. 2019 Mar 5;321(9):858–868. doi: 10.1001/jama.2019.0556

Table 2. Effect of Supplements and Food-Related Behavioral Activation Therapy on Major Depressive Disorder Onset.

	No. of Patients	No. of Events	No. of Follow-up Dropouts^a	Imputed OR (95% CI)^b	P Value^b
Primary Outcome: 12-mo MDD Onset
Placebo^c	513	51	113	1 [Reference]
Supplements^d	512	54	126	1.06 (0.87-1.29)	.57
Food-related behavioral activation therapy
No^e	513	57	122	1 [Reference]
Yes^f	512	48	117	0.93 (0.76-1.13)	.47
Supplements by therapy^g	NA	NA	NA	0.93 (0.76-1.14)	.48
Post Hoc Outcome: Time to MDD Onset
	Time to Event or Censoring (Person-Months)			HR (95% CI)
Placebo	4656	51	113	1 [Reference]
Supplements	4488	54	126	1.05 (0.86-1.27)	.65
Food-related behavioral activation therapy
No	4503	57	122	1 [Reference]
Yes	4641	48	117	0.91 (0.75-1.10)	.32
Supplements by therapy^g	NA	NA	NA	0.91 (0.75-1.11)	.36

Abbreviations: HR, hazard ratio; MDD, major depressive disorder; NA, not applicable; OR, odds ratio.

^{^a}

No. that did not complete 12 mo follow-up and had no MDD at previous follow-up measurements.

^{^b}

Derived from multiple imputed data. Models were adjusted for interventions, site and history of MDD. Food-related behavioral activation therapy and pills were effect-coded (−1,1).

^{^c}

Twenty-five of 257 participants (9.7%) in the placebo without therapy group and 26 of 256 (10.2%) in the placebo group with therapy.

^{^d}

Thirty-two of 256 participants (12.5%) in the supplements without therapy group and 22 of 256 (8.6%) in the supplements without therapy group.

^{^e}

Twenty-five of 257 participants (9.7%) in the placebo group without therapy and 32 of 256 (12.5%) in the supplements without therapy group.

^{^f}

Twenty-six of 256 participants (10.2%) in the placebo with therapy group and 22 of 256 (8.6%) in the supplements with therapy group.

^{^g}

Interaction between therapy and supplements.