Five markers that can alert us to diabetes risk before glucose levels are elevated

And I begin with something important: none of these values, in isolation, can reliably predict who will develop diabetes over the next ten years.

What we do is interpret the whole picture: medical history, body composition, waist circumference, blood pressure, diet, physical activity and various biomarkers.

Because in preventive medicine we do not want to wait for disease to appear. We want to identify an unfavourable trajectory while we still have the opportunity to change it.

1. Fasting insulin: when the pancreas is working harder than it seems

Let us imagine two people with a blood glucose level of 90 mg/dL. On paper, both have the same glucose. But one may need a considerably larger amount of insulin to keep it at that level.

Why? Because when tissues begin to respond less effectively to insulin, the pancreas can compensate by increasing its secretion. This is what we call compensatory hyperinsulinaemia.

For a period of time, this compensation can keep glucose apparently normal. That is why, in certain patients, knowing both fasting glucose and fasting insulin simultaneously can give us additional metabolic information.

But insulin should not be interpreted in isolation either: there is variability between laboratory methods and there is no single universal cut-off point applicable to everyone.

2. HOMA-IR: approximating insulin resistance

When we have both fasting glucose and fasting insulin we can calculate HOMA-IR, an index used to estimate insulin resistance.

And here I want to correct an idea that circulates widely on social media: HOMA-IR does not directly measure insulin resistance. It estimates it.

Furthermore, there is no universal value of 1.5 above which we can automatically state that a person has a problem. Cut-off points can vary depending on population, age, methodology and clinical context.

So why does it interest me? Because, correctly interpreted and within a comprehensive metabolic assessment, it can help us detect that the body is requiring more insulin than expected to maintain glucose homeostasis. And that can appear before overt diabetes.

3. Glycated haemoglobin: I do not want to know only your glucose reading from this morning

HbA1c, or glycated haemoglobin, gives us information about average glucose exposure over approximately the last two to three months. This is very different from measuring fasting glucose on one particular morning alone.

A single blood glucose reading is a photograph. HbA1c is more like a film of what has happened over the preceding weeks.

But here too we must be rigorous. An HbA1c of 5.4% does not mean that prediabetes exists. The diagnostic criteria commonly used place prediabetes from an HbA1c of 5.7%, and diabetes, when properly confirmed, from 6.5%.

So why pay attention before that point? Because in preventive medicine we do not necessarily wait for a person to cross a diagnostic threshold before we begin talking about nutrition, physical activity, body composition, sleep, or visceral fat reduction. A laboratory value is not a switch that suddenly flips from health to disease. There is a metabolic continuum.

4. Triglyceride/HDL ratio: a wealth of information hidden in a standard blood test

This is one of my favourite parameters for its simplicity. Triglycerides. HDL. Both are probably already on your blood test.

The relationship between the two can provide indirect information about the metabolic context and has been associated in various populations with insulin resistance and cardiometabolic risk.

But, again, be careful about turning this into a universal rule: a ratio above 2 does not diagnose insulin resistance. Interpretation depends, among other things, on the units used, sex, the population studied, and the individual's metabolic context.

Why is it interesting, then? Because elevated triglycerides combined with low HDL — especially when they coincide with abdominal obesity, hypertension, glycaemic disturbances, or other metabolic signs — can form part of a pattern that deserves our attention. And we can obtain that information from two extraordinarily common parameters.

5. High-sensitivity CRP: metabolism and inflammation are connected

Finally, we have high-sensitivity C-reactive protein, or hs-CRP. It is a marker of low-grade systemic inflammation. We know that chronic inflammation and metabolic dysfunction are deeply interrelated.

Visceral obesity, for example, does not simply consist of storing too much energy. Adipose tissue is biologically active and can produce inflammatory mediators that play a role in the metabolic disturbances associated with obesity.

That is why a persistently elevated high-sensitivity CRP can provide interesting information about the patient's inflammatory and cardiovascular context.

But a high-sensitivity CRP above 1 mg/L does not automatically mean that inflammation is causing insulin resistance. A recent infection, an inflammatory disease, trauma, and many other circumstances can alter it. That is why I never interpret a CRP result in isolation without knowing the patient.

So, what are the actual diagnostic values?

This distinction is fundamental. To diagnose prediabetes and diabetes we have established clinical criteria that include, principally: fasting plasma glucose; HbA1c; and, when indicated, the oral glucose tolerance test.

The five parameters I am discussing with you today do not replace those diagnostic criteria. They help us do something different: broaden our view of metabolism.

Because my goal is not solely to answer: 'Do you have diabetes, yes or no?' I also want to ask myself: 'In which direction is this patient's metabolism heading?'

Context matters more than an isolated number

Let us imagine a patient with still-normal glucose, relatively elevated fasting insulin, an unfavourable HOMA-IR, rising triglycerides, falling HDL, progressively increasing HbA1c, visceral fat, little physical activity and a family history of type 2 diabetes.

I am not going to tell that person they have diabetes if they do not meet the diagnostic criteria. But neither am I going to say: 'Everything is perfect, come back in five years.' That is where true preventive medicine begins, in my view.

What do we do when we detect an unfavourable metabolic trajectory?

This is not about chasing numbers or medicalising a healthy person. It is about acting on what we know can modify risk.

Over the following months we can work on the quality of nutrition, an adequate intake of protein and fibre, the reduction of ultra-processed foods, daily physical activity, strength training, cardiorespiratory fitness, the reduction of visceral fat where there is excess, sleep, alcohol consumption and other individual factors.

We then repeat certain parameters when it is clinically indicated. And we observe the trend, not just an isolated result.

When I evaluate a patient's metabolic health, I do not only want to know whether they have diabetes today: I want to detect where their metabolism is heading so that I can try to change that trajectory while we still have time.