To be a scientific term, a term must have the same meaning every time it is used. Is causation a scientific term? No – in actual usage, it does not have the same meaning. There is no definition that fits the following two statements:
- The motion of the eight ball was caused by a collision with the cue ball (almost true).
- Smoking three or more packs of cigarettes per day causes lung cancer (nonsense, there is a contingency table of possible outcomes, and 20% of all lung cancers occur in non-smokers).
But here is a scientific definition, illustrated with two related phenomena, hyperuricemia and gout.
By definition, hyperuricemia causes gout if there is SIN:
- S = sufficiency. No one with hyperuricemia does not have gout. Hyperuricemia is sufficient to produce gout all by itself. AND:
- I = immediacy. The instant that uric acid concentration exceeds the solubility limit, it precipitates and causes gout. AND:
- N = necessity. There is no route to gout that does not precede through the cause, hyperuricemia.
Looking at the data from NHANES 2007-2008 (Arthritis Rheum. 2011 Oct;63(10):3136-41. doi: 10.1002/art.30520. PMID: 21800283):
- S = sufficiency is not met: 21.2% of American men have hyperuricemia (> 7mg/dL), while 5.9% have been diagnosed with gout. The correlation is strong, but it is not causal.
- I = immediacy is not met by the same data. If it were, there would be 21.2% gout among American men.
- N = necessity is not met because roughly a quarter of people with acute gout do not have hyperuricemia. Ref: Schlesinger N et al. Serum urate during acute gout. J Rheumatol 2009 Jun; 36:1287. “Of 339 patients who presented with acute gout, 14% had serum uric acid levels ≤6 mg/dL, and 18% had levels between 6 mg/dL and 8 mg/dL. Mean uric acid level was about 8.3 mg/dL (7.2 mg/dL in patients taking allopurinol, and 8.5 mg/dL in patients not taking allopurinol).” – as summarized by Dr. Allan Brett, https://www.jwatch.org/jw200906300000002/2009/06/30/uric-acid-levels-during-acute-gout-attacks). Sorry, this reference takes 8 mg/dL as the definition of hyperuricemia. To be consistent, as an estimate, let us divide by two the 18% with serum uric acid in the 6-8 mg/dL range. This suggests that roughly 23% of all acute gouty attacks occur in people with serum uric acid in the normal range. In the discrepant cases, did the uric acid spike just before the attack and then subside? I don’t know. But even if it did:
There is no SIN. There is definitely no S, definitely no I, and apparently no N – there is no causation if one of the three is missing.
There is a contingency table of outcomes, gout vs no gout, hyperuricemia vs normouricemia. There is no contingency table describing the outcome of a collision between a cue ball and an eight ball. This physics experiment nearly has all three aspects of causation nailed down. It almost has SIN. Hyperuricemia does not have SIN in the etiology of gout. Biology is a bit more complicated than physics, as in every biological process, many factors oppose many other factors. In biology, outcomes are always a vector sum of the opposing forces and are described by contingency tables. Such is life.
By definition, any phenomenon with a contingency table of outcomes is not causal. All by themselves (sufficiency) causes necessarily (necessity) and immediately (immediacy) produce their effects.