Coronary calcium scans, fasting insulin, HbA1c, blood pressure, LDL cholesterol, ApoB and BMI all compete for the title of the one number to watch for heart attack risk. This article compares what each one actually measures and where it falls short, then makes the case for a measurement that needs only a tape measure.
If you want to avoid a heart attack, and you can only measure one thing, what would it be? It’s probably not ApoB. The most important measurement is likely something most doctors don’t routinely do, and something you can organise yourself.
Table of Contents
- Why Not Just Look for the Plaque?
- The Blood Sugar Tests
- Blood Pressure
- Cholesterol, the Cause
- Why Not BMI?
- The Number to Steer By: Waist-to-Height Ratio
- How the Rest Rank
- Bringing It Down
- The Bottom Line
- References
Why Not Just Look for the Plaque?
You might think the answer is measuring arterial plaque directly. If heart attacks come from plaque, why not look for the plaque?
The best-known test is a coronary artery calcium (CAC) scan, a quick CT that counts the calcium in the heart's arteries. Plaque usually starts soft, a pocket of cholesterol and inflammation in the artery wall, and over years the body lays calcium over some of it to try and stabilize it so that it doesn’t rupture and lead to a heart attack [1].
The CAC scan only sees that hardened part. In autopsy hearts, the plaque covered about five times the area the calcium did [2].
And it's usually a soft plaque, not a hard one, that sets off the clot behind a heart attack [3, 4].
So a zero score doesn't mean zero plaque [5, 6, 7].
So you might think the answer is a CT coronary angiogram, with dye through a drip, which can show the soft plaque too [8].
Or even an ultrasound of your neck and groin arteries, which finds plaque earlier, and in far more people, than a calcium scan does [9, 10].
These are good tests.
But every one of them waits for the disease to show up, and a picture of plaque doesn't tell you what caused it.
The goal is a measurement that points at a cause, and something that can be addressed early enough so that plaque doesn’t start building up.
The Blood Sugar Tests
So what about blood tests for what drives the plaque? Plenty of people say fasting insulin or fasting glucose is the number to watch, because insulin resistance matters. And it does. In that same Spanish study, people with a high insulin-resistance score had more plaque, even after accounting for their long-term blood sugar [11].
The problem is measuring it. In a small study of healthy volunteers, three days without food made both fasting glucose and insulin look better, while their muscles became more resistant to insulin [12].
And the readings swing on their own: a common insulin-resistance score built from fasting insulin and glucose moved by about a quarter from week to week in the same healthy people [13].
No diabetes guideline diagnoses anything on fasting insulin because of how variable the readings are from day to day. Fasting glucose is an official way to diagnose diabetes, but a high result has to be confirmed [14, 15].
HbA1c fixes most of that. It's the sugar stuck to the haemoglobin inside your red blood cells, a rolling average over their roughly four-month life, weighted to the last few weeks. A weekend of not eating barely moves it [14].
In about 11,000 American adults without diabetes, an HbA1c of 6 to 6.5% went with about 78% more coronary heart disease than 5 to 5.5% [16].
But a normal HbA1c doesn't prove you're insulin sensitive, because the pancreas can hold blood sugar normal for years by pumping out more insulin [17].
And it only tells you about blood sugar. It can be perfect while your blood pressure is quietly too high.
Blood Pressure
Ideally, the measurement should influence insulin sensitivity and blood pressure, rather than looking at these metrics in isolation.
Especially because the blood pressure targets keep getting lower.
In 2017 the American guidelines moved the line for high blood pressure from 140 down to 130. That isn't a drug-company conspiracy. A lot of it came from the SPRINT trial [18].
One group aimed for a systolic blood pressure of around 120, and the other group accepted a blood pressure of 140, which was the standard blood pressure at the time [19].
Long story short, the trial had to be stopped early because the group who aimed for a blood pressure of 120 had about a quarter fewer major heart events compared to the 140 group, and later trials pointed the same way [19, 20].
Doctors do have to individualize this, because frail patients might need a slightly higher blood pressure so that they don’t faint or feel dizzy.
Across 123 studies, every 10 points lower meant about a fifth fewer major heart and stroke events [21]. So it works. But it's one risk factor. You can have perfect blood pressure and a cholesterol problem.
Cholesterol, the Cause
With cholesterol, the evidence goes a step further.
Every cell in our body can produce its own cholesterol, and cholesterol is vital for our cell membranes, hormone production, and much more.
Where we run into trouble is when there’s too much cholesterol in our blood stream.
We know this from an overwhelming amount of data such as this 2017 analysis from the European Atherosclerosis Society, that pulled together more than 200 studies covering over two million people: genetic studies, long-term population studies and randomised trials. People born with lower LDL-cholesterol get far less heart disease, and the further a trial lowers LDL-cholesterol, the fewer heart attacks [22].
Because of this data, the panel concluded, “Consistent evidence from numerous and multiple different types of clinical and genetic studies unequivocally establishes that LDL causes ASCVD [atherosclerotic cardiovascular disease]” [22].
