Discordant Obesity Severity Classification Between the Edmonton Obesity Staging System and the Lancet Commission Model

This study utilizing UK Biobank data reveals that while the Edmonton Obesity Staging System and the Lancet Commission Diagnostic Model show high concordance for established disease, they differ significantly in early-stage classification due to distinct operational definitions of organ dysfunction, suggesting that integrating both frameworks could enhance obesity risk stratification and intervention targeting.

The Gist

The Big Picture: Why "Weight" Isn't the Whole Story

Imagine you are looking at a house. For decades, doctors have judged how "dangerous" a house is just by looking at its square footage (how big the footprint is). In the medical world, this is BMI (Body Mass Index).

But here's the problem: A massive mansion with a brand-new roof, strong foundation, and working plumbing might be safer than a small, cramped shack that is on fire, has a leaking roof, and is full of mold.

This study asks: Does a big house (high BMI) always mean the house is in trouble? And how do we actually measure the "damage" inside?

The researchers compared two different "inspection manuals" used to grade the severity of obesity:

1. The EOSS (Edmonton System): Think of this as the "Damage Control" inspector. It focuses on what is already broken (heart disease, diabetes, joint pain). It asks, "Is the house currently on fire?"
2. The DMO (Lancet Model): Think of this as the "Early Warning" inspector. It looks for the first signs of trouble (early organ stress, subtle metabolic issues) before the house is fully burning down. It asks, "Are the pipes starting to rust?"

The Experiment: Checking 411,000 Houses

The researchers took data from the UK Biobank, a massive database of about 411,000 people. They ran every single person through both "inspection manuals" to see if they agreed on who was in trouble.

Here is what they found:

1. The "Fire" vs. The "Rust"

• Agreement on the Big Fires: When people were very sick (advanced obesity with major organ damage), both inspectors agreed. They both said, "This house is in critical condition."
• Disagreement on the Early Stages: This is where they differed.
  • The EOSS (Damage Control) was strict. If your organs were working okay, even if you were heavy, it said, "You're fine for now." It mostly ignored people with "rusty pipes" but no fires yet.
  • The DMO (Early Warning) was more sensitive. It caught people who had excess fat and early signs of organ stress, even if they didn't have full-blown diabetes or heart failure yet. It identified a "pre-clinical" group that the other system missed.

2. The "Re-Classification" Surprise

The study also tested what happens if you change the rules for who gets labeled "Obese."

• The Old Rule: Just look at the square footage (BMI > 30).
• The New Rule: Look at the square footage plus where the weight is stored (Waist size, body fat percentage).

The Result: Changing the rules changed the list of "at-risk" people by about 50%.

• Analogy: Imagine a school that only admits kids over 6 feet tall. If you change the rule to "6 feet tall AND wearing red shoes," you suddenly get a totally different group of students. Some very tall kids get kicked out (because they aren't wearing red), and some shorter kids get in (because they are wearing red).
• The study found that adding waist measurements or body fat scans moved half the people into different categories. This proves that BMI alone is a blunt instrument.

3. The "Age" and "Visceral Fat" Factor

The researchers found that the "damage" inside the body wasn't just about how heavy you were.

• Age: The older you get, the more likely you are to have "damage," regardless of your weight.
• Visceral Fat: This is the fat stored deep inside your belly, wrapping around your organs (like grease around an engine). The study found that people with high levels of this "deep fat" had much worse health outcomes than people with the same weight but more "subcutaneous fat" (the soft fat you can pinch on your arm).

What Does This Mean for You?

1. The "Obese" Label is Too Simple
Just because someone has a high BMI doesn't mean they are sick, and just because someone has a "normal" BMI doesn't mean they are healthy. You need to look under the hood.

2. We Need a Better Checklist
The paper suggests that doctors should stop relying only on the scale. They should combine:

• The Weight (BMI)
• The Shape (Waist size)
• The Damage (Blood pressure, blood sugar, organ function)

3. Two Tools for Two Jobs

• Use the EOSS approach when you need to know: "How urgent is this? Does this patient need immediate, intensive treatment?" (Focus on the fire).
• Use the DMO approach when you want to know: "Who is at risk before it's too late? Who needs lifestyle changes to stop the rust?" (Focus on the early warning).

The Bottom Line

Obesity isn't just about being heavy; it's about how that weight affects your body's machinery. This study shows that we need to stop using a single ruler (BMI) to measure a complex problem. By combining different tools—checking the weight, the shape, and the actual health of the organs—we can catch the "rust" before the house burns down, and treat the right people at the right time.

Technical Summary

1. Problem Statement

Current obesity classification relies heavily on Body Mass Index (BMI), which fails to capture the heterogeneity of obesity-related health risks, organ dysfunction, and functional impairments. Two major frameworks have emerged to address this:

• Edmonton Obesity Staging System (EOSS): Focuses on prognostic severity, staging obesity based on established comorbidities, end-organ damage, and functional limitations.
• Lancet Commission Diagnostic Model for Obesity (DMO): Proposed in 2025, this model distinguishes between "preclinical" (excess fat without current dysfunction) and "clinical" obesity (presence of organ/tissue dysfunction), emphasizing a multidimensional assessment beyond BMI.

The Gap: It remains unclear how these two frameworks overlap, whether they identify the same high-risk phenotypes, and how the inclusion of additional anthropometric measures (beyond BMI) alters obesity categorization and staging.

