Alternative treatments for any diagnosis requires knowing "alternative to what?" I assume you mean an alternative to medication or do you mean alternative like "far out there". Alternative methods, no matter how you defined it, probably worked for some people, but the science of psychology and psychiatry require statistical significance. According to some published research, even the standard treatments cannot claim statistical significance. For example, if you followed many research reports it would not make sense to take any other medication than Prozac for depression, but from a practical practice point of view, that would be false. Many people cannot take Prozac and I have seen many people who swear by the other medications. Everyone is different and human, not a statistic.
Could aroma therapy work? Sure. The olfactory lobes have more direct connection to the cerebrum and limbic system than other sensory organs, but music therapy also works for some people, as does visual meditative techniques. But no method works for everyone.
Statistically, the therapeutic techniques with the most success are Cognitive Behavior Therapy (CBT) techniques, and their evolutionary cousin, Action & Commitment Therapy (ACT). ACT and Dialectical Behavior Therapy use mindfulness techniques which ultimately put treatment into the person's hands to use in life. Sometimes these techniques take a long time to learn and practice. You can never eliminate the thoughts or events that produced your depression, or any psychological problem, because they are naturally occurring events in people's lives. It is our response to these events that leads to problems and further problematic stimuli. We have to learn to cope with these events in more effective ways. Iknow it sounds simple, and I know it is difficult to do. We have have ways of managing situations that we would like to change, myself included.
If you want to know more about procedures for alternative methods, I am sure you can find them on the internet. Look up meditation and self-hypnosis.