Table 2.
Association of average DM across childhood (intercept) and increase of DM over time (slope) with metabolic syndrome score and its components in children and adolescents
| Metabolic outcomes | DM use | Analysis populationa | Accelerometer –sample b | |||||
|---|---|---|---|---|---|---|---|---|
| N | Crude β (95%CI) | N | Adjusted β (95%CI) a | N | Adjusted β (95% CI) b | Adjusted β (95% CI) c | ||
| z_WC d | Intercept | 10,301 | 0.33 (0.28, 0.39) | 10,153 | 0.15 (0.11, 0.19) | 4258 | 0.19 (0.13, 0.26) | 0.19 (0.12, 0.25) |
| Slope | -0.05 (-0.40, 0.29) | 0.19 (-0.04, 0.43) | 0.26 ( -0.10, 0.63) | 0.26 (-0.10, 0.63) | ||||
| z_BP e | Intercept | 10,099 | 0.04 (0.01, 0.08) | 9409 | 0.04 (0.00, 0.07) | 4073 | 0.02 (-0.03, 0.08) | 0.02 (-0.03, 0.07) |
| Slope | -0.05 (-0.27, 0.16) | 0.09 (-0.10, 0.30) | 0.18 (-0.13, 0.49) | 0.16 (-0.14, 0.47) | ||||
| z_TRG | Intercept | 7398 | 0.11 (0.07, 0.15) | 6193 | 0.08 (0.03, 0.12) | 2683 | 0.06 (-0.01, 0.13) | 0.06 (-0.01, 0.13) |
| Slope | 0.17 (-0.08, 0.44) | 0.02 (-0.24, 0.30) | 0.01 (-0.40, 0.42) | 0.00 (-0.41, 0.41) | ||||
| z_HDL-c | Intercept | 7766 | -0.10 (-0.14, -0.06) | 6506 | -0.08 (-0.12, -0.04) | 2857 | -0.07 (-0.14, -0.002) | -0.07 (-0.14, -0.00) |
| Slope | -0.35 (-0.63, -0.06) | -0.28 (-0.54, -0.02) | -0.32 (-0.72, 0.07) | -0.33 (-0.73, 0.06) | ||||
| z_HOMA | Intercept | 6293 | 0.16 (0.11, 0.22) | 3435 | 0.12 (0.05, 0.19) | 1688 | 0.15 (0.05, 0.26) | 0.15 (0.04, 0.25) |
| Slope | 0.19 (-0.13, 0.51) | 0.64 (0.21, 1.08) | 0.58 (-0.01, 1.18) | 0.59 (0.00, 1.19) | ||||
| z_MetS | Intercept | 5770 | 0.16 (0.11, 0.21) | 2973 | 0.07 (0.02, 0.13) | 1476 | 0.07 (-0.01, 0.15) | 0.06 (-0.02, 0.14) |
| Slope | 0.14 (-0.15, 0.44) | 0.54 (0.20, 0.88) | 0.51 ( 0.04, 0.97) | 0.49 (0.02, 0.95) | ||||
a Models are adjusted for age (continuous), sex, pubertal status, HDAS, snack consumption, parental ISCED, observation period, (age at follow-up – age at baseline), country and baseline z-scores of the respective outcome. Bold significance is provided via confidence limits
b Models are based on the accelerometer sample and are adjusted for same confounders as in the main analysis. N varied due to missing values for each outcome
c Models based on sample with accelerometer data are further adjusted for MVPA, SED and valid accelerometer wear time
d WC- waist circumference, BP-blood pressure, TRG- triglycerides, HDL-c– high density lipoprotein cholesterol, HOMA- homeostasis model assessment for insulin resistance, MetS- metabolic syndrome, DM- digital media
e Models for the z-scores of BP, HDL-c, TRG and HOMA-IR are additionally adjusted for z-score of WC at the last measurement point. The number of participants varied for metabolic outcomes due to missing values