Causal inference is unfortunately not that simple. If you find differences in cancer rates, it could be due to alcohol consumption or any number of other factors that differ between Islamic societies and other societies. You can try to control those differences statistically but since the set of possible differences is pretty much open-ended, it’s incredibly difficult to say the least.