Summary
Background
Obstructive sleep apnoea is a common disorder that is associated with major public health and economic burden across the USA. Previous studies have assessed the prevalence of the condition. In the present study, we aimed to estimate the burden of obstructive sleep apnoea across the USA from 2020 to 2050, to guide public health policies and management pathways.
Methods
In this prospective modelling study, historical data on obstructive sleep apnoea prevalence in the USA were extracted from a previously published longitudinal cohort study. US population characteristics (age and sex) were obtained from relevant and validated population data sources, and data on BMI were obtained from the National Health and Nutrition Examination Survey and the Wisconsin Sleep Cohort. To project the obstructive sleep apnoea burden (cases and prevalence) into 2050, we developed an open cohort dynamic population simulation model.
Findings
On the basis of projected changes in US age, sex, and BMI population distributions, the model predicts a significant rise in obstructive sleep apnoea over the next three decades. By 2050, the prevalence of obstructive sleep apnoea (apnoea–hypopnoea index ≥5/h) is expected to show a relative increase of 34·7%, from 34·3% (95% uncertainty interval [UI] 34·0–34·4) to 46·2% (46·0–46·4), resulting in 76·6 million cases. We estimate that females will see a larger relative increase than males, with a 65·4% relative increase in prevalence, from 22·8% (22·5–23·0) to 37·7% (37·4–38·0) reaching a total of 30·4 million cases. Males are projected to show a more moderate relative increase of 19·3%, from 45·6% (45·4–46·0) to 54·4% (54·2–54·7), reaching 45·9 million cases.
Interpretation:
Projections indicate that obstructive sleep apnoea will affect 76·6 million adults aged 30–69 years across the USA in 2050, with a disproportionate growth among females compared with males. These findings highlight the urgent need for targeted public health strategies and revised access to diagnosis and follow-up pathways to address the growing prevalence of obstructive sleep apnoea, particularly among females.
Funding
Resmed.
Introduction
Obstructive sleep apnoea is a sleep disorder affecting more than 936 million individuals globally, with approximately 40% having moderate-to-severe obstructive sleep apnoea.1 In the USA, obstructive sleep apnoea affects a large proportion of the population, with estimates suggesting that between 30 million2 and up to 60 million1 adults might have the condition, depending on diagnosis criteria. Obstructive sleep apnoea is associated with symptoms including excessive daytime sleepiness and fatigue and is a well established independent risk factor for cardiovascular disease.3,4 The condition imposes a substantial burden on both health systems and societal costs.5 A white paper by the American Academy of Sleep Medicine estimated that the total societal costs of obstructive sleep apnoea in the USA exceeded US$150 billion annually, including $86·9 billion from lost workplace productivity, $30 billion in increased health-care use, $26·2 billion from motor vehicle crashes, and $6·5 billion due to workplace accidents and injuries.6
Despite this impact, long-term projections are scarce. Accurate forecasts are crucial for resource planning and targeted interventions, especially in the context of an ageing population, lifestyle changes, and increasing obesity. Early intervention is essential, because obstructive sleep apnoea often goes undiagnosed and is linked to cardiovascular, metabolic, and mental health comorbidities.7
Obstructive sleep apnoea treatment involves opening the upper airway with mechanical supports, such as positive airway pressure therapy, which is recognised as the gold standard treatment.8 Obesity is a key reversible risk factor for obstructive sleep apnoea that contributes to airway collapse during sleep. Obesity (BMI ≥30 kg/m2) affects 40% of adults with obstructive sleep apnoea.9 Research suggests that glucagon-like peptide-1 (GLP-1) receptor agonists, such as tirzepatide and semaglutide, might aid weight loss and improve obstructive sleep apnoea outcomes, including reducing the apnoea–hypopnoea index (AHI) and hypoxic burden.10 These medications are increasingly prescribed in the USA because of their effectiveness in promoting weight loss, lowering appetite, and improving insulin sensitivity, all of which can help mitigate obstructive sleep apnoea symptoms.
Alongside rising obesity rates, demographic shifts (especially population ageing) are expected to shape the future burden of obstructive sleep apnoea. As life expectancy continues to increase, the proportion of older adults in the USA will rise,11 with middle and older age being a key risk factor for obstructive sleep apnoea because of the changes in muscle tone, airway structure, and respiratory control that occur with ageing.12 Moreover, although obstructive sleep apnoea increases with age in both males and females, the incidence and severity of the condition differs between males and females, with men being more likely to develop obstructive sleep apnoea at a younger age, whereas the prevalence in women increases significantly after menopause.7,13,14 These age-related and sex-related factors underscore the importance of considering demographic trends and sex-specific strategies when projecting the future burden of obstructive sleep apnoea, ensuring that health-care planning and interventions adequately address the needs of these populations.
