Supplements: doing me any good?

Are the supplements I'm taking really doing me any good?  It's possible the positive effects I'm experiencing are just placebo, or follow from my healthier diet, or are due entirely to my exercise regime.  Finding out if they are worth my money and time will involve an experiment. 

Ideally I'd have a window into those alternate universes where I'm doing everything the same, except taking supplements.  Lacking that window makes the determination of their effect tricky.  Once I've hit my targets I could stop taking them, to see if anything changes; but that won't tell me if they're doing me any good right now.  From the perspective of my personal experience I have no means of knowing whether the supplements I'm taking are helping, hindering, or having no effect in the pursuit of my targets.

So I have no choice: I must trust something other than my personal experience. The options I can think of are:

• scientific literature on the topic
• respectable science reportage (e.g. New Scientist)
• claims made by legally regulated suppliers / manufacturers
• claims made by supposed users of specific products
• anecdotes from friends and colleagues

(As you can probably guess, that list is in my order of preference.)

None of these sources are entirely satisfactory.  The raw literature, when it is publically available at all, is generally written by professionals for fellow professionals: i.e. unreadable by us laypeople.  Hence New Scientist and similar publications fullfill a vital role; but they are subject to the journalistic paradox; i.e. the most important news is rarely the most interesting.  Suppliers and manufacturers of supplements have to be considered biased (and the more hyped the product, the more cynical I tend to be).  Reviews by users are seductive, but caution: are they what they seem to be?  Anecdotes from friends and colleagues are probably more reliable, if only because you have some knowledge of the opinionator. 

In practice, I'm bumbling along using as many of these sources as happen to be easily available at the point of decision.  It's a far from perfect approach...