· Lee Chee Cheow, M.Sc. Dip CMP · Metabolic Health · 4 min read
Obesity and Diabetes Risk: The BMI Connection
The relationship between BMI and diabetes risk isn't linear — it's closer to exponential, and it doesn't affect men and women equally. Here's what the research shows, and how the major weight-management approaches compare.
Of all the modifiable risk factors for type 2 diabetes, body weight is the one most consistently singled out — and the research behind it is more striking than the general advice to “maintain a healthy weight” suggests. The relationship between BMI and diabetes risk isn’t a gentle slope; it curves sharply upward, and it does so differently for men and women.
How steeply risk rises with BMI
Long-standing cohort data (Chan et al., Diabetes Care, 1994; Colditz et al., Annals of Internal Medicine, 1995) tracked age-adjusted relative risk of type 2 diabetes across BMI bands, separately for men and women:
| BMI (kg/m²) | Women — relative risk | Men — relative risk |
|---|---|---|
| Under 22 | 1.0 | 1.0 |
| Under 23 | 2.9 | 1.0 |
| 23-23.9 | 4.3 | 1.0 |
| 24-24.9 | 5.0 | 1.5 |
| 25-26.9 | 8.1 | 2.2 |
| 27-28.9 | 15.8 | 4.4 |
| 29-30.9 | 27.6 | 6.7 |
| 31-32.9 | 40.3 | 11.6 |
| 33-34.9 | 54.0 | 21.3 |
| 35 and above | 93.2 | 42.1 |
Two things stand out. First, the curve is closer to exponential than linear — relative risk roughly doubles with each BMI band increase in the higher ranges, rather than climbing steadily. Second, in this dataset, women showed a markedly steeper risk increase than men at equivalent BMI levels — for example, at a BMI of 35 and above, women’s relative risk (93.2) was more than double men’s (42.1) compared to their respective baselines.
A separate, more recent systematic review and meta-analysis (International Journal of Obesity, 2022) grouped risk by standard weight categories rather than fine BMI bands, and found a similarly steep pattern:
| Weight category | Relative risk of diabetes |
|---|---|
| Severe obesity (BMI over 35) | 22.97 |
| Obesity (BMI 30-34.9) | 4.56 |
| Overweight (BMI 25-29.9) | 2.24 |
| Underweight (BMI under 18.5) | 0.93 |
Severe obesity was associated with nearly a 23-fold increase in relative risk compared to the reference group — among the largest single-factor risk multipliers found anywhere in this body of research.
What actually causes obesity
Obesity itself is described in the research as “a complex and multifactorial condition” — the causes vary considerably from person to person, and rarely come down to one factor alone. A 2018 umbrella review in PLoS One identified the main contributing categories:
- Overeating and poor diet — consuming more calories than the body needs, particularly from high-calorie, low-nutrient foods
- Sedentary lifestyle — insufficient physical activity
- Genetics and family history — obesity tends to run in families, pointing to a genetic component
- Medical conditions — such as hypothyroidism or Cushing’s syndrome
- Medications — certain antidepressants and steroids can cause weight gain as a side effect
This list matters because it means “eat less, move more” doesn’t address every case — a medical or medication-related cause needs a doctor’s involvement, not just a diet change.
Comparing weight-management approaches
For people managing weight through diet, a 2020 network meta-analysis in the BMJ compared major dietary strategies across effectiveness, time to results, and downsides:
| Strategy | Effectiveness for weight loss | Time to results | Side effects | Contraindications |
|---|---|---|---|---|
| High-protein, fibre-rich | Moderate (improves satiety) | Gradual — months to years | Bloating, constipation | Kidney disease (due to high protein intake) |
| Mediterranean / DASH | Moderate (improves metabolic health) | Slow — needs long-term adherence | Generally well tolerated | None major |
| Intermittent fasting | Moderate to high (improves metabolism) | Rapid — weeks to months | Hunger, dizziness, irritability | Hypoglycaemia, eating disorders, pregnancy |
| Low-carb ketogenic | High, especially with insulin resistance | Rapid — weeks to months | Nutrient deficiencies | Kidney disease, liver disease, pregnancy |
No single approach wins on every dimension. The faster-acting strategies (intermittent fasting, low-carb ketogenic) carry more restrictive contraindications, while the slower, more sustainable approaches (Mediterranean/DASH) have the fewest downsides but require longer-term commitment to see comparable results.
The takeaway
BMI’s relationship with diabetes risk is steep enough, and different enough between men and women, that “a bit overweight” and “obese” are not close cousins on a risk chart — they’re categories separated by a several-fold difference in relative risk. Given the multiple possible causes of obesity, the right approach depends on which of those factors is actually driving the weight gain in the first place, which is a conversation worth having with a doctor rather than assuming diet and exercise alone will always be the answer.
Frequently asked questions
Does BMI affect diabetes risk the same way in men and women? Not according to this cohort data — women showed a steeper relative-risk increase than men at comparable BMI levels, particularly in the higher BMI bands.
Which weight-loss approach works fastest? Intermittent fasting and low-carb ketogenic diets were rated rapid (weeks to months) in this comparison, versus gradual-to-slow for high-protein/fibre-rich and Mediterranean/DASH approaches — but the faster options carry more contraindications, including pregnancy and existing kidney or liver disease.
Is being underweight also a diabetes risk factor? No — this data shows being underweight (BMI under 18.5) was associated with a relative risk of 0.93, essentially no different from the reference group, unlike overweight and obese categories which showed clearly elevated risk.
This article is for general information only and does not constitute medical advice. If you are managing your weight or have a diagnosed diabetic condition, consult your doctor before starting any new diet or exercise programme.