Underweight in Type 2 Diabetes Raises Cardiovascular Risk — Severe Leanness Carries Greater Risk Than Obesity

Highlight

– A Korean nationwide cohort of >2 million adults with Type 2 diabetes (T2DM) found that underweight status was associated with significantly higher risk of cardiovascular disease (CVD) compared with normal weight; risks were greatest in severe underweight (BMI <16 kg/m2).

– Adjusted hazard ratios for composite CVD were 1.49 (95% CI 1.28–1.73) for severe, 1.47 (1.33–1.63) for moderate, and 1.19 (1.13–1.25) for mild underweight versus normal weight; this exceeded risk for very high BMI (≥35 kg/m2: 1.14 [1.08–1.22]).

Background

Clinical attention in cardiometabolic medicine has traditionally focused on overweight and obesity as dominant drivers of Type 2 diabetes and cardiovascular disease. However, an appreciable and clinically important subgroup of people with T2DM are lean or underweight. In East Asian populations, lean T2DM is increasingly recognized, and body composition differences (less subcutaneous fat but early beta-cell dysfunction) influence disease course and complication risk. Existing epidemiologic work on body mass index (BMI) and mortality often demonstrates a J- or U-shaped relationship: both low and high BMI predict worse outcomes. Until now, large-scale, BMI-stratified analyses in people with T2DM that disaggregate underweight by severity have been limited.

Study design

Park et al. analyzed data from the Korean National Health Insurance Service (NHIS), performing a population-based cohort study of 2,064,406 adults with T2DM who underwent standardized health examinations between 2015 and 2016. Participants were categorized by BMI into underweight, normal weight, overweight, and obese groups. Underweight was further stratified as severe (<16.0 kg/m2), moderate (16.0–16.9 kg/m2), and mild (17.0–18.4 kg/m2). Primary outcomes were incident cardiovascular events, namely myocardial infarction (MI) and ischemic stroke, with follow-up through administrative records over a mean of 5.7 years. Time-to-event associations were assessed using Cox proportional hazards models adjusted for demographic variables, clinical comorbidities, and lifestyle factors including smoking and alcohol intake.

Key findings

Population and outcomes

Over a mean 5.7-year follow-up, there were 111,522 CVD events among the cohort, including 55,622 MIs and 62,099 ischemic strokes. Underweight individuals—though a minority of the cohort—experienced disproportionately high rates of CVD events.

Main adjusted associations

After multivariable adjustment, underweight was consistently associated with higher CVD risk compared with normal weight. The adjusted hazard ratios (aHRs) for composite CVD (with 95% confidence intervals) were:

• Severe underweight (<16.0 kg/m2): aHR 1.49 (1.28–1.73)
• Moderate underweight (16.0–16.9 kg/m2): aHR 1.47 (1.33–1.63)
• Mild underweight (17.0–18.4 kg/m2): aHR 1.19 (1.13–1.25)

For comparison, extreme obesity (BMI ≥35 kg/m2) carried a smaller adjusted excess risk: aHR 1.14 (1.08–1.22). The pattern produced a reverse J-shaped association between BMI and CVD risk, with nadir risk in the normal–overweight range and rising risk at both extremes, but steeper on the underweight side in this T2DM population.

Secondary and subgroup results

The associations were stronger in younger participants and in non-smokers. These stratified findings argue against smoking as the only driver of the low-BMI risk signal and suggest that underweight confers independent cardiovascular vulnerability particularly among younger adults with diabetes. Event-specific analyses showed increased risk for both MI and ischemic stroke among underweight groups.

Robustness

Associations persisted after adjustment for many conventional confounders and in sensitivity analyses reported by the authors, supporting a robust signal. The large sample size and national administrative capture reduced random error and improved the precision of estimates, permitting fine stratification within the underweight range.

Expert commentary and interpretation

The Park et al. study makes an important contribution by quantifying a graded and clinically meaningful increase in cardiovascular risk at low BMI levels among people with T2DM in a large East Asian population. Several points merit emphasis for clinicians and researchers.

Potential mechanisms

Low BMI in T2DM likely represents a heterogeneous set of phenotypes. Mechanistic pathways plausibly include:

• Sarcopenia and low muscle mass: reduced skeletal muscle is linked to insulin resistance, impaired glucose disposal, and adverse cardiac outcomes.
• Undernutrition and micronutrient deficiencies: can impair cardiac reserve and promote arrhythmia or ischemic vulnerability.
• Frailty and comorbidity burden: unintentional weight loss may reflect occult illness (cancer, chronic inflammatory disease) or frailty that increases CVD risk.
• Adverse lipid and hemostatic profiles: despite lower total adiposity, adverse metabolic phenotypes (e.g., ectopic fat, hepatic steatosis) can exist and modify risk.

