Predictors of BMI-SDS
| Effect | Univariable regression* | Multivariable regression† | ||||
|---|---|---|---|---|---|---|
| Effect size estimate (95% CI) | p Value | Male | Female | |||
| Effect size estimate (95% CI) | p Value | Effect size estimate (95% CI) | p Value | |||
| Age at onset | <0.001 | <0.001 | <0.001 | |||
| 15–≤20 | −0.106 (−0.151 to −0.061) | <0.001 | −0.148 (−0.222 to −0.074) | <0.001 | 0.021 (−0.079 to 0.121) | 0.951 |
| 10–<15 | 0.004 (−0.018 to 0.026) | <0.708 | −0.074 (−0.113 to −0.036) | <0.001 | 0.093 (0.051 to 0.135) | <0.001 |
| 5–<10 | −0.082 (−0.103 to −0.060) | <0.001 | −0.104 (−0.142 to −0.066) | <0.001 | −0.057 (−0.096 to −0.018) | 0.001 |
| <5 (ref.) | 0 | – | 0 | – | 0 | – |
| Diabetes duration | <0.001 | <0.001 | <0.001 | |||
| ≥10 years (ref.) | 0 | – | 0 | – | 0 | – |
| 5–<10 years | −0.008 (−0.027 to 0.010) | 0.365 | 0.110 (0.076 to 0.144) | <0.001 | −0.099 (−0.134 to −0.064) | <0.001 |
| 2–<5 years | 0.031 (0.012 to 0.049) | 0.001 | 0.210 (0.174 to 0.246) | <0.001 | −0.097 (−0.134 to −0.060) | <0.001 |
| <2 years | −0.016 (−0.035 to 0.003) | 0.109 | 0.203 (0.163 to 0.242) | <0.001 | −0.149 (−0.190–0.109) | <0.001 |
| Sex | ||||||
| Male (ref.) | 0 | – | 0 | |||
| Female | 0.123 (0.107 to 0.139) | <0.001 | 0.149 (0.130 to 0.168)‡ <0.001 | |||
| Insulin regimen | <0.001 | <0.001 | ||||
| CSII (ref) | 0 | – | 0 | |||
| 4+injections | −0.028 (−0.040 to −0.017) | <0.001 | −0.010 (−0.029 to 0.008)§ 0.463 | |||
| 3 injections | −0.062 (−0.077 to −0.048) | <0.001 | −0.045 (−0.068 to −0.022) <0.001 | |||
| ≤2 injections | −0.050 (−0.068 to −0.031 | <0.001 | −0.031 (−0.060 to −0.002) 0.029 | |||
| Insulin dose (IU/kg) | 0.059 (0.042 to 0.077) | <0.001 | 0.086 (0.059 to 0.113) | <0.001 | 0.042 (0.014 to 0.070) | 0.003 |
| BMI-SDS at baseline | 0.676 (0.667 to 0.684) | <0.001 | −0.318 (−0.327 to −0.309)§ <0.001 | |||
| Calendar year at onset | −0.002 (−0.004 to 0.0001) | 0.068 | 0.0009 (−0.002 to 0.003) | 0.516 | −0.007 (−0.009 to −0.004) | <0.001 |
| Use of short-acting analogue | 0.030 (0.021 to 0.039) | <0.001 | 0.007 (−0.007 to 0.021) | 0.358 | 0.047 (0.032 to 0.061) | <0.001 |
| Use of long-acting analogue | −0.014 (−0.024 to −0.005) | 0.004 | −0.037 (−0.051 to −0.022) | <0.001 | 0.001 (−0.014 to 0.017) | 0.883 |
*Separate model for each effect adjusting for BMI-SDS at baseline, per unit increase for continuous variables.
†Independent effects in the model: age at onset, diabetes duration, sex, insulin regimen, insulin dose, BMI-SDS at baseline, calendar years at onset, use of short-acting analogue, use of long-acting analogue and interaction terms between sex and age at onset, diabetes duration, insulin dose, calendar year at onset and use of short-acting and long-acting analogue.
‡Estimated from least squares means.
§No significant difference between sexes.