And it matters even without the other risk factors. In the PESA study, people with normal blood pressure, normal blood sugar and normal HbA1c were more likely to have plaque the higher their LDL [10].
Measuring LDL-cholesterol has a problem though. In people with insulin resistance or high triglycerides LDL can look fine, even though the number of particles carrying cholesterol that can penetrate the arterial wall is high [23].
So a better measurement is ApoB.
Every particle that can carry cholesterol into your artery wall carries exactly one ApoB, so ApoB counts the particles themselves [24].
Why Not BMI?
If the question were which single blood test best predicts a heart attack, ApoB would be the king.
But this is a slightly different question.
What is the best overall measurement to avoid a heart attack?
Ideally, we want something that we can act upon before plaque starts building up in our blood vessels.
- Something that influences insulin sensitivity, blood sugar levels, blood pressure, and ApoB
- Something that is cheap and easy to do at home
Body mass index (BMI) might seem like the answer. It’s your weight in kilograms divided by height in meters squared (kg/m²).
Get that right and a lot follows. For every kilogram people lose, the top blood-pressure number comes down by about a point [25]. And in people heading towards diabetes, a lifestyle programme aiming for 7% weight loss cut new cases of type 2 diabetes by 58% [26].
Get body mass index right, and insulin sensitivity follows. Blood pressure follows. ApoB follows unless you are on a carnivore diet or particular keto diets.
But there’s a glaring problem with BMI.
It often misclassifies muscly people as overweight.
And for some people with a normal BMI, they might have central obesity where their legs and arms are skin and bone because they don’t exercise, but their waist is large.
In a large US survey, people with a normal BMI but a big waist compared with their hips had the worst long-term survival of any group, worse even than people with obesity by BMI whose fat wasn't around the middle [27, 28].
The Number to Steer By: Waist-to-Height Ratio
So, arguably, here’s the most important measurement: your waist-to-height ratio.
Here’s why, and then how to bring it down, because the management has recently changed based on new data.
It's your waist measurement divided by your height, and the target is simple: keep your waist to less than half your height, a ratio under 0.5.
Waist-to-height ratio is what the UK's guideline body, NICE, tells doctors to use [29]. And the American Diabetes Association now uses the same line, alongside BMI, to diagnose obesity [30].
Stand up. Find the bottom of your ribs and the top of your hip bone, at your side. Wrap the tape around your middle, halfway between the two. That's usually just above your belly button. Breathe out normally, then measure.
It tracks risk, too. In a British study that followed civil servants for about 16 years, people at 0.5 or above had about two and a half times the risk of developing type 2 diabetes [31].
So why does the waist do so much? Genetic studies suggest that carrying fat around your middle actually causes heart disease, rather than just travelling with it [32].
People whose genes push their fat towards the belly have more type 2 diabetes and more coronary heart disease [32]. And these studies can also estimate how. When researchers split up the ways extra weight raises coronary risk, most of it ran through diabetes and blood pressure. Very little ran through cholesterol [33].
How the Rest Rank
A very close second is ApoB.
A slim waist doesn't completely protect you from a high ApoB, as evidenced by the PESA study [10]. And Lp(a) is another inherited cholesterol particle that your waist doesn't change, and it usually only needs testing once [34].
Third, on equal footing, HbA1c and blood pressure. A waist under half your height makes high blood pressure less likely. Not impossible. People with the most blood-pressure-raising genes run about 10 points higher than those with the fewest [35].
And about one in seventeen people with high blood pressure in general practice turns out to have primary aldosteronism, where the adrenal glands make too much of a salt-retaining hormone [36].
So this isn't a reason to skip the other tests. Measure them. But if only one could be picked, it would be waist-to-height ratio.
What about a DEXA scan (a low-dose X-ray that measures body fat and bone) instead?
It probably isn’t needed, unless you are screening for osteoporosis. In about 470,000 people in the UK Biobank, the link between body-fat percentage and heart attacks and strokes was mostly explained by waist-to-height ratio [37].
Bringing It Down
So, how do you get your waist-to-height ratio down to below 0.5?
Diet and exercise, of course. But in one big lifestyle trial, in people with type 2 diabetes, modest weight loss over almost ten years didn't cut heart attacks or strokes [38]. For many people diet and exercise alone aren't enough, and this is where some recent breakthroughs have changed the picture.
GLP-1 medications have been shown to reduce major heart events in people without diabetes [39].
But we’ve moved forward from just GLP-1 meds such as semaglutide.
We now have tirzepatide, which combines GLP-1 with GIP.
Retatrutide, the next drug along, adds a third hormone. It isn't approved anywhere yet; the company plans to ask the FDA early next year [40]. In its main published trial that was just released, people on the top dose lost about a quarter of their body weight, counting everyone who started [41].
But the trials so far had too few heart attacks and strokes to tell whether it prevents them, and the proper heart trial is due to finish around 2029 [40, 42].
That’s three hormones that we’ve discussed, and now there’s a fourth on the horizon: amylin.
Your pancreas releases it alongside insulin, and it acts on its own receptors in the brain to signal that the meal is over [43].