2. Methodology

• Cohort: The study utilized data from the UK Biobank (UKBB), a large prospective cohort.
  • Sample Size: $N \approx 411,000$ participants (restricted to White British/Irish backgrounds to minimize ethnic bias; exclusions applied for cancer, liver disease, pregnancy, etc.).
  • Demographics: 54% women, mean age $56.5 \pm 8$ years.
• Classification Systems Applied:
  1. Modified EOSS: Adapted from the Clinical Obesity Chronic Disease Dashboard.
     • Criteria: BMI $\ge 30$ kg/m². Stages 0–3 were assigned based on the presence of risk factors, established chronic diseases (e.g., T2D, hypertension), end-organ damage (e.g., heart failure, osteoarthritis), and functional/mental health limitations (ADL, GAD-7, PHQ-9).
     • Note: Stage 4 (severe disability) was not assessed due to lack of data in population cohorts.
  2. Lancet Commission DMO:
     • Criteria: BMI $\ge 30$ kg/m² AND body fat percentage >25% (men) or >30% (women), OR BMI >40 kg/m² (where adiposity is assumed).
     • Staging: Classified as Preclinical (excess fat only) or Clinical (presence of organ/tissue dysfunction in metabolic, cardiovascular, musculoskeletal, respiratory, renal, or CNS systems, or limitations in daily activities).
• Statistical Analysis:
  • Cross-classification of individuals under both systems.
  • Analysis of the impact of varying anthropometric thresholds (BMI alone vs. BMI + Waist Circumference [WC], Waist-to-Hip Ratio [WHR], Waist-to-Height Ratio [WHtR], or Body Fat %).
  • Correlation of staging with age, MRI-derived visceral adipose tissue (VAT), and standard anthropometrics.
  • Software: R (v4.4.0) on the UKBB Research Analysis Platform.

3. Key Contributions

• Systematic Comparison: First large-scale comparison of the EOSS and the newly proposed DMO framework within a single population.
• Reclassification Analysis: Demonstrated that the choice of diagnostic thresholds (BMI alone vs. BMI + anthropometrics) significantly alters the prevalence of obesity, reclassifying up to ~50% of individuals depending on the criteria used.
• Phenotypic Differentiation: Revealed that while both systems agree on advanced disease, they diverge significantly in early-stage identification, with DMO capturing a broader spectrum of subclinical dysfunction compared to EOSS's focus on established disease.

4. Key Results

• Prevalence and Distribution:
  • Approximately 24% of the cohort was classified as having obesity under both frameworks.
  • EOSS Distribution: Among those with obesity, 0.4% were Stage 0 (no risk), 8.4% Stage 1 (subclinical risk), 42.3% Stage 2 (established disease), and 48.9% Stage 3 (end-organ damage).
  • DMO Distribution: Clinical obesity was 8.6 times more frequent than preclinical obesity. The most common dysfunctions were metabolic (73%), cardiovascular (58%), and musculoskeletal (51%).
• Discordance and Concordance:
  • High Concordance in Advanced Stages: Both systems effectively identified individuals with established chronic diseases and end-organ damage.
  • Divergence in Early Stages:
    • DMO captured a broader spectrum of mild/subclinical organ dysfunction (e.g., early renal abnormalities, hyperglycemia without diabetes).
    • EOSS was stricter, requiring established disease (e.g., diagnosed T2D, hypertension) for higher stages.
    • Example: Of the 10,005 individuals classified as "preclinical obesity" by DMO, only 279 were assigned to EOSS Stage 1. Conversely, 279 of 420 EOSS Stage 0 individuals were classified as preclinical obesity by DMO.
• Impact of Anthropometric Thresholds:
  • Combining BMI $\ge 30$ kg/m² with Body Fat %, WC, or WHtR resulted in stable classifications similar to BMI alone.
  • Combining BMI with WHR was the least restrictive, increasing obesity prevalence the most.
  • Lowering the BMI threshold to $\ge 25$ kg/m² (overweight range) combined with additional parameters caused the most significant reclassification, highlighting the sensitivity of early-stage assignment to diagnostic definitions.
• Correlates of Severity:
  • Obesity severity (in both systems) correlated more strongly with age and MRI-derived Visceral Adipose Tissue (VAT) than with BMI, WC, or WHR.
  • VAT showed a clear stepwise increase across severity stages, whereas standard anthropometrics remained relatively stable across stages.

5. Significance and Implications

• Complementary Roles: The study concludes that EOSS and DMO offer complementary perspectives. EOSS provides prognostic granularity for established disease, while DMO offers a multidimensional approach for early detection and functional assessment.
• Clinical Practice: Integrating both frameworks could improve risk stratification, allowing clinicians to:
  • Identify "preclinical" individuals (via DMO) for early intervention before organ damage occurs.
  • Prioritize "clinical" or advanced-stage patients (via EOSS) for intensive management.
• Beyond BMI: The findings reinforce that BMI alone is insufficient. The substantial heterogeneity within BMI categories (e.g., Class II/III obesity containing many "preclinical" or low-stage individuals) necessitates the routine use of anthropometric measures (WC, WHtR) or body fat assessment.
• Standardization Need: The study highlights the urgent need for harmonized diagnostic criteria to avoid inconsistencies in research and clinical practice, particularly regarding the definition of "clinical obesity" and the thresholds for early-stage dysfunction.

Limitations Noted:

• Inability to definitively prove causality between adiposity and comorbidities in the UKBB.
• Reliance on self-reported functional limitations (potential bias).
• Underdiagnosis of liver disease and certain organ dysfunctions in population cohorts.
• The "healthy volunteer" bias of the UKBB may limit generalizability to the general population.