HTML
-
A total of 158 participants aged 25-59 years were eligible for analysis. Ten subjects (3 men and 7 women) were excluded from our study because of incomplete data. Demographic and SES information of study participants are shown in Table 1. According to our sampling procedure, participants were almost evenly split by gender, with similar age distribution between the sexes. Women tended to have lower BMI (P = 0.002) and WHtR (P= 0.030) than men. As for socioeconomic status, education attainment and income were not significantly different between genders. The percentage of unemployed was higher in women because of different legal retirement age requirements.
Items Men (N = 81) Women (N = 77) P Demographic Info Age (y) 42.9 ± 10.2 43.5 ± 10.3 0.729 Height (cm) 173.1 ± 5.2 160.5 ± 4.5 < 0.001 Weight (kg) 72.3 ± 8.7 58.6 ± 8.9 < 0.001 Waist circumference (cm) 85.7 ± 6.8 77.1 ± 7.8 < 0.001 BMI (kg/m2) 24.1 ± 2.6 22.7 ± 2.9 0.002 WHtR 0.50 ± 0.04 0.48 ± 0.05 0.030 SES Info Education beyond primary school (n, %) 79 (97.5%) 70 (90.9%) 0.073 Unemployment (n, %) 4 (4.9%) 22 (28.6%) < 0.001 Per capita disposal annual income (10 k RMB) (percent of missing values on income variable) 3.1 ± 1.7 (17.2) 3.3 ± 1.8 (14.3) 0.410 Table 1. Demographic and SES Information of Study Participants by Gender
Table 2 shows the dimensions and intensities of PA measured by the IPAQ-L and pedometer. Among four dimensions of PA, men tended to participate in more work-related and less domestic PA than women in both surveys (P ≤ 0.001). Among the three intensities of PA, men tended to spend more time in vigorous intensity PA than women in both surveys (both P < 0.001). No significant difference was found in the total amount of PA between men and women. In the first and second surveys, 7.5% and 9.4% of all participants reported themselves as physically inactive (not shown in the table).
Items Total (N = 158) Men (N = 81) Women (N= 77) P First Survey (day 1) (MET-min/week) Dimensions of PA Work-related PA 330 (1, 659) 417 (2, 049) 198 (746) < 0.001 Transportation PA 660 (1, 224) 459 (1, 151) 870 (1, 362) 0.097 Housework PA 323 (540) 180 (450) 420 (1, 080) < 0.001 Leisure Time PA 281 (891) 306 (983) 240 (893) 0.073 Intensities of PA Walking 990 (1, 716) 957 (1, 485) 1320 (2, 071) 0.843 Moderate intensity 865 (1, 676) 810 (1, 260) 1080 (1, 970) 0.092 Vigorous intensity 0 (480) 0 (720) 0 (0) < 0.001 Total PA 2, 460 (3, 540) 2, 389 (3, 676) 3050 (4, 191) 0.597 Second Survey (day 9) (MET-min/week) Dimensions of PA Work-related PA 344(1, 223) 594 (1, 884) 198 (576) 0.001 Transportation PA 486 (1, 221) 413 (1, 233) 594 (1, 188) 0.002 Housework PA 270 (660) 150 (540) 360 (1, 080) < 0.001 Leisure time PA 297 (722) 438 (847) 231 (737) 0.016 Intensities of PA Walking 883 (1, 493) 858 (1, 535) 891 (1, 485) 0.841 Moderate intensity 870 (1, 650) 840 (1, 470) 900 (1650) 0.022 Vigorous intensity 0 (290) 0 (720) 0 (0) < 0.001 Total PA 2, 725 (3, 469) 2, 738 (3, 153) 2, 591 (3, 392) 0.341 Pedometer (day 2 to day 8) Total (N = 80) Men (N = 41) Women (N = 39) Ex (MET-min/ week) 1, 141 (764) 1, 287 (810) 990 (674) 0.841 Walking distance (km)a 26.1 ± 12.6 29.6 ± 13.8 22.4 ± 10.3 0.061 Note. a: Mean ± SD was used for walking distance. Table 2. IPAQ-L and pedometer measured physical activity (median with interquartile range)
-
Test-retest reliability results are presented in Table 3. Among the four dimensions of PA, the ICC for leisure time PA was the lowest (ICC = 0.37; 95% CI: 0.23, 0.50), and the remaining dimensions performed quite well. The housework PA showed the biggest ICC of 0.73. ICC of total PA was 0.67. The analysis of PA by intensity showed at least a fair level of repeatability for each intensity. The test-retest reliability was highest for walking (ICC = 0.71), followed by moderate PA (ICC = 0.65) and was lowest for vigorous PA (ICC = 0.56). The percentage agreement for the distribution of low, moderate and high categories between two surveys was 72.2%, with a kappa statistic of 0.515 (P < 0.001).
