Trials of the same nutrient frequently reach opposite conclusions. A large part of that inconsistency comes from a single design question, which is whether the participants were short of the nutrient to begin with.
The dose-response curve is not a straight line
For most nutrients, the relationship between intake and function rises steeply from a deficient state and then flattens once requirements are met.
On the flat part of the curve, additional intake changes little, because the processes that use the nutrient are already operating at capacity.
The same amount of a supplement therefore produces a large effect in one person and none in another, without either result being wrong.
Trial populations determine the answer
A trial recruiting from a well-nourished population is mostly enrolling people on the flat part of the curve, and will tend to find no effect.
A trial in a population with widespread shortfall is enrolling people on the steep part, and is far more likely to find one.
Both results can be reported as findings about the nutrient in general, which is how the appearance of contradiction arises from consistent underlying biology.
Deficiency is defined by consequence, not by intake
Reference intakes are population-level figures set to cover the requirements of nearly everyone, which means most individuals need less than the stated number.
Clinical deficiency is defined by a measurable consequence, whether a blood marker outside range or a functional impairment attributable to the shortfall.
An intake below the reference figure is therefore not the same as a deficiency, though the two are routinely conflated in marketing.
More is not simply safer
Several nutrients have a defined upper level above which harm becomes more likely, and for fat-soluble vitamins the margin between requirement and excess is narrower than for water-soluble ones.
Some effects only appear over long periods, which makes them hard to attribute and easy to miss in short trials.
Interactions with medication and with other nutrients add further complexity, since some minerals compete for the same absorption pathways.
Why testing precedes the question of supplementation
Because the effect depends on starting status, the informative step is establishing status rather than reasoning from symptoms, which are non-specific.
Certain groups have well-recognised elevated requirements or absorption difficulties, and those are identified clinically rather than through self-assessment.
Decisions about testing, and about what to do with a result, belong with a clinician, since both the thresholds and the appropriate response vary by circumstance.