Table 4.
Effect of potentially confounding parameters (a) and stimulation parameters (b) on changes in nonfasted blood glucose concentrations (mg/dL) from baseline to follow‐up in patients with VNS.
| (a) Identification of potentially confounding parameters | |||
|---|---|---|---|
| Parameter | Mean ± SEM | Effect Size | Significance |
| Gender (♂ = 1, ♀ = −1) | 0.05 ± 0.24 | N/A | n.s. |
| Glucose at baseline (mg/dL) | 95.2 ± 4.8 | −0.69 | P < 0.001 |
| Age at implantation (years) | 39.9 ± 2.7 | +1.17 | P < 0.001 |
| Duration of VNS (years) | 4.4 ± 0.5 | N/A | n.s. |
| BMI at follow‐up (kg/m2) | 28.4 ± 1.6 | N/A | n.s. |
| Intercept | N/A | +24.44 | n.s. |
| (b) Effect of stimulation parameters on changes in nonfasted blood glucose concentrations (mg/dL) from baseline to follow‐up, adjusted for glucose concentrations at baseline and age at implantation. | |||
|---|---|---|---|
| Parameter | Mean ± SEM | Effect size | Significance |
| Frequency (Hz) | 23.7 ± 1.1 | N/A | n.s. |
| Current (mA) | 1.9 ± 0.2 | N/A | n.s. |
| On time (s) | 25.8 ± 2.3 | +0.92 | P < 0.01 |
| Off time (s) | 96.6 ± 17.6 | ‐0.076 | P = 0.06 |
| Intercept | N/A | −11.47 | P = 0.09 |
N/A, not applicable; n.s., not significant.
(a) The multiple linear regression analysis is based on 19 subjects. The effects of gender, duration of VNS, and BMI (body mass index) at follow‐up were not significant and were removed from the model. Model statistics: multiple R 2: 0.80; adjusted R 2: 0.77; F: 31.43 on 2 and 16 degrees of freedom; P < 0.001.
(b) The multiple linear regression analysis is based on 19 subjects. The effects of the stimulation frequency and stimulation current were not significant and were removed from the model. Model statistics: multiple R 2: 0.38; adjusted R 2: 0.30; F: 4.94 on 2 and 16 degrees of freedom; P < 0.05.