Items ICCa (n = 158) 95% CI Spearman's ρ by Age Spearman's ρ by Gender Age < 40 (n = 64) Age ≥ 40 (n = 94) Z Score P Men (n = 81) Women (n = 77) Z Score P Dimensions of PA Work-related PA 0.69 (0.59, 0.76) 0.60 0.73 1.42 0.156 0.72 0.64 0.92 0.358 Transportation PA 0.44 (0.30, 0.56) 0.53 0.60 0.62 0.535 0.70 0.46 2.28 0.023 Housework PA 0.73 (0.65, 0.80) 0.71 0.69 0.24 0.810 0.71 0.69 0.24 0.810 Leisure time PA 0.37 (0.23, 0.50) 0.48 0.49 0.24 0.810 0.58 0.38 1.62 0.105 Intensities of PA Walking 0.71 (0.62, 0.78) 0.62 0.66 0.41 0.682 0.75 0.51 2.53 0.011 Moderate intensity 0.65 (0.55, 0.73) 0.78 0.64 1.74 0.082 0.74 0.65 1.08 0.280 Vigorous intensity 0.56 (0.45, 0.66) 0.40 0.69 2.56 0.011 0.60 0.48 1.05 0.294 Total PA 0.67 (0.57, 0.75) 0.62 0.76 1.64 0.101 0.77 0.68 1.18 0.238 Note.a: One-way random effect single measure ICC, with all P < 0.001. Table 3. Test-retest Reliability of IPAQ-L
Spearman's ρ for subjects aged < 40 years tended to indicate lower reliability for vigorous PA (0.40 vs. 0.69, P = 0.011) compared to subjects aged ≥ 40 years. Compared to women, men tended to have higher reliability for transportation-related PA (0.70 vs. 0.46, P = 0.023) and walking behaviour (0.75 vs. 0.51, P = 0.011). No significant difference was found in total PA between genders.
-
Table 4 shows the results of criterion validity using pedometer data as a reference. Spearman's rank correlation showed walking, vigorous PA, and total PA were fairly related to the Ex of the pedometer, with Spearman's ρ of 0.36, 0.38, and 0.32, respectively. After adjusting for gender, age, BMI, education attainment and employment status, walking, vigorous and total PA showed only a moderate but still significant correlation, with partial r of 0.32, 0.26, and 0.27. Correspondingly, similar correlation results were observed between walking, vigorous exercise, and total PA with the indicator of Walking Distance from the pedometer. Before or after the adjustment, the correlation of IPAQ-L with Walking Distance were always lower than correlations with the Ex.
IPAQ-L (MET-min/week) Spearman's ρ Partial Correlation ra Ex (MET-min/ week) P Walking Distance (km) P Ex (MET-min/ week) P Walking Distance (km) P Walking 0.36 0.001 0.32 0.004 0.32 0.005 0.29 0.013 Moderate PA 0.19 0.091 0.15 0.179 0.04 0.739 0.10 0.391 Vigorous PA 0.38 0.001 0.33 0.003 0.26 0.024 0.29 0.011 Total PA 0.32 0.004 0.31 0.005 0.27 0.020 0.31 0.007 Note.a: Adjusted for gender, age, BMI, education attainment, employment. Table 4. Criterion validity of IPAQ-L using pedometer data as standard (n = 80)
The Bland Altman Plot of IPAQ-L and pedometer Ex indicator for total PA (Figure 1) showed a small mean difference (-91 MET-min/week), but a 95% limit of agreement (-4, 615 to 4, 433 MET-min/week) was enlarged because of the existence of five outliers. When the outliers were removed, the difference in total PA presented a normal distribution with much narrower 95% limits of agreement (-2, 435 to 1, 222 MET-min/week). Figure 2 shows the Bland Altman Plot of walking PA for IPAQ-L and pedometer Ex indicator. The mean difference was small (58 MET-min/week). Similar to Figure 1, the 95% limit of agreement (-2, 319 to 2, 435 MET-min/week) was affected by an outlier. The mean difference was almost near zero (-0.70 MET-min/week), and the 95% limit of agreement was-2, 159 to 2, 157 MET-min/weekafter removing the outlier.