Overweight, Obesity May Lead to More UK Hospitalizations than Previously Estimated

Using Mendelian randomization, a new study found in particular that waist-hip ratio may be a more important predictor of hospitalization than BMI.

The impact of excess weight on hospital admissions in the United Kingdom (UK) may be larger than conventional epidemiologic analyses report, according to new research from investigators at the University of Bristol.

Importantly, this first Mendelian randomization study of adiposity’s role in hospitalization, morbidity, and mortality found that the relationship was driven primarily by regional adiposity (indexed by waist-to-hip ratio [WHR]) and not by overall body mass index (BMI), according to study authors.

Led by Audinga-Dea Hazewinkel from the University’s Bristol Medical School and Population Health Sciences Institute, the researchers write in Economics and Human Biology, that understanding the causal impact of adiposity on inpatient admissions is essential to understanding the larger influence of “adverse weight profiles” on the health system and cost of health care. They add, however, that existing findings from observational assessments of the association are “challenged by endogeneity attributable to unobserved confounding and reverse causation" that precludes the ability to accurately infer causality.

Their solution: Mendelian randomization.

For their analysis, Hazewinkel and colleagues used UK Biobank hospital inpatient admission data from 310 471 participants and more than 550 000 hospital admissions. Baseline Biobank measures were used for weight, bio-impedance, height, waist circumference, and hip circumference.

Participants had an average age of 57.4 years, a mean BMI of 27.38 kg/m2, and mean WHR of 0.87. The cohort was 54% women. Average follow-up was 6.5 years.

Using this data, the researchers compared estimates from conventional epidemiological analyses using Mendelian randomization, leveraging changes in the genome linked to body composition to estimate the causal effect of being overweight on a health outcome and to remove potential confounding by factors that may jointly influence body composition and rates of hospital admission.

Their results found evidence for a direct causal effect of higher BMI and WHR on higher yearly hospital admission rates, with estimates that were larger than those obtained from existing research. One of the team’s most striking discoveries showed the relationship was largely driven by adverse fat distribution in a certain area (measured by waist-hip ratio) rather than overall BMI.

Investigators report that participants were 16% to 26% more likely to be admitted to hospital per each 0.09-unit higher WHR vs 8% to 16% more likely with each 4.74 kg/m2 higher BMI.

For example, for a woman in this study of average height (5’3”) and weight (143 lbs) and with average waist (34”) and hip (40”) measurements, the change would be the equivalent of gaining 3.7” in waist circumference, and just under 28 lbs in weight, respectively. For a man in this study of average height (5’8”) and weight (174 lbs), and average measurements of waist (37”) and hip (40”), this would correspond to a 3.6-inch increase in waist circumference and an increase in weight of 33 lbs.

“We live in increasingly obesogenic environments with the World Health Organisation identifying 39% of men and 40% of women as being overweight, and 11% of men and 15% of women as obese worldwide,” said Hazewinkel in a University of Bristol statement. “Finding causal effect estimates between fatty tissue and hospital admissions larger than those previously reported in existing studies emphasises the necessity of exploring policies aimed at reducing obesity in the population.

“The results also suggest that a preference should be given to waist-hip ratio as a measure of body fat over BMI as this may be more important for predicting hospital admissions.”

Reference: Hazewinkel A-D, Richmond RC, Wade KH, et al. Mendelian randomization analysis of the causal impact of body mass index and waist-hip ratio on rates of hospital admission. Econ Human Biol. 2022;44. https://doi.org/10.1016/j.ehb.2021.101088