I often mention ‘sexy’ science topics in my blog, because these are the issues that make it into the popular press. Many of these examples of late are focused on connections between commonly used medications and dementia.
In the studies, there is a correlation between increased use of the offending medication and development of dementia and this implies that the medication must somehow cause the dementia, but correlation is not the same as causation. There is a tongue-in-cheek website that makes this point. In it, all sorts of statistics are correlated, but there is no reason to think there is a connection between them. Take a look >>>HERE<<<
That patients with dementia may use sleeping pills more because they have difficulty with sleep more than the average person, or perhaps they have skin conditions that cause itching and this is why they use anti-histamines more often. These are alternative explanations for the correlation, but they do not imply causation.
Making a causal connection is much more difficult and requires a placebo controlled double-blind study to eliminate statistical variations and to prospectively prove the desired connection. Although we’d like to believe these correlations, we must remember how weak these connections can be and avoid our desire to go with biases that we have; we need to be more objective and open our minds to possibilities that refute your prejudices as opposed to reinforcing them.