To address this knowledge gap, we developed a modelling study to project the 30-year burden of obstructive sleep apnoea in the USA from 2020 to 2050, considering key risk factors such as BMI, age, and sex. The study integrates a comprehensive literature review to estimate obstructive sleep apnoea prevalence, its association with BMI, and the relationship between AHI and BMI. By incorporating therapeutic trends, our analysis provides a more nuanced forecast of the future burden of obstructive sleep apnoea, offering insights into how advancements in obesity management could influence the disease trajectory over the coming decades. Although obstructive sleep apnoea is a global public health concern, this study focuses on the USA because of the availability of high-quality, publicly available population-level data on obstructive sleep apnoea prevalence, risk factors, GLP-1 medication use, and health-care effects, serving as a robust and well characterised setting that provides a foundational structure for future regional and global extensions of this work.
Methods
For this prospective modelling study, we developed a probabilistic simulation model to project obstructive sleep apnoea prevalence over time, incorporating demographic and clinical factors. The model accounted for interactions between age, BMI, and obstructive sleep apnoea risk. Model development and the performance of analyses were done in Python version 3.8.5. A visual of the model and a table containing model inputs and their sources can be found in the appendix (p 2). Bootstrapping was applied to generate 95% uncertainty intervals (UIs) for obstructive sleep apnoea prevalence estimates.
Primary analysis input
Demographic analyses included US-specific population estimates, stratified by sex and age. This information was obtained from the World Bank.15 At each model step, cohort size was determined by the 2020 US fertility rate.16
Obstructive sleep apnoea prevalence was retrieved from a 30-year longitudinal sleep cohort study.17 Given the scarcity of data for obstructive sleep apnoea incidence in the USA, prevalence was used as a proxy to determine incidence by sex and age group. Mortality included all-cause mortality,18 mortality attributable to obstructive sleep apnoea,19,20 and mortality attributable to BMI.21 Mortality was modelled as a function of age and sex.18
BMI was used as a categorisation of weight classes. The initial BMI was sampled at random using the BMI distribution for the USA, by sex and age groups as reported by the National Health and Nutrition Examination Survey Data (NHANES) from 2017 to 2018. Because of data availability, BMI was only assigned to individuals aged 20 years or older. BMI was advanced prospectively according to BMI-category-stratified regression models (appendix p 1), informed by longitudinal data from the Wisconsin Sleep Cohort, accounting for sex, age, and time. BMI trajectories were specific to both initial BMI category and sex, ensuring a more nuanced representation of weight-change patterns.
Obstructive sleep apnoea diagnosis was based age, sex, BMI, and AHI, a metric reflecting the number of apnoea and hypopnoea events per hour of sleep,22 with individuals considered to have obstructive sleep apnoea if their AHI was 5/h or higher. For people diagnosed with obstructive sleep apnoea, the initial AHI was assigned according to a regression model (appendix p 1) informed by the Wisconsin Sleep Cohort,23 adjusting for age, sex, and BMI. People without obstructive sleep apnoea were not assigned an AHI. If AHI was lower than 5/h during the simulation, they were no longer considered to have obstructive sleep apnoea. At each step, AHI was adjusted on the basis of changes in BMI, because AHI is known to correlate with BMI.24
Sensitivity analysis input
To explore the potential effect of shifting obesity trends due to emerging GLP-1 usage, we did a sensitivity analysis incorporating GLP-1 use. This analysis used the same inputs as the primary analysis, with the addition of data specifically regarding GLP-1 usage and access.
Eligibility for GLP-1 was defined as individuals aged 20–65 years with a BMI of 30 kg/m2 or higher. Access to GLP-1 was modelled on the basis of potential health-care-coverage scenarios. Specifically, 70% of eligible males and 80% of eligible females were considered to have access to treatment. These assumptions were based on several factors, including the gradual increase in treatment availability as more indications are approved, and the differences in treatment uptake between sexes. Historically, females tend to be more engaged in seeking health-care treatments, particularly for obesity and metabolic conditions.25 Furthermore, a 5% increase in access with each 5-year time step was implemented throughout the simulations to reflect potential future growth in access and interest in treatment as well as increased health-care coverage in the USA.
