Insulin Resistance & Alzheimer's Disease

Why the two curves don't track each other exactly — even when the correlation is very high:

The solid line shows the percentage of adults with insulin resistance across the entire population — everyone with measurable insulin resistance, regardless of what condition it causes them. Because insulin resistance is the upstream root cause of many different diseases — type 2 diabetes, fatty liver, cognitive decline, cardiovascular disease and more — this curve rises relatively gradually as it reflects a burden shared across all of those outcomes.

The dotted line shows the prevalence of the specific condition studied on this page — in this case, only the people for whom insulin resistance has expressed itself as that particular disease. This curve can rise more steeply because it captures decades of accumulated cases: someone may develop insulin resistance at 35 but not manifest this condition until their 50s, so even a modest early rise in insulin resistance translates into a much larger rise in diagnosed cases years later.

The r value (e.g. r = 0.97) is a correlation coefficient. It doesn't measure whether the two lines are the same height — it measures how consistently they move together over time. An r of 0.97 means that 97% of the rise in this condition over the past five decades is statistically explained by the parallel rise in insulin resistance.

What the r value tells you:
0.50–0.70 — Modest connection. The two trends are related but other factors are involved.
0.70–0.90 — Strong connection. Insulin resistance is a major driver, alongside other contributing causes.
0.90 and above — Dominant connection. Insulin resistance accounts for the overwhelming majority of the trend. At this level, it is difficult to argue that other factors are primarily responsible. The values seen across these studies — consistently 0.90 to 0.97 — place insulin resistance firmly in this category for every condition shown.