Eloralintide prefers one of the amylin receptors over the calcitonin receptor about twelve to one [43]. In theory, the more targeted drug should cause fewer side effects, and fewer side effects would mean more people stay on it.
When eloralintide was added to tirzepatide, so a triple hormone therapy, people with type 2 diabetes lost up to 23 percent over 48 weeks if they stayed on it, against 15 percent on tirzepatide alone [44].
The main issue though with amylin agonists is the nausea.
If the side effects can be tamed, tirzepatide plus an amylin drug, with no glucagon, could become another long-term option. And further out, you could even see eloralintide added to retatrutide, four hormones in one treatment.
Plus these drugs seem to have benefits beyond their weight loss effect.
They reduce heart disease, fatty liver, kidney disease, and there’s emerging evidence that they may help to regrow cartilage [39, 45, 46, 47].
In a 24 week trial where the knee cartilage was measured in an MRI machine, the semaglutide group showed an average 17% increase in cartilage thickness, suggesting regeneration. The control group: less than 1% [47]. The thickened cartilage was visible in weight-bearing areas of the knee, the areas that take the most punishment.
Patients also experienced reduced pain and improved joint function [47].
Side effects with these drugs are mostly in the gut, and with tirzepatide they were mostly mild to moderate and came while the dose was being raised [48, 49]. And the weight doesn't stay off on its own: in one tirzepatide trial, people switched to placebo after about eight months put back about 14% of their body weight over the next year [50].
The biggest side effect is cost, but prices are coming down and there are now oral versions that are further reducing costs [51].
The Bottom Line
So if there can only be one number, it's your waist, under half your height. Then get your ApoB checked, and keep an eye on your HbA1c and blood pressure.
References
1. https://pubmed.ncbi.nlm.nih.gov/24247483/
2. https://pubmed.ncbi.nlm.nih.gov/7554196/
3. https://pubmed.ncbi.nlm.nih.gov/16631505/
4. https://pmc.ncbi.nlm.nih.gov/articles/PMC6447371/
5. https://pubmed.ncbi.nlm.nih.gov/37821352/
6. https://pubmed.ncbi.nlm.nih.gov/19520336/
7. https://pubmed.ncbi.nlm.nih.gov/33129734/
8. https://pubmed.ncbi.nlm.nih.gov/36075621/
9. https://pubmed.ncbi.nlm.nih.gov/32273027/
10. https://www.sciencedirect.com/science/article/pii/S0735109721051159
11. https://pmc.ncbi.nlm.nih.gov/articles/PMC10731750/
12. https://pmc.ncbi.nlm.nih.gov/articles/PMC4971894/
13. https://pubmed.ncbi.nlm.nih.gov/38987271/
14. https://pmc.ncbi.nlm.nih.gov/articles/PMC12690183/
15. https://pubmed.ncbi.nlm.nih.gov/17646610/
16. https://pmc.ncbi.nlm.nih.gov/articles/PMC2872990/
17. https://pmc.ncbi.nlm.nih.gov/articles/PMC2726723/
19. https://pmc.ncbi.nlm.nih.gov/articles/PMC4689591/
20. https://pubmed.ncbi.nlm.nih.gov/38945140/
21. https://pubmed.ncbi.nlm.nih.gov/26724178/
22. https://pmc.ncbi.nlm.nih.gov/articles/PMC5837225/
23. https://pmc.ncbi.nlm.nih.gov/articles/PMC4501252/
24. https://pmc.ncbi.nlm.nih.gov/articles/PMC7369156/
25. https://pubmed.ncbi.nlm.nih.gov/12975389/
26. https://pmc.ncbi.nlm.nih.gov/articles/PMC1370926/
27. https://pubmed.ncbi.nlm.nih.gov/18283284/
28. https://pmc.ncbi.nlm.nih.gov/articles/PMC4995595/
30. https://pmc.ncbi.nlm.nih.gov/articles/PMC13266248/
31. https://pubmed.ncbi.nlm.nih.gov/40404050/
32. https://pmc.ncbi.nlm.nih.gov/articles/PMC5571980/
33. https://pmc.ncbi.nlm.nih.gov/articles/PMC8236409/
34. https://pmc.ncbi.nlm.nih.gov/articles/PMC9639807/
35. https://pmc.ncbi.nlm.nih.gov/articles/PMC6284793/
36. https://pubmed.ncbi.nlm.nih.gov/28385310/
37. https://pmc.ncbi.nlm.nih.gov/articles/PMC11196863/
38. https://pubmed.ncbi.nlm.nih.gov/23796131/
39. https://www.nejm.org/doi/full/10.1056/NEJMoa2307563
41. https://www.nejm.org/doi/full/10.1056/NEJMoa2604169
42. https://clinicaltrials.gov/study/NCT06383390
43. https://pmc.ncbi.nlm.nih.gov/articles/PMC12640043/
45. https://pubmed.ncbi.nlm.nih.gov/40305708/
46. https://pubmed.ncbi.nlm.nih.gov/38785209/
47. https://www.cell.com/cell-metabolism/fulltext/S1550-4131(26)00008-2
49. https://www.nejm.org/doi/full/10.1056/NEJMoa2206038