Published data indicate a 29·9% persistence rate, defined as the proportion of patients who continue the drug without discontinuation, over 2 years26 when injectable medications were used. In our model, in line with real-world evidence,27 treatment persistence was assumed to be 50%. Given the nature of the drug and increasing availability and interest in the treatment, we assumed that persistence would improve compared with typical oral medication persistence rates. On the basis of clinical trials on GLP-1 and weight loss among people with obstructive sleep apnoea,10 we assumed a 22·5% reduction in BMI as a result of persistent GLP-1 treatment, which could rebound following treatment discontinuation.28
Primary simulation model for prevalence of obstructive sleep apnoea
We built a probabilistic simulation model using a Monte Carlo simulation with 200 random iterations to project obstructive sleep apnoea prevalence and cases up until 2050. In each simulation, 10 000 representative units of the population were randomly initialised with sex, age, and BMI. Sex was randomly assigned with equal probability, and the age and BMI of individuals were sampled from US-specific distributions.
Obstructive sleep apnoea diagnosis status was determined by sampling from a probability distribution constructed with base prevalence, stratified by age and sex, obtained from relevant prevalence studies. The odds of acquiring obstructive sleep apnoea were then adjusted on the basis of BMI, and obstructive sleep apnoea status was sampled from the resulting probability. The simulation moved forward in time steps of 5 years. At each step, age, BMI, obstructive sleep apnoea status, and mortality status were updated according to the associated transition probabilities and distributions. AHI was adjusted relative to changes in BMI, derived from past research on the relationship between BMI and AHI.29
If the person did not have obstructive sleep apnoea, the probability of developing the condition was determined by age and sex, modified by risk ratios associated with the BMI of the participant. Mortality was determined by sex, age, and BMI as the simulation moved forward in time. Among people with obstructive sleep apnoea, the probability of mortality was calculated from obstructive sleep apnoea mortality and all-cause mortality across sex and age. For people without obstructive sleep apnoea, the probability of mortality was calculated using all-cause mortality only. Population growth was accounted for by adding individuals to the population probabilistically, according to the crude fertility rate for the USA.
Simulation model for sensitivity analysis of GLP-1 treatment and obstructive sleep apnoea prevalence
Participants aged 20–65 years with a BMI of 30 kg/m2 or higher were evaluated for GLP-1 treatment eligibility. Of those found eligible, 70% of males and 80% of females were assumed to have access to treatment. Once an individual was deemed eligible, persistent, and to have accessed treatment, they received a one-time BMI reduction of 22·5%. This BMI reduction resulted in an adjustment to their AHI. If a person was determined to be persistent at subsequent steps, their BMI changed according to BMI trajectories provided by the Wisconsin Sleep Cohort. At each step of the simulation, treatment continuation or discontinuation was determined on the basis of probabilities for GLP-1 eligibility, persistence, and accessibility. If treatment was discontinued, the 22·5% BMI reduction was reversed. If GLP-1 treatment was discontinued, the person was assigned a probability to restart the treatment at subsequent steps. On the basis of eligibility and the sex-specific access probabilities, whether a person who was previously on GLP-1 treatment would restart therapy or not was determined.
Role of the funding source
This work represents an academic and industry partnership. The study sponsor was involved in data collection, data analysis, data interpretation, writing of the manuscript, and the decision to submit for publication.
Results
Sex was assigned probabilistically, with a 50% likelihood for male and 50% for female in the simulated population. Projected estimates and their 95% bootstrapped UIs for the primary analysis can be found in table 1 and figure 1. Given that the prevalence estimates of obstructive sleep apnoea were model simulated, 95% bootstrapped UIs were calculated for the prevalence data. By contrast, the case counts were derived from the US population size and simulated prevalence rates. Because the case counts are dependent on population size and the simulated prevalence, UIs were not calculated for case counts. Additionally, the prevalence estimates exhibited relatively narrow UIs because of the large, simulated sample size, which limits the variability observed across bootstrapped samples. This outcome is common when prevalence is based on large-scale population models.
