Call it regulatory capture.
There is a clear (admitted) bias in Western nutrition science. Willet et al. are known for P hacking and curve fitting the same datasets over and over to find ideologically supportive correlations. This one only has an HR of 1.7 even after the data is parsed supportively.
I agree that science and academic/intellectual circles are sometimes biased due to other influences and economic incentives. A quick search on the said Dr revealed he is quite active in bashing some studies while promoting his own ideas (biases).