |
Input frequency of physical activity information |
−0.642 |
−1.079 to −0.205 |
0.004 |
Mean daily caloric intake |
0.001 |
0.001 to 0.002 |
<0.001 |
Mixed effects model (second part)a
|
Time |
−0.0178 |
−0.0181 to −0.0175 |
<0.001 |
Pollution index×time |
1.5×10−5
|
9.01×10−6 to 2×10−5
|
<0.001 |
Sex (ref=male) |
0.704 |
0.391 to 1.019 |
<0.001 |
Input frequency of dinner information |
−1.732 |
−2.168 to −1.296 |
<0.001 |
Age |
0.026 |
0.015 to 0.036 |
<0.001 |
Input frequency of physical activity |
−0.918 |
−1.391 to −0.445 |
<0.001 |
Mean daily caloric intake |
0.004 |
0.003 to 0.006 |
<0.001 |