Table 1:
Projected prevalence and number of cases of obstructive sleep apnoea in the USA up to 2050, assuming no access to GLP-1 therapies
| AHI ≥5, total population | AHI ≥15, total population | AHI ≥30, total participants | AHI ≥5, female population | AHI ≥15, female population | AHI ≥30, female population | AHI ≥5, male population | AHI ≥15, male population | AHI ≥30, male population | |
|---|---|---|---|---|---|---|---|---|---|
|
| |||||||||
| Prevalence of obstructive sleep apnoea | |||||||||
| 2020 | 34·3% (34·0– 34·4) | 23·3% (23·1– 23·5) | 11·8% (11·7– 11·9) | 22·8% (22·5– 23·0) | 15·4% (15·1– 15·6) | 7·7% (7·6–7·9) | 45·6% (45·4– 46·0) | 31·2% (31·0– 31·6) | 15·9% (15·7– 16·1) |
| 2025 | 35·6% (35·4– 35·8) | 23·9% (23·8– 24·2) | 12·1% (11·9– 12·3) | 24·6% (24·3– 25·0) | 16·5% (16·3– 16·7) | 8·3% (8·2–8·5) | 46·4% (46·1– 46·9) | 31·4% (31·2– 31·8) | 15·7% (15·6– 16·1) |
| 2030 | 37·3% (37·2– 37·6) | 25·1% (24·9– 25·4) | 12·7% (12·6– 12·9) | 26·8% (26·6– 27·1) | 17·9% (17·7– 18·1) | 9·1 (8·9–9·2) | 47·6% (47·3– 48·0) | 32·1% (31·9– 32·6) | 16·3% (16·0– 16·6) |
| 2035 | 39·1% (38·9– 39·3) | 26·4% (26·2– 26·6) | 13·4% (13·3– 13·6) | 29·1% (28·8– 29·4) | 19·6% (19·3– 19·8) | 9·9 (9·7– 10·0) | 48·8% (48·5– 49·1) | 33·0% (32·8– 33·4) | 16·8% (16·6– 17·2) |
| 2040 | 41·3% (41·0– 41·5) | 28·0% (27·9– 28·3) | 14·4% (14·3– 14·6) | 31·9% (31·7– 32·2) | 21·6% (21·4– 21·8) | 11·0 (10·8– 11·2) | 50·3% (50·0– 50·9) | 34·3% (34·0– 34·8) | 17·7% (17·5– 18·0) |
| 2045 | 43·9% (43·7– 44·0) | 30·0% (29·9– 30·2) | 15·7% (15·5– 15·8) | 35·0% (34·8– 35·3) | 23·8% (23·6– 24·0) | 12·3% (12·1– 12·5) | 52·4% (52·1– 52·7) | 36·1% (35·8– 36·4) | 19·0% (18·7– 19·1) |
| 2050 | 46·2% (46·0– 46·4) | 32·0% (31·8– 32·1) | 17·0% (16·8– 17·1) | 37·7% (37·4– 38·0) | 25·8% (25·5– 26·0) | 13·5% (13·3– 13·7) | 54·4% (54·2– 54·7) | 37·9% (37·6– 38·3) | 20·3% (20·1– 20·5) |
| Percent change from 2020–50 | +34·7% | +37·3% | +44·1% | +65·4% | +67·5% | +75·3% | +19·3% | +21·5% | +27·7% |
| Cases of obstructive sleep apnoea, millions | |||||||||
| 2020 | 56·9 | 38·7 | 19·6 | 18·9 | 12·8 | 6·5 | 37·8 | 26·1 | 13·2 |
| 2025 | 60·7 | 41·0 | 20·6 | 20·9 | 14·0 | 7·1 | 39·8 | 25·9 | 13·7 |
| 2030 | 63·9 | 42·9 | 21·7 | 22·7 | 15·2 | 7·7 | 41·1 | 27·0 | 14·3 |
| 2035 | 66·1 | 44·6 | 22·7 | 24·2 | 16·4 | 8·3 | 41·7 | 28·3 | 14·6 |
| 2040 | 69·4 | 47·1 | 24·3 | 26·3 | 17·9 | 9·2 | 42·9 | 29·2 | 15·3 |
| 2045 | 73·4 | 50·3 | 26·2 | 28·6 | 19·6 | 10·1 | 44·6 | 30·7 | 16·2 |
| 2050 | 76·6 | 52·9 | 28·1 | 30·4 | 21·0 | 11·0 | 45·9 | 31·9 | 17·2 |
| Percent change from 2020–50 | +34·6% | +36·7% | +43·4% | +60·8% | +64·0% | +69·2% | +21·4% | +22·2% | +30·3% |
Data are % (95% uncertainty interval) or n unless otherwise stated. AHI=apnoea–hypopnoea index. GLP-1=glucagon-like peptide-1. AHI ≥5 is inclusive of AHI ≥15 and AHI ≥30.
Figure 1: Projected prevalence and number of obstructive sleep apnoea cases (in millions) in the USA from 2020 to 2050, stratified by AHI severity, without GLP-1 intervention.
AHI=apnoea–hypopnoea index. GLP-1=glucagon-like peptide-1.
