In regard to practical interventions, concurrent glycogen loading has been noted to increase creatine stores by 37-46% regardless of whether the tissue was exercised prior to loading phase. It is important to note, however, that creatine levels in response to the creatine loading protocol were compared in one glycogen-depleted leg to the contralateral control leg, which was not exercised. This does not rule out a possible systemic exercise-driven increase in creatine uptake, and the increase in creatine noted above was larger than typically seen with a loading protocol (usually in the 20-25% range). Consistent with an exercise-effect, others have reported that exercise itself increases creatine uptake into muscle, reporting 68% greater creatine uptake in an exercised limb, relative to 14% without exercise.
Even human studies can be misinterpreted or manipulated. A few years ago a then-esoteric trace mineral called boron was found to increase testosterone, but only in older women, as it turned out. When boron was tested in young men engaged in weight training, it proved worthless for increasing testosterone. That doesn't mean boron is useless; it helps the body use the minerals calcium and magnesium, and it appears to increase mental alertness.
It is known that intracellular energy depletion (assessed by a depletion of ATP) stimulates AMPK activity in order to normalize the AMP:ATP ratio, and when activated AMPK (active in states of low cellular energy and colocalizes with creatine kinase in muscle tissue) appears to inhibit creatine kinase via phosphorylation (preserving phosphocreatine stores but attenuating the rate that creatine buffers ATP). While phosphocreatine technically inhibits AMPK, this does not occur in the presence of creatine at a 2:1 ratio. It seems that if the ratio of phosphocreatine:creatine increases (indicative of excess cellular energy status) that AMPK activity is then attenuated, since when a cell is in a high energy status, there is less AMP to directly activate AMPK.