These mechanisms are not mutually exclusive and likely interact with glycemic control, medication effects (e.g., some glucose-lowering drugs can cause weight loss), and socioeconomic determinants of health.

Confounding and reverse causality

Observational data cannot prove causality. Low BMI may be a marker rather than a cause of higher CVD risk: preexisting illness could cause weight loss and increase near-term event risk. The stronger associations in younger and non-smoking subgroups somewhat reduce—but do not eliminate—concerns about confounding by smoking or age-related frailty. Detailed data on weight trajectory, body composition, inflammatory biomarkers, and cause-specific mortality would help disentangle causality.

Generalizability

These results are drawn from a Korean population, with distinct body composition and diabetes phenotypes compared with Western populations. Absolute risks and BMI thresholds may not translate directly to other ethnic groups, but the qualitative finding—that severe underweight is not benign in T2DM—likely has broad applicability and warrants attention across settings.

Clinical implications

For clinicians, the study prompts re-evaluation of routine care pathways that prioritize weight reduction without parallel strategies to identify and manage underweight patients. Practical actions include:

• Systematic screening: identify underweight patients with T2DM using BMI and, where available, measures of muscle mass and function (e.g., grip strength, gait speed).
• Assess for causes: evaluate for unintentional weight loss, malnutrition, eating disorders, gastrointestinal disease, malignancy, and medication-related effects.
• Nutritional and rehabilitative interventions: consider referral for dietitian-led assessment, high-quality protein supplementation, resistance exercise programs, and treatment of micronutrient deficiencies.
• Individualized risk management: do not withhold established cardioprotective therapies (statins, antihypertensives, SGLT2 inhibitors/GLP‑1 receptor agonists where indicated) based solely on low BMI—underweight patients may benefit equally or more from guideline-based secondary prevention.

Research priorities

Key gaps include prospective studies that incorporate body composition, muscle function, markers of malnutrition and inflammation, and longitudinal weight change; interventional trials to test whether nutritional and exercise interventions reduce CVD events in underweight T2DM; and mechanistic work to link sarcopenia and cardiac vulnerability.

Limitations of the study

Important limitations include the observational design with potential residual confounding and reverse causality; reliance on BMI without direct measures of body composition or nutritional status; limited information on intentional versus unintentional weight loss; and potential misclassification from administrative coding of events. Geographic and ethnic specificity to Korea may limit global generalizability.

Conclusion

Park et al. provide compelling population-level evidence that underweight status in people with Type 2 diabetes is associated with substantially increased cardiovascular risk, particularly at severely low BMI. Clinicians should recognize underweight as a high-risk phenotype in diabetes care, undertake directed evaluation for malnutrition and sarcopenia, and ensure that underweight patients receive guideline-based cardiovascular risk management. Public health and guideline bodies should consider including explicit recommendations on the assessment and management of underweight T2DM patients. Further research is needed to determine causality and to evaluate interventions that could mitigate this excess risk.

Funding and clinicaltrials.gov

The published report used data from the Korean National Health Insurance Service. Specific study funding and trial registration were not reported in the summary citation provided. This analysis appears to be an observational cohort study of administrative data rather than an interventional trial and therefore would not typically have a clinicaltrials.gov registration.

References

1. Park JH, Han K, Choe HJ, Jang HN, Kwon H, Park SE, Rhee EJ, Moon SJ, Lee WY. Underweight and cardiovascular risk in Type 2 diabetes: a Korean nationwide study. Eur Heart J. 2025 Nov 25:ehaf903. doi: 10.1093/eurheartj/ehaf903. Epub ahead of print. PMID: 41288350.

2. Global BMI Mortality Collaboration. Body-mass index and all-cause mortality: individual-participant-data meta-analysis of 239 prospective studies in four continents. Lancet. 2016 Aug 20;388(10046):776–786. doi:10.1016/S0140-6736(16)30175-1.

Note: For practical guidance, clinicians should consult contemporary diabetes and cardiovascular guidelines (e.g., ADA Standards of Care; ESC/EASD guidance) for specific recommendations about risk factor management in patients with diabetes.