Dataset Information

Modelling future trajectories of obesity and body mass index in England.

ABSTRACT:

Background

Obesity is a leading risk for poor health outcomes in England. We examined best- and worst-case scenarios for the future trajectory of the obesity epidemic.

Methods

Taking the last 27 years of Health Survey for England data, we determined both position and shape of the adult body mass index (BMI) distribution and projected these parameters 20 years forward in time. For the best-case scenario, we fitted linear models, allowing for a quadratic relationship between the outcome variable and time, to reflect a potential reversal in upwards trends. For the worst-case scenario, we fitted non-linear models that applied an exponential function to reflect a potential flattening of trends over time. Best-fitting models were identified using Monte Carlo cross-validation on 1991-2014 data, and predictions of population prevalence across five BMI categories were then validated using 2015-17 data.

Results

Both linear and non-linear models showed a close fit to observed data (mean absolute error <2%). In the best-case scenario, the proportion of the population at increased risk (BMI≥25kg/m2) is predicted to fall from 66% in 2017 to 53% (95% confidence interval: 41% to 64%) in 2035. In the worst-case scenario, this proportion is likely to remain relatively stable overall- 64% (37% to 90%) in 2035 -but with an increasing proportion of the population at highest risk (BMI≥35kg/m2).

Conclusions

While obesity prediction depends on chosen modelling methods, even under optimistic assumptions it is likely that the majority of the English population will still be at increased risk of disease due to their weight until at least 2035, without greater allocation of resources to effective interventions.

SUBMITTER: Cobiac LJ

PROVIDER: S-EPMC8172072 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:BackgroundAccurate obesity classification is important so that appropriate intervention can be instituted to modify metabolic risk factors. Commonly utilized body mass index (BMI) and percentage body fat (PBF) are influenced by lean mass whereas fat mass index (FMI) measures only body fat. This study compares the prevalence of obesity and metabolic risk factors with FMI, BMI and PBF using DXA (dual-energy x-ray absorptiometry).Methods489 women randomly recruited from the electoral roll were stratified into 4 age groups; 40-49, 50-59, 60-69 and 70-79 years from 2000 to 2001. Clinical data and DXA body composition were obtained. Statistical analyses were performed using Medcalc v15 (Ostend, Belgium) with significance level at p = 0.05 (two-tailed).ResultsThere was higher prevalence of obesity using PBF compared to BMI and FMI (p<0.001). This difference was greater from age 50-59 (p<0.05) which may be explained by age-related lean mass loss. PBF over-classified obesity in over 35% of normal and 95% of overweight categories compared to FMI and BMI. BMI has a sensitivity of 78.9% and specificity of 98.3% for obesity using FMI as the standard. BMI under-classified obesity in the overweight category by 14.9% compared to FMI. There was no difference in diabetes, dyslipidemia, hypertension and metabolic syndrome prevalence within the BMI-obesity and FMI-obesity categories (p>0.05).ConclusionPBF classified more obesity than BMI and FMI because of its low pre-determined threshold. The greater difference with PBF compared to BMI and FMI from the 50-59 decade onwards can be attributed to age-related lean mass loss. BMI had the lowest sensitivity for obesity diagnosis. BMI under-classified obesity in the overweight category compared to FMI due to its inability to differentiate lean mass. However, there was no significant difference in the prevalence of metabolic risk factors between BMI and FMI-obesity categories indicating that fat location may influence metabolic dysregulation.