Table 3.
Multivariate analysis using categorical testosterone levels within the PCOS cohort.
| Linear regression | β coefficient | Standard error | p value | N |
|
| ||||
| Mean ± SD 24 hr urine panel (range): | ||||
| Calcium (mg/day) | −32.4 | 38.8 | 0.406 | 65 |
| Citrate (mg/day) | −246.3 | 117.4 | 0.041 | 54 |
| Creatinine (g/day) | −0.1 | 0.1 | 0.691 | 56 |
| Oxalate (mg/day) | 1.8 | 4.8 | 0.705 | 58 |
| Sodium (mmol/day) | −36.7 | 18.9 | 0.058 | 52 |
| Uric acid (mg/day) | −28.4 | 57.2 | 0.621 | 60 |
| Volume (ml/day) | −449.5 | 238.7 | 0.066 | 53 |
| Mean ± SD stone composition % (range): | ||||
| CaOMH | 11.9 | 13.6 | 0.392 | 24 |
| CaODH | 3.6 | 10.9 | 0.743 | 24 |
| CaPh | 2.5 | 12.9 | 0.851 | 24 |
| CaHPh | −14.4 | 9.5 | 0.144 | 24 |
| MAP | −9.6 | 6.3 | 0.145 | 24 |
| UA | 1.8 | 2.4 | 0.481 | 24 |
|
| ||||
| Logistic regression | Odds ratio | 95% CI | p value | N |
|
| ||||
| No. abnormality (%): | ||||
| Hypercalciuria (>200 mg/day) | 1.0 | 0.32–3.33 | 0.955 | 65 |
| Hypocitraturia (<550 mg/day) | 0.2 | 0.04–0.94 | 0.042 | 54 |
| Hyperoxaluria (>40 mg/day) | 1.9 | 0.50–7.14 | 0.340 | 58 |
| Hypernatriuria (>150 mmol/day) | 13.3 | 1.49–100 | 0.021 | 52 |
| Hyperuricosuria (>750 mg/day) | 6250 | 4 × 1022–9 × 10−16 | 0.693 | 60 |
| Low volume (<2000 ml/day) | 0.2 | 0.04–1.19 | 0.079 | 53 |
| No. greater than 70% mixed (%)∗ | 1.0 | 0.11–9.09 | 0.978 | 24 |
Note: confounding variables adjusted for age, BMI, and metformin status with high or normal testosterone classification as an independent variable. Odds ratio is reported as the likelihood of a high testosterone PCOS patient having a given urine or stone analysis finding. Negative β coefficient indicates an elevation of that parameter in the high testosterone cohort. CI = confidence interval.