Point is that the paper where you post your picture about (
https://www.ncbi.nlm.nih.gov/pubmed/8942407), showed absolutely nothing
- a non-significant increase in total testosterone favoring the high fat group
- a non-significant increase in free testosterone favoring the low fat group
- NO correction for multiple comparisons problem (this is wildly problematic on top of the non-significant stuff)
The second paper (or the first of what you link here):
https://www.ncbi.nlm.nih.gov/pubmed/6538617 didn't only just decrease fat intake. It slightly decreased calorie intake too:
"The analysis of total energy consumption during these three periods revealed that the energy intake during intervention was approx. 15% lower than during baseline and switch-back periods. The chemical quantitation, however, showed that the method based on the Finnish food composition tables and a
computer program [21] overestimated about 10% of the total energy consumption in the basal and switchback diet but not that of the intervention diet. Thus, the possible slight change was not more than 5%. No differences were found in the quality of diets between the two methods. "
Indeed, they lost weight over the 6-week period:
"The monitoring of body weight revealed a small irreversible loss of weight (mean 1.1 kg) after transferring from the baseline to the intervention period."
(And yes, I'm aware there is a paper showing no difference in T with a 15% energy deficit, but it only lasted 2 weeks, which might have been too short to demonstrate an effect.)
And besides decreasing total fat intake, they also increased the intake of unsaturated fatty acids by 3-fold (see Table 1).
Finally, this was not a randomized trial.
The third paper:
https://www.ncbi.nlm.nih.gov/pubmed/15741266 is better, they randomized part of the subjects, I'll be focusing on that. Again, bodyweight decreased slightly (1 kg) over the course of 8 weeks.
What's interesting about the randomized part (see Table 3) is that switching from baseline to low fat decreased testosterone by 1.8 nmol/l, but then switching back to high fat didn't really change it at all (+0.2 nmol/l). What's more; free T didn't change.
Then if you look at the ones who went from baseline to high fat and then to low fat: those found no significant increase from baseline to high fat (+0.8 nmol/l), but did find a significant decrease then switching from high fat to low fat (-2.1 nmol/l). However, again, free T did not change significantly.
So eating a lot more fat (going from low fat to high fat) did NOT change testosterone (from 13.4 to 13.6 nmol/l, free unchanged 0.15 nmol/l). And going from baseline to high fat did not change total/free T either. So how does this paper support the claim "eat more fat for more testosterone"? The only thing they found was that going from high fat or baseline to low fat decreased total T (with unchanged free T). So this study did not demonstrate 'eat more fat for higher testosterone'. At best it demonstrated; eat less fat for lower testosterone. For which it still leaves open that it could be due to the high fiber, but I suppose you could argue to some extent that that's pragmatic about it. If you switch from high-fat to low-fat you'll likely end up eating more fiber.
And then on top of that, you now claim: "I think this is enough evidence to point people into the direction of higher fat intakes for optimal hormones."
Can you point me out on which clinical endpoints you base that this would be for "optimal hormones"?
In regard to body composition, one of your papers favors the low fat group:
"The BMI also decreased from 27.8 0.6 to 26.5 1.1 kg/m2 after low-fat diet (P 0.0033). The body composition parameters measured by dual-energy x-ray absorptiometry (Fig. 2) showed a significant decrease in body mass (P < 0.016) that was due mainly to the decrease in body fat (P < 0.05)."