# 33. Statistical manipulation - False causality ## 33.5. Methodology/Refinements/Sub-species ### 33.5.1. Regression fallacy The regression fallacy is a failure to prove a causal relationship despite a strong correlation in a regression analysis. The regression fallacy is a failure to take into account natural fluctuations in data when ascribing causes and it is also cause by overlooking other variables which may really have true causal links to otherwise unrelated data. This phenomenon can be used to deliver a manipulative payload by attributing correlation between two sets of data, without proving any causal relationship. The world is full of obscure and irrelevant correlations, but to prove real cause and effect is a different matter. #### 33.5.1.1. Quack medical examples Many people are led to believe in the causal effectiveness of worthless remedies, simply because of the regression fallacy. The intensity and duration of pain from arthritis, chronic backache, gout, etc., fluctuates. A remedy such as a chiropractic spinal treatment is usually sought when the pain is at its worst. The pain in most cases would begin to lessen after it has peaked. It is easy to deceive ourselves into thinking that the remedy we sought caused our reduction in pain. It is because of the ease with which we can deceive ourselves about causality in such matters that scientists do controlled experiments to test these kinds of causal claims. #### 33.5.1.2. Big Pharma examples A complex area of investigation at the moment is the use of drugs designed for one treatment and being used for an entirely different medical purpose. For the large pharmaceutical companies, this concept is a very attractive business opportunity because the costs of testing and licensing an existing drug for use in a different application are much, much lower than introducing a completely new drug. A simple instance is the use of aspirin in the control of heart disease, but there are many more complex cases, like the use of Statin drugs in the possible control of various cancers. The practice of prescribing a drug normally used for one illness for a completely different ailment is called "Off Label Use", because its use is not described in the drugs specification on the label.33.1 For the pharmaceutical companies the practice of "off label use" is highly desirable because it opens up huge alternative markets for uses of their existing regulated drugs at very low cost. The use of false causality as a manipulative method is very tempting in these circumstances. Correlations between a particular drug use and alternative medical benefits are being detected in patients that already have particular - pre-existing - conditions, such as high levels of cholesterol. The use of cholesterol-lowering drugs and a correlation with a lower incidence of colon cancer does not necessary prove a link between the drug and the benefit. It may well be that users of the Statin drugs are also receiving other related treatments which are actually the cause of the reduction in cancer incidence. It may well be that the lower levels of bad cholesterols is symptomatic of lower cancer propensities etc. Extreme caution is required in responding to any correlative claims which cannot demonstrate causality.