Determining the cutoff level of body mass index for diagnosing obesity in Asian people with chronic spinal cord injury: a cross-sectional diagnostic accuracy study

Technical Summary: Determining BMI Cutoffs for Obesity Screening in Asian Individuals with Chronic Spinal Cord Injury

Problem Statement
Obesity is a critical concern for individuals with chronic spinal cord injury (SCI), significantly increasing the risk of premature cardiovascular disease (CVD), a leading cause of mortality in this population. While Body Mass Index (BMI) is the standard tool for diagnosing obesity, its application in chronic SCI populations is problematic. Individuals with SCI often exhibit "sarcopenic obesity"—characterized by higher fat mass and lower skeletal muscle mass—leading to an underestimation of obesity when using standard BMI cutoffs. Furthermore, existing BMI thresholds for obesity vary by ethnicity (e.g., ≥30 kg/m² for Western populations vs. ≥25 kg/m² for Asian populations) and by gender. Currently, there is a lack of validated BMI cutoffs specifically for the overall Asian SCI population that incorporate both body composition and cardiovascular risk, particularly for Thai individuals.

Methodology
This study employed a cross-sectional design with prospective data collection involving 142 Thai individuals with chronic SCI (injury duration >1 year) aged 20–70 years. Participants were recruited from the Department of Rehabilitation Medicine at Chiang Mai University.

• Data Collection: Researchers gathered demographic data, medical history, and metabolic variables. Anthropometric measurements included weight, height, waist/hip circumference, and four-site skinfold thickness (biceps, triceps, subscapular, suprailiac) to calculate Percentage Body Fat (%BF).
• Reference Standards:
  1. Cardiovascular Risk: The Thai Cardiovascular (CV) Risk Score was calculated to stratify participants into "at-risk" (≥10% 10-year CVD risk) and "not at-risk" groups.
  2. Body Composition: %BF served as the primary reference for defining obesity status.
• Statistical Analysis:
  • Sex-specific %BF cutoffs were determined using Receiver Operating Characteristic (ROC) analysis against the Thai CV Risk Score, utilizing the Youden Index to optimize sensitivity and specificity.
  • A BMI cutoff for screening obesity was subsequently derived using ROC analysis, with the newly established %BF cutoffs serving as the reference standard.
  • Linear regression was used to validate the association between the derived BMI cutoff and the Thai CV Risk Score.

Key Results

• Participant Characteristics: The cohort (n=142) consisted of 73.9% males, with a mean BMI of 22.0 kg/m² and a mean %BF of 24.1%. Approximately 24.6% of participants were classified as "at-risk" for 10-year CVD events.
• Sex-Specific %BF Cutoffs:
  • Females: A %BF cutoff of 34% was selected, yielding a sensitivity of 75.00% and specificity of 62.07% against the CV risk score.
  • Males: While initial data analysis suggested a lower cutoff, the study adopted a 24% cutoff based on prior literature to avoid overestimation and poor specificity observed in the male cohort.
• Proposed BMI Cutoff: Based on the sex-specific %BF thresholds, the optimal BMI cutoff for screening obesity in Thai individuals with chronic SCI was determined to be 22.0 kg/m².
  • Diagnostic Accuracy: This cutoff demonstrated a sensitivity of 84.91% and a specificity of 70.79% (AuROC = 0.88).
  • Validation: The obesity status defined by this 22.0 kg/m² cutoff showed a significant positive correlation with the Thai CV Risk Score (Beta coefficient = 2.96; p = 0.025).

Significance and Claims
The study concludes that a BMI cutoff of 22 kg/m² is an effective screening tool for identifying obesity in Asian individuals with chronic SCI. The authors claim this finding is significant for several reasons:

1. Clinical Utility: The proposed cutoff aligns with cardiovascular risk profiles, suggesting it can effectively identify individuals at higher risk for CVD, thereby facilitating early intervention and management strategies.
2. Cross-Ethnic Consistency: The result (22 kg/m²) is consistent with a study by Laughton et al. conducted in Western (Canadian) SCI populations, implying that ethnicity may play a less crucial role than SCI status itself in determining body composition thresholds for obesity.
3. Differentiation from General Populations: The study reinforces that the standard Asian cutoff (≥25 kg/m²) is too high for the SCI population, where obesity is often masked by muscle atrophy.

Limitations Noted by Authors
The authors acknowledge several constraints:

• Study Design: The cross-sectional nature limits causal inference, and the Thai CV Risk Score was used as a surrogate for actual cardiovascular events rather than a longitudinal outcome.
• Measurement Challenges: Skinfold measurements can be difficult in individuals with high subcutaneous adiposity, though the method was validated by a single operator.
• Generalizability: As a single-center study with a relatively small sample size, the results require external validation in multi-center settings and diverse Asian subpopulations.
• Subgroup Analysis: The sample size precluded the establishment of specific cutoffs stratified by lesion level (e.g., paraplegia vs. tetraplegia).