Table 4.
Univariate and multivariate binary logistic regression to predict adult subfertility by adolescent menstrual cycle pattern classified at least 2 years after menarche.
| Adolescent oligomenorrhea | ||||||
|---|---|---|---|---|---|---|
| Univariate analysis | B | S.E. | Wald | Odds Ratio | 95% CI | P-value |
| Gynaecological age* | 0.58 | 0.35 | 2.76 | 1.78 | 0.90–3.51 | 0.10 |
| BMI | 0.36 | 0.14 | 6.12 | 1.43 | 1.01–1.89 | 0.01 |
| Waist-Hip ratio | 3.51 | 4.42 | 0.63 | 33 | 0.00–189 | 0.43 |
| mF&G score | −1.19 | 0.22 | 0.73 | 0.83 | 0.54–1.27 | 0.39 |
| Acne | −1.55 | 0.763 | 4.10 | 0.21 | 0.05–0.95 | 0.04 |
| LH | 0.195 | 0.095 | 4.22 | 1.21 | 1.00–1.46 | 0.04 |
| Testosterone | −2.69 | 0.892 | 0.91 | 0.76 | 0.13–4.39 | 0.76 |
| Androstostenedione | 0.35 | 0.117 | 3.82 | 1.41 | 1.00–2.00 | 0.05 |
| DHEAS | −0.10 | 0.116 | 0.008 | 0.99 | 0.79–1.24 | 0.93 |
| AMH | −0.003 | 0.68 | 0.001 | 1.003 | 0.79–1.27 | 0.98 |
| Insulin/glucose ratio | −0.35 | 0.28 | 1.53 | 0.71 | 0.41–1.22 | 0.22 |
| PCOM | 0.86 | 0.28 | 1.21 | 2.36 | 0.51–10.9 | 0.27 |
| Multivariate analysis | ||||||
| BMI—adjusted for Gynaecological age, Acne, LH, and Androstenedione | ||||||
| 0.32 | 0.16 | 3.86 | 1.38 | 1.00–1.89 | 0.05 | |
| BMI—adjusted for Gynaecological age and smoking | ||||||
| 0.44 | 0.17 | 6.82 | 1.56 | 1.12–2.17 | 0.009 | |
|
| ||||||
| Adolescent regular menstrual cycles | ||||||
|
| ||||||
| Univariate analysis | B | S.E. | Wald | Odds Ratio | 95% CI | P-value |
|
| ||||||
| Testosterone | 0.90 | 0.48 | 3.49 | 2.47 | 0.96–6.35 | 0.06 |
| BMI | 0.48 | 0.11 | 0.19 | 1.05 | 0.84–1.31 | 0.66 |
Other physical characteristics, endocrine, or ovarian ultrasound parameters did not contribute to improve the prediction for subfertility. Significant outcomes (P < 0.05) are presented in bold.
*
Gynaecological age (age in months minus age in months at menarche).
mF&G, modified Ferriman and Gallwey score; DHEAS, dehydroepiandrosterone; AMH, anti-Mullerian hormone.