Table 2.

Association of depression history and severity characteristics with low levels of anti-müllerian hormone (≤1.4 ng/mL), Harvard Study of Moods and Cycles, 1995–1998

	AMH Level		Age-adjusted	Multivariable Ɨ
	≤1.4 ng/mL
	n	%	PR	Adjusted PR (95% CI)
Model 1
History of depressiona
No	132	57.9	1.00	1.00	Reference
Yes	81	57.5	0.92	0.90	(0.75, 1.08)
Model 2
Antidepressant use
Depressed, never used antidepressants	31	58.5	0.93	0.91	(0.71, 1.15)
Depressed, ever used antidepressants	50	56.8	0.92	0.89	(0.72, 1.11)
Model 3
Depression & anxiety
Depression only	41	56.2	0.91	0.89	(0.71, 1.12)
Depression and anxiety	40	58.8	0.94	0.91	(0.73, 1.13)
Model 4
Hamilton Depression Rating Scale (HAM-D)b
History of depression, HAM-D < 8	19	54.3	0.86	0.81	(0.59, 1.11)
History of depression, HAM-D ≥ 8	46	58.2	0.94	0.93	(0.75, 1.15)
Model 5
Age of depression onset, yearsc
< 25	42	59.2	1.03	0.97	(0.78, 1.20)
25–29	14	58.3	0.95	0.90	(0.64, 1.27)
≥ 30	22	53.7	0.83	0.81	(0.61, 1.09)
Model 6
Number of depressive episodesc
1–2	46	56.8	0.89	0.87	(0.71, 1.07)
≥ 3	32	58.2	0.98	0.97	(0.76, 1.24)
Model 7
Total duration of depressive episodes, monthsc
< 12	38	52.8	0.86	0.85	(0.68, 1.07)
12–35	16	61.5	1.01	1.00	(0.74, 1.36)
≥ 36	23	62.2	0.97	0.94	(0.72, 1.24)

Ɨ Models adjusted for age, age at menarche, cycle irregularity, parity, oral contraceptive use, smoking status at baseline, smoking pack-years, and body mass index (BMI). Analyses were restricted to those with AMH < 5.0 ng/mL, a potential indicator of elevated AMH due to polycystic ovary syndrome (PCOS)

a. Non-depressed serve as reference group for all exposure variables (displayed in first row only)

b. Hamilton Depression scores were missing for n = 27 (apx. 7% from both depressed and non-depressed cohort)

c. ‘Age of onset’ variable missing 5 responses; ‘number of episodes’ and ‘duration’ variables missing 6 responses