Advertisement
This ad is displayed using third party content and we do not control its accessibility features.
Close Banner
Advertisement
This ad is displayed using third party content and we do not control its accessibility features.

A 20,000-Person Study Found A Hidden Clue To Diabetes Risk

Zhané Slambee
Author:
October 20, 2026
Zhané Slambee
mindbodygreen editor
Longevity
Image by filadendron / iStock
October 20, 2026

You get your annual blood test results back, and everything looks fine. Your BMI is in the normal range, too, so diabetes risk probably isn't on your radar.

But weight alone can't tell you how well your body manages blood sugar or what your cholesterol levels look like.

In a recent study of nearly 20,000 adults1, researchers explored whether combining these blood measurements could help spot people at risk of prediabetes or type 2 diabetes, even if they weren't considered overweight.

About the study

Researchers followed nearly 20,000 adults in China, all aged 45 or older, who had a normal BMI and no prediabetes or diabetes at the start. Over an average of three years, they tracked who developed either condition.

They focused on the CHG index, a measurement that combines total cholesterol, fasting blood sugar, and HDL cholesterol (often called "good" cholesterol) into one number.

They wanted to see whether these three blood measurements together could offer clues about future diabetes risk that weight alone might miss.

This blood marker was linked to more than double the diabetes risk

People with higher CHG index values were more likely to develop prediabetes or type 2 diabetes during the study. Even after researchers accounted for other factors, the index was linked to more than double the risk.

Risk rose more sharply as the index increased through a certain range, then leveled off. So, the relationship wasn't as simple as higher numbers always meaning proportionally higher risk.

The study doesn't prove that a higher CHG index causes diabetes, but it suggests that looking at cholesterol and blood sugar together may reveal useful clues about metabolic health.

Why a normal BMI doesn't tell the whole story

BMI can be a useful screening tool, but it doesn't measure blood sugar or cholesterol, or distinguish between muscle and fat.

That means someone can have a normal BMI and still have metabolic risk factors worth discussing with their health care provider.

This doesn't make BMI useless. It just means body size is one part of a bigger picture.

Other research, including a study of nearly 480,000 adults, has also explored diabetes risk factors that aren't obvious from body size alone.

There are a few caveats. This was an observational study, so researchers couldn't establish cause and effect.

The participants were also adults aged 45 and older in China, meaning the findings may not apply equally to younger people or other populations. The CHG index isn't yet an established clinical screening tool.

Four ways to support your metabolic health

You don't need to calculate a new index to support your metabolic health. These everyday habits are a good place to start.

  • Keep up with checkups: Ask your health care provider what your blood sugar and cholesterol results mean for your individual risk, whatever your BMI.
  • Move regularly: Walking, cycling, and other physical activities help your body use glucose and support metabolic health.
  • Build balanced meals: Fiber-rich foods, protein, and unsaturated fats help support healthy blood sugar and cholesterol levels.
  • Strength train: Your muscles use glucose for energy, making resistance exercise a useful way to support blood sugar regulation.

The takeaway

The scale can't tell you everything that's happening under the hood.

Your blood work can fill in some of the blanks, so get to know your cholesterol and blood sugar numbers. The CHG index needs more research, but it's a good reminder that metabolic health is about more than your weight.

Advertisement
This ad is displayed using third party content and we do not control its accessibility features.