Ray, as you say it's dificult to draw any reliable conclusions of isolated dietary studies with small samples. Even your hypothetical example with twins is not infallible, since there can still be multiple differences between the twins and their lifestyles (i.e. one lives in a house with hardwood floor and windows that can be opened, with a friendly and accommodating wife, whereas the other lives in an apartment building with sealed windows and thick broadloom, and went through three unhappy marriages).
That's true, and that is why I wrote,
".... and persuade them to lead, as far as possible, an identical lifestyle,"For the experiment to be scientifically sound, one would need to recruit a large number of twins and conduct the experiment over a number of decades, perhaps 30 or 40 or even 50 years. If, during the course of the long experiment, it was observed that a few of the twin pairs diverged significantly in their lifestyles, one of them perhaps not physically exercising nearly as much as the other, or experiencing a lot more emotional stress than the other, then those twins would have to be excluded from the results.
However, when comparing larger studies, i.e. effect of dairy intake in Scandinavia or New Zeeland, it becomes quickly obvious that the harmful ingredients in milk have some effect on bone health.
Sweden - annual dairy consumption 355 kg and 802 bone fractures,
Denmark - 295 and 853,
Norway - 261 and 563,
Finland - 361 and 440,
New Zeeland - 110 and 288
Mexico - 155 and 169
I don't think the above study is useful without more information about the lifestyles of the people included in the results, and details about how the information was collected. For example, what does a consumption of 355 kg of dairy products include. If person 'A' drinks one litre of milk per day, and takes no other dairy products, that still amounts to a yearly consumption of approximately 365 kg of dairy products. If person 'B' eats 1 kg of cheese day every day, and no other dairy products, then that represents a lot more energy than a litre of milk, but would still be described as a yearly consumption of 365 kg of dairy products.
There is also a significant anomaly in the above results you've quoted. Sweden with an annual dairy consumption of 355 kg and 802 bone fractures, and Finland with a slightly higher consumption of dairy products, 361 kg, yet a very significantly lower number of bone fractures, only 440.
That seems to me a very clear example that other significant factors are involved, which are not addressed in the study. Perhaps the major factor not addressed is the overweight and obesity problem.
The following study shows there is a strong connection between obesity, or simply being overweight, and the need for hip or knee replacement.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2267551/If you were to check the overweight and obesity percentages in those countries listed in your study, which have low rates of hip replacement which seem to correlate with low consumption of dairy products, you'd probably find the obesity rates are significantly lower than they are in Western countries, although obesity seems to be on the rise everywhere.
I suspect that the real reason for the correlation between dairy consumption and bone fractures, is due to the
over-consumption of dairy products and the
over-consumption of meat and fat products. Too much of any type of food can be unhealthy. There have been instances of people even dying from drinking too much water in order to quench a serious thirst after running a marathon.
My personal view is that dairy products consumed in moderation are healthy. I always chose full-cream milk and prefer butter to margarine. I also eat natural and pure Yoghurt, unadulterated with sugar and other flavors, full-fat cream, sour cream, cheese, and coconut oil. I'm also as fit as a fiddle.