Table 3.
Unadjusted and adjusted odds ratio of significant variables explaining lipohypertrophy.
| Variables | Univariate binary logistic regression | Multivariate logistic regression | |||
|---|---|---|---|---|---|
| Unadjusted odds ratio (95% CI) | Significance | Adjusted odds ratio (95% CI) | Significance | ||
| BMI (kg/m2) (Reference-obese >27.5) | Underweight | 12.10 (3.03–48.56) | <0.0005 | 13.01 (2.25–75.24) | 0.004 |
| Normal | 7.56 (2.10–27.21) | 0.002 | 11.76 (2.52–54.79) | 0.002 | |
| Overweight | 5.02 (1.24–20.30) | 0.024 | 7.25 (1.27–41.38) | 0.026 | |
| Injection area size (Reference-post card) | Credit card | 37.39 (17.02–82.16) | <0.0005 | 23.18 (9.07–59.24) | <0.0005 |
| Playing card | 3.25 (1.53–6.88) | 0.002 | 2.30 (0.97–5.47) | 0.059 | |
| Rotation (Reference-followers) | Non-followers | 14.28 (8.56–23.81) | <0.0005 | 6.31 (3.35–11.88) | <0.0005 |
| Injection regimen (Reference-rapid and long acting analogs) | Regular plus Long | 1.91 (1.14–3.18) | 0.013 | 3.17 (1.49–6.78) | 0.003 |
| Regular plus NPH | 2.77 (1.13–6.81) | 0.026 | 4.57 (1.35–15.54) | 0.015 | |
| Conventional premixed | 6.31 (2.57–15.53) | <0.0005 | 4.62 (1.36–15.73) | 0.014 | |
| Needle syringe length (Reference-4 mm) | 5 mm | 0.44 (0.17–1.19) | 0.108 | Dropped (0.096)a | |
| 6 mm | 1.63 (1.00–2.66) | 0.049 | |||
| 8 mm | 1.98 (0.75–5.22) | 0.166 | |||
| Injection device (Reference-pen) | Syringe | 2.62 (1.45–4.73) | 0.001 | Dropped (0.200)a | |
| Pen and syringe combined | 2.76 (0.99–7.66) | 0.051 | |||
| Total daily dose (units) | 0.98 (0.97–0.99) | 0.001 | Dropped (0.224)a | ||
| Gender (Reference-female) | Male | 1.77 (1.16–2.70) | 0.008 | Dropped (0.326)a | |
| HbA1c (%) | 1.13 (1.03–1.23) | 0.007 | Dropped (0.938)a | ||
a
These variables became insignificant after adjusting for other variables and was dropped in sequential order to obtain the final model.