Our model projected a substantial increase in the prevalence, number of cases, and severity of obstructive sleep apnoea over the next three decades, with the most notable changes occurring in the moderate (AHI ≥15/h) and severe (AHI ≥30/h) AHI categories. By 2050, with the total US population expected to reach 389 million, obstructive sleep apnoea prevalence across the mild category (AHI ≥5/h) is expected to show a relative increase of 34·7%, from 34·3% (95% UI 34·0–34·4) to 46·2% (46·0–46·4), resulting in a total of 76·6 million cases. Notably the category of people with an AHI of 15/h or higher is projected to show a relative increase of 37·3%, from 23·3% (23·1–23·5) to 32·0% (31·8–32·1), a total of 52·9 million cases, and those with an AHI of 30/h or higher by 44·1%, from 11·8% (11·7–11·9) to 17·0% (16·8–17·1), a total of 28·1 million cases by 2050.
When examining trends by sex, females showed a higher relative increase in all AHI categories. That is, the category of females with an AHI of 5/h or higher is expected to show a relative increase of 65·4%, from 22·8% (22·5–23·0) to 37·7% (37·4–38·0), a total of 30·4 million cases by 2050, whereas those with an AHI of 15/h or higher are projected to rise by 67·5%, from 15·4% (15·1–15·6) to 25·8% (25·5–26·0), a total of 21·0 million cases, and those with an AHI of 30/h or higher by 75·3%, from 7·7% (7·6–7·9) to 13·5% (13·3–13·7), a total of 11.0 million cases up until 2050. By contrast, males exhibit more moderate increases across the same categories. The projected rise for males in the category with an AHI of 5/h or higher is 19·3%, from 45·6% (45·4–46·0) to 54·4% (54·2–54·7), a total of 45·9 million cases, for those with an AHI of 15/h or higher it is a 21·5% relative increase from 31·2% (31·0–31·6) to 37·9% (37·6–38·3), a total of 31·9 million cases, and for those with an AHI of 30/h or higher it is a 27·7% relative increase, from 15·9% (15·7–16·1) to 20·3% (20·1–20·5), a total of 17·2 million cases into 2050. Overall, these findings underscore a significant upward trend in the severity of obstructive sleep apnoea, with females particularly affected by the growing burden of the disease through 2050.
Projected estimates and their 95% bootstrapped UIs for the sensitivity analysis can be found in table 2 and figure 2. When accounting for GLP-1s, similar trends were observed. However, compared with the analysis without GLP-1s, the increases in the prevalence and severity of obstructive sleep apnoea over the next three decades were more modest. The projections continue to show a significant rise in the prevalence, cases, and severity of obstructive sleep apnoea up until 2050, particularly in the higher AHI categories (≥15/h and ≥30/h). By 2050, the prevalence of obstructive sleep apnoea in people with an AHI of 5/h or higher is expected to show a relative increase of 29·7%, from 34·3% (34·1–34·5) to 44·5% (44·3–44·7), a total of 73·7 million cases. People with an AHI of 15/h or higher are expected to show a relative increase of 32·2%, from 23·3% (23·2–23·5) to 30·8% (30·6–31·0), a total of 51·0 million cases, and those with an AHI of 30/h or higher of 38·1%, from 11·8% (11·7–12·0) to 16·3% (16·2–16·5), a total of 27·0 million cases.
Table 2:
Projected prevalence and number of cases of obstructive sleep apnoea in the USA up to 2050, assuming population-level access to GLP-1 therapies
| AHI ≥5, total population | AHI ≥15, total population | AHI ≥30, total participants | AHI ≥5, female population | AHI ≥15, female population | AHI ≥30, female population | AHI ≥5, male population | AHI ≥15, male population | AHI ≥30, male population | |
|---|---|---|---|---|---|---|---|---|---|
|
| |||||||||
| Prevalence of obstructive sleep apnoea | |||||||||
| 2020 | 34·3% (34·1–34·5) | 23·3% (23·2–23·5) | 11·8% (11·7–12·0) | 22·8% (22·7–23·1) | 15·4% (15·2–15·6) | 7·7% (7·6–7·9) | 45·6% (45·4–45·9) | 31·2% (31·0–31·5) | 15·9% (15·7–16·1) |
