The key insight is twofold: first confirm that this is a treatable illness, then decide how to pick the drug.
A woman with low mood, anhedonia, poor appetite and insomnia lasting a full year clearly has a major depressive episode. The fact that a business loss triggered it is irrelevant to whether she needs treatment; a sustained one-year episode is not normal bereavement and must be treated, which immediately rules out the do-nothing option.
Now consider drug selection. The various antidepressant classes are roughly equal in overall efficacy, so no class can be crowned the most efficacious for everyone; that disqualifies the blanket SSRI claim. Starting two antidepressants together is not how a first episode is managed and only multiplies adverse effects, so combination therapy is wrong at the outset.
Since efficacy is comparable across agents, the rational basis for choosing one is its side-effect and safety profile matched to the patient. Therefore antidepressant therapy should be initiated according to the side-effect profile.
\[\boxed{\text{Start an antidepressant based on the side-effect profile}}\]