This is actually easily explainable and it isn't the "dunk" or whatever some people think it is.
At the most fundamental level, linoleic acid (LA) concentration and percentage are separate biological variables.
You can see this with a simple hypothetical: the first person is metabolically healthy while the second person isn't.
Healthy person:
Total fatty acids = 4 mmol/L
LA = 1.2 mmol/L
LA = 30% of total
Unhealthy person:
Total fatty acids = 8 mmol/L
LA = 1.6 mmol/L
LA = 20% of total
The LA concertation increased by 33% in the unhealthy person, but the percentage decreased from 30% to 20%.
And this would result in the exact thing you see among these graphs, where higher metabolic risk correlates with both higher LA concentrations and lower LA percentage.
This is almost certainly what was happing in the NHANES cohort from this study because total circulating fatty acid concentrations were positively correlated to triglyercides, LDL, cholesterol, insulin, glucose, blood pressure, BMI, and waist circumference.
And LA increased as total fatty acids increased, but not nearly as quickly, meaning that percentage of LA fell while concentration of LA increased as someone became more metabolically unhealthy.
Basically, LA concentration may simply be inheriting the association between total circulating fatty acids and metabolic dysfunction.
Just take a step back here and imagine someone with insulin resistance. These people have a larger circulating pool of fatty acids overall, including LA. So of course LA concentration tracks with insulin, just like nearly every other fatty acid.
So, if anything, these results aren't evidence that LA concentration is itself harmful.
HUGE
#seedoil News: Science studies typically only use fatty acid percentage of the whole but when actual concentrations are measured, the associations to disease outcomes flip, especially for omega-6 fatty acids found in seed oils! Authors include anti seed oil investigators Chilton and Brenna
The Influence of Fatty Acid Expression Methods on Associations with Cardiometabolic Biomarkers: Evidence from NHANES 2011–2012
@trikomes @TuckerGoodrich @SeedOilDsrspctr @SBakerMD @drcateshanahan @ChrisMasterjohn