| 2025 | 34·2% (34·0–34·3%) | 22·9% (22·8–23·1) | 11·6% (11·5–11·7) | 23·0% (22·8–23·2) | 15·3% (15·1–15·5) | 7·8% (7·6–7·9) | 45·1% (44·8–45·4) | 30·4% (30·2–30·7) | 15·4% (15·2–15·5) |
| 2030 | 34·8% (34·7–35·0) | 23·3% (23·2–23·5) | 11·8% (11·7–12·0) | 24·0% (23·8–24·2) | 16·0% (15·8–16·2) | 8·1% (7·9–8·2) | 45·4% (45·2–45·7) | 30·5% (30·3–30·8) | 15·5% (15·3–15·7) |
| 2035 | 36·3% (36·1–36·5) | 24·5% (24·3–24·7) | 12·6% (12·5–12·7) | 25·8% (25·5–26·1) | 17·3% (17·1–17·5) | 8·9% (8·7–9·0) | 46·4% (46·2–46·8) | 31·5% (31·3–31·8) | 16·3% (61·1–16·5) |
| 2040 | 38·4% (38·3–38·6) | 26·2% (26·0–26·4) | 13·7% (13·5–13·8) | 28·7% (28·4–28·9) | 19·4% (19·2–19·6) | 10·0% (9·8–10·1) | 48·0% (47·7–48·2) | 32·8% (32·6–33·1) | 17·3% (17·1–17·5) |
| 2045 | 41·3% (41·2–41·5) | 28·4% (28·2–28·5) | 14·9% (14·7–15·0) | 32·1% (31·9–32·4) | 21·8% (21·6–22·1) | 11·3% (11·2–11·5) | 50·3% (50·0–50·6) | 34·7% (34·4–35·0) | 18·3% (18·1–18·5) |
| 2050 | 44·5% (44·3–44·7) | 30·8% (30·6–31·0) | 16·3% (16·2–16·5) | 35·7% (35·4–36·0) | 24·4% (24·2–24·6) | 12·7% (12·5–12·9) | 53·0% (52·7–53·2) | 37·0% (36·7–37·2) | 19·8% (19·6–20·0) |
| Percent change from 2020–50 | +29·7% | +32·2% | +38·1% | +56·6% | +58·4% | +64·9% | +16·2% | +18·6% | +24·5% |
| Cases of obstructive sleep apnoea, millions | |||||||||
| 2020 | 56·9 | 38·7 | 19·7 | 18·9 | 12·7 | 6·4 | 38·0 | 26·0 | 13·3 |
| 2025 | 58·2 | 39·1 | 19·8 | 19·4 | 12·9 | 6·6 | 38·9 | 26·2 | 13·2 |
| 2030 | 59·5 | 39·9 | 20·2 | 20·3 | 13·5 | 6·8 | 39·3 | 26·4 | 13·4 |
| 2035 | 61·2 | 41·6 | 21·3 | 21·5 | 14·4 | 7·4 | 39·7 | 26·9 | 13·9 |
| 2040 | 64·7 | 44·1 | 23·0 | 23·8 | 16·1 | 8·3 | 40·9 | 28·0 | 14·7 |
| 2045 | 69·1 | 47·4 | 24·9 | 26·5 | 18·0 | 9·3 | 42·7 | 29·4 | 15·6 |
| 2050 | 73·7 | 51·0 | 27·0 | 29·1 | 19·9 | 10·4 | 44·6 | 31·2 | 16·7 |
| Percent change from 2020–50 | +29·5% | +31·8% | +37·1% | +54·0% | +56·7% | +62·5% | +17·4% | +20·0% | +25·6% |
Data are % (95% uncertainty interval) or n unless otherwise stated. AHI=apnoea–hypopnoea index. GLP-1=glucagon-like peptide-1. AHI ≥5 is inclusive of AHI ≥15 and AHI ≥30.
Figure 2: Projected prevalence and number of obstructive sleep apnoea cases (in millions) in the USA from 2020 to 2050, stratified by AHI severity, assuming the introduction of GLP-1 therapy.
AHI=apnoea–hypopnoea index. GLP-1=glucagon-like peptide-1.
When considering sex-specific trends, females were projected to have a more pronounced increase across all AHI categories, even with GLP-1 access. Specifically, females with an AHI of 5/h or higher were expected to show a relative increase of 56·6%, from 22·8% (22·7–23·1) to 35·7% (35·4–36·0), a total of 29·1 million cases by 2050. Females with an AHI of 15/h or higher were projected to show a relative increase of 58·4%, from 15·4% (15·2–15·6) to 24·4% (24·2–24·6), a total of 19·9 million cases, and those with an AHI of 30/h or higher to increase by 64·9%, from 7·7% (7·6–7·9) to 12·7% (12·5–12·9), a total of 10·4 million cases. By comparison, the increase in obstructive sleep apnoea prevalence and cases among males was more moderate. Males with an AHI of 5/h higher were projected to show an increase of 16·2%, from 45·6% (45·4–45·9) to 53·0% (52·7–53·2), a total of 44·6 million cases, whereas those with an AHI of 15/h or higher were expected to show an increase of 18·6%, from 31·2% (31·0–31·5) to 37·0% (36·7–37·2), a total of 31·2 million cases, and those with an AHI of 30/h or higher by 24·5%, from 15·9% (15·7–16·1) to 19·8% (19·6–20·0), a total of 16·7 million cases by 2050. Overall, by 2050, estimates with GLP-1 access suggest slightly lower prevalence than projections without GLP-1s, with differences of 3·7% for people with an AHI of 5/h or higher, 3·8% for those with an AHI of 15/h or higher, and 4·1% for those with an AHI of 30/h or higher.
Discussion
This study projects the 30-year burden of obstructive sleep apnoea in the USA from 2020 to 2050, considering risk factors such as age, sex, and BMI, and accounting for use of GLP-1s. The findings of our primary model indicate a significant projected increase in the prevalence, number of cases, and severity of obstructive sleep apnoea over the next three decades. By 2050, the prevalence of obstructive sleep apnoea in people with AHI of 5/h or higher is expected to show a relative increase of 34·7%, from 34·3% (34·0–34·4) to 46·2% (46·0–46·4), reaching a total of 76·6 million cases. The projected growing obstructive sleep apnoea burden might have substantial wider clinical consequences, including higher incidence of cardiovascular and metabolic diseases, resistant hypertension, and potential effects on reproductive health, such as fertility issues in females.4,30
Females are expected to have a higher increase in obstructive sleep apnoea prevalence across all AHI categories than males. Specifically, females with an AHI higher than or equal to 5/h is projected to show a relative increase of 65·4%, from 22·8% (22·5–23·0) to 37·7% (37·4–38·0), rising to a total of 30·4 million cases. By contrast, males are projected to have more moderate increases, with those with an AHI of 5/h or higher rising by 19·3%, from 45·6% (45·4–46·0) to 54·4% (54·2–54·7), rising to a total of 45·9 million cases by 2050.
Accounting for GLP-1s, the results also demonstrate continued increase in obstructive sleep apnoea and growing severity of the condition over the next few decades. Despite the potential benefits of GLP-1s in managing obesity and metabolic disorders, both of which are risk factors for obstructive sleep apnoea, the overall trend remains largely unchanged. Females, in particular, are expected to bear a disproportionately heavy burden of the disease, with a more pronounced rise in both the prevalence and severity of obstructive sleep apnoea compared with males. This finding suggests that even as GLP-1 therapies might help mitigate some of the underlying risk factors for obstructive sleep apnoea, the broader demographic shifts, including ageing populations and rising obesity prevalence, will probably continue to drive the overall increase in obstructive sleep apnoea cases, with females particularly affected. The projected higher increase in prevalence among females can also be attributed to several factors beyond BMI, such as hormonal changes during perimenopause and after menopause, smaller airway size, and sex-specific symptom presentation, leading to underdiagnosis.31
Given that BMI is an important predictor of obstructive sleep apnoea prevalence in our model, we compared our obesity projections to those recently published,32 reporting obesity trends across the USA until 2050. Both studies predict significant increases in obesity over the coming decades. In 2021, the prevalence of overweight and obesity exceeded 40% across all states for both sexes combined. The recently published study estimates that by 2050, 58·8% of females will have obesity, whereas for males, 45 states are expected to have obesity surpassing 50·0%. In our study, we report a slightly lower obesity prevalence of 39·6% in 2020 overall for both sexes. However, our 2050 projections suggest a higher prevalence among females, with an estimated 68·7% of females expected to have obesity, compared with 53·4% of males. The higher obesity prevalence among females in our study reflects trends in the Wisconsin Sleep Cohort, where the average BMI in the highest category is higher than in the general US population.
Additionally, The Lancet Commission on obesity33 has highlighted that lower BMI thresholds, combined with symptoms (such as obstructive sleep apnoea), might be more important than BMI alone in determining obesity-related health risks. If these new standards are adopted, they could further refine our understanding of the role of obesity in the prevalence of obstructive sleep apnoea.
In terms of methods, we used a simulation to estimate the future burden of obstructive sleep apnoea. A key strength of this approach is its capacity to produce granular, sex-specific projections under varying scenarios, such as the inclusion of GLP-1s. Additionally, this model structure allows for transparent sensitivity analyses and scenario testing.
We acknowledge several limitations. First, as with any modelling work, we were reliant on assumptions and published data used in the modelling. However, we did a sensitivity analysis, and we were generally conservative in our parameter choices, for example by selecting lower-bound effect sizes, and avoiding overly aggressive estimates that could inflate prevalence projections. Second, we relied on BMI as the measure of obesity and as primary predictor, given that it is a major reversible risk factor for obstructive sleep apnoea. Beyond BMI, we also recognise that factors such as diabetes, smoking, and air pollution can significantly influence obstructive sleep apnoea prevalence. For example, smoking can exacerbate obstructive sleep apnoea severity,34 yet it is not directly included in the base model assumptions. Furthermore, additional clinical factors, including physical comorbidities (eg, cardiovascular disease and diabetes), insomnia, depression, anxiety, and sleep parameters (eg, sleep duration) can affect obstructive sleep apnoea burden estimates. These factors were not incorporated because of the increased complexity they would introduce, including the need for granular data and more assumptions. Adding such features could reduce model transparency and divert from the primary goal of producing a clear, policy-relevant projection based on widely available risk factors. Nonetheless, we recognise their importance and encourage future modelling efforts to explore their role using more detailed datasets. Third, our study was limited by the absence of race-specific data. Although racial and ethnic disparities in obstructive sleep apnoea prevalence and severity are well documented, most available data report aggregated figures without sufficient granularity for use in dynamic population models. As such, our projections did not account for race-based variations in obstructive sleep apnoea burden. This is a notable limitation, because racial and ethnic differences in obesity prevalence, access to diagnosis, treatment uptake, and socioeconomic determinants could substantially influence both baseline obstructive sleep apnoea prevalence and future trends. We encourage future modelling efforts to incorporate race and ethnicity explicitly, using more detailed, disaggregated datasets where available, to improve precision and equity in burden estimation and health planning. Fourth, although we anticipate a significant rise in obstructive sleep apnoea among females, we recognise that GLP-1s might lead to greater weight loss in women than in men. As real-world data accumulate, our understanding of the long-term effects of GLP-1s on obstructive sleep apnoea will continue to evolve. We encourage further research in this area, including the sex-specific effects of obesity. Fifth, we recognise that other factors might change over time including air pollution and smoking prevalence, which could affect obstructive sleep apnoea incidence. Sixth, our model is based on US data and assumptions, limiting its generalisability to other countries because of differences in population structure, risk factors, health-care access, and diagnostic practices. Finally, although we primarily used NHANES and the Wisconsin Sleep Cohort to estimate the future obstructive sleep apnoea burden, we acknowledge the potential value of other datasets such as the Sleep Heart Health Study, MrOS Sleep, and the Multi-Ethnic Study of Atherosclerosis. Future research efforts could leverage these cohorts to validate or expand our projections.
This study highlights a substantial upward trend in both the prevalence and severity of obstructive sleep apnoea, with females expected to bear a relatively larger increase in the coming decades. The findings suggest that demographic factors, including sex, ageing, population growth, and obesity will continue to drive the rising prevalence of obstructive sleep apnoea, even with advancing anti-obesity medications.
As obstructive sleep apnoea prevalence rises, especially among females, policy makers should prioritise access to early screening, affordable treatment options, and targeted obesity prevention. Integrating obstructive sleep apnoea into broader chronic disease management plans is essential to mitigate the future health-care burden. Although this study focuses on the USA, it is important to note that the burden of obstructive sleep apnoea is rising globally, driven by similar trends in obesity, ageing populations, and underdiagnosis, underscoring the need for coordinated international public health strategies.
Supplementary Material
Research in context.
Evidence before this study
Obstructive sleep apnoea is a prevalent sleep-related breathing disorder with serious health and economic consequences. Most previous studies have focused on estimating the prevalence of the condition. Although these efforts have helped to quantify the present-day burden, they fall short in projecting how obstructive sleep apnoea prevalence might evolve in response to shifting population dynamics, such as ageing and rising obesity prevalence. To inform long-term health planning, we identified a need for future-oriented modelling studies that integrate demographic and metabolic risk trends in the US population. We searched PubMed and Embase for articles published in English language from inception using the terms “adult”, “sleep disordered breathing”, “obstructive sleep apnoea”, “prevalence”, “population”, and “United States” to identify studies on obstructive sleep apnoea and its burden. The search was done in March 14, 2024, and rechecked in Feb 14, 2025, with no new studies found. To support the modelling inputs, we also searched for US population characteristics, including BMI distribution, apnoea–hypopnoea index (AHI) severity, obstructive sleep apnoea mortality, BMI-related mortality, fertility, and all-cause mortality.
Added value of this study
To the best of our knowledge, our study is the first to project the future burden of obstructive sleep apnoea in the USA up until 2050 using an open-cohort dynamic population model informed by data on age, sex, and BMI. By simulating a representative US population aged 30–69 years, we estimate that the total number of adults with obstructive sleep apnoea (AHI ≥5/h) will grow by over a third, reaching 76·6 million cases by 2050. Notably, although males will continue to represent the majority of cases, females are projected to have a disproportionately