Synergy vs Additivity
Additivity is the combined effect two agents are expected to produce from their individual potencies alone, while synergy is a combined effect that exceeds that prediction under a stated reference model.
Synergy is definable only against a null model of no interaction, and there is more than one. The Loewe additivity model treats the agents as dilutions of each other and predicts the combination from dose equivalence, tested with an isobologram or combination index. The Bliss independence model assumes independent mechanisms and multiplies fractional effects. A combination can exceed one reference and not the other, so a synergy claim without a named model is not interpretable.
Establishing synergy needs a dose-response surface, not a single combination arm: components must be tested across a range of doses and ratios so the observed effect can be compared with the prediction at matched effective doses. Dual incretin agonism is the case in point. Tirzepatide activates both GIP and GLP-1 receptors and, in the SURPASS-2 head-to-head trial, produced greater HbA1c and weight reductions than semaglutide 1 mg — superiority over a comparator, not a formal demonstration that the two activities interact supra-additively.
The distinction changes what a combination is worth. Genuine synergy means a lower dose of each component reaches the same effect, which can widen a therapeutic window by keeping each below its own toxicity threshold. Additivity means the combination buys nothing a higher dose of the more tolerable agent would not.
The near-universal error is inferring synergy because a combination outperformed each single agent. Two drugs given below their individual ceilings will nearly always do that, since both curves still have room to rise; that result is the definition of additivity, not evidence against it. Stacking anecdotes make the same leap without even a controlled comparison.