Is it even posible? Not really according to his paper

This study is really interesting. As you see, “In 457 (96.2%) displaced intertrochanteric fractures and in 221 (93.2%) displaced subcapital fractures, radiological shadow of the penis was turned to the fractured side (c2 = 17.1, P<0.01). ” So statistically speaking,  the penis is located at the fractured side during x-ray.

I am so glad the author did not concule that “the weight of the penis lead to the fracture”. If you read some of the omic studies, you see that many authers will see the side effect as cause.

So, when we are doing statistiscally analysis on data that show good correlation, you are just observing 2 events occur concurrently with high probability. It have nothing to do with the cause and effect.