How Does the Carbohydrate-Insulin Model aka TWICHOO, Explain the Obesity Epidemic?

A short post ... something I hit on again in the process of writing another one. Below is a plot of the absolute amounts of protein, carb and fat that Americans claimed to consume from 1965 through 2011 (adapted from Hite & Economic Litigation Consultants, 2015 , blog post).   All but 1965 is NHANES data, but I'll even "give them" the inappropriate toss-in of the 1965 Sesame-Street-song data point.

I've properly lined up the timelines for prevalence of BMI > 25 curve  (that scale -- not shown -- goes from just over 45% to just under 75%. ).  This is the same curve repeated by Adele Hite and so, so many others -- often immediately after chastising Ancel Keys for equating correlation with causation (keep ignoring thqt he didn't do that with his infamous plot) -- in an attempt to implicate the Dietary Guidelines in the obesity epidemic.    Even IF one could pin the obesity epidemic on an increase in carbohydrate consumption, two things are clear:
  • Absolute fat intake stayed consistent with baseline
  • Calories increased 

Q:  How does TWICHOO  explain this?

                   (Taubes Wrong Insulin-Carbohydrate Hypothesis Of Obesity)

A:   It doesn't!                                                                          

The only time ANY effect at all has been seen for LC vs. LF, it has been when carb restriction is extreme, protein is not controlled for, calories are not controlled for, the low fat is of negligible difference from the baseline diet, LC is whole real food vs. processed carbs and sugar for the LF, or some combination of several-to-all of these confounders.  Even with the deck stacked, any effect seen is relatively short lived and can't be attributed to increased satiety of fat or clinically significant reduction in insulin levels.  

If you take that together with data we're supposed to accept at face value -- NHANES -- there can be no support for an increase in carbohydrate from around 225 to 275 g/day resulting in hyperinsulinemia induced fat cells gone wild.   It doesn't square.  

Why is anyone spending another research dollar to test this idea that doesn't even fit the observations?