Modeling Obesity in Norway (The MOON Study): A Decision-Analytic Approach—Prevalence, Costs, and Years of Life Lost

Abstract

Background Limited knowledge exists on the expected long-term effects and cost-effectiveness of initiatives aiming to reduce the burden of obesity. Aim To develop a Norwegian obesity-focused disease-simulation model: the MOON model. Material and Methods We developed a Markov model and simulated a Norwegian birth cohort’s movement between the health states “normal weight,”“overweight,”“obese 1,”“obese 2,” and “dead” using a lifetime perspective. Model input was estimated using longitudinal data from health surveys and real-world data (RWD) from local and national registers ( N = 99,348). The model is deterministic and probabilistic and stratified by gender. Model validity was assessed by estimating the cohort’s expected prevalence, health care costs, and mortality related to overweight and obesity. Results Throughout the cohort’s life, the prevalence of overweight increased steadily and stabilized at 45% at 45 y of age. The number of obese 1 and 2 individuals peaked at age 75 y, when 44% of women and 35% of men were obese. The incremental costs per person associated with obesity was highest in older ages and, when accumulated over the lifetime, higher among women (€12,118, €9,495–€15,047) than men (€6,646, €5,252–€10,900). On average, obesity shortened the life expectancy of women/men in the whole cohort by 1.31/1.08 y. The life expectancy for normal-weight women/men at age 30 was 83.31/80.31. The life expectancy was reduced by 1.05/0.65 y if the individual was overweight, obese (2.87/2.71 y), or obese 2 (4.06/4.83 y). Conclusion The high expected prevalence of obesity in the future will lead to substantial health care costs and large losses in life-years. This underscores the need to implement interventions to reduce the burden of obesity; the MOON model will enable economic evaluations for a wide range of interventions.

Registry data for use in health technology assessments in Norway – Opportunities and challenges

“…One possible future direction is to increase collaboration between epidemiologists and HTA researchers, in order to fully integrate information from registers and population studies into the evidence base used for decisions. Some examples exist (51,52), but there is clearly a potential for further interdisciplinary collaborations to bring forth decision relevant information.…”

Novel Insights from Stratified Cost-Effectiveness Analysis of Sugar-Sweetened Beverage Tax Policy in Norway: A Comparative Modeling Study

“…In total, a study population of 55,120 individuals (i.e., the size of the Norwegian birth cohort in 2018 [30]) were simulated from 2 years of age until death or 100 years old, whichever came first. Since the prevenance of obesity and healthcare costs varied markedly by age and sex, males (52% of initial study population) and females (48% of initial study population) were modeled separately along with yearly (i.e., cycle length) time-varying transition probabilities between health states.…”

“…Since the prevenance of obesity and healthcare costs varied markedly by age and sex, males (52% of initial study population) and females (48% of initial study population) were modeled separately along with yearly (i.e., cycle length) time-varying transition probabilities between health states. The underlying model structure was built upon the very recent MOON (Modeling Obesity in Norway) study [30], which estimated transition probabilities between as well as relevant health care costs for each health state, according to longitudinal data from health surveys and Norwegian registers. However, there was no consideration of policy interventions because it centered around extrapolating the secular trend of obesity prevalence from historical data.…”

“…The background mortality rates for each age and gender came from the national life table for Norwegian population in 2017 [38]. As used in the MOON study [30], the age-and BMI-specific mortality rates for each health state were the product of background mortality rates and hazard ratios, which were derived from the Global BMI Mortality collaboration [39].…”

“…Formal healthcare costs consisted of costs pertaining to somatic care, psychiatric care, private specialist, multi-disciplinary drug and alcohol treatment, and general practitioners [30]. Secondly, an extended perspective from the government should take not only formal healthcare costs but also government tax collection costs and SSB tax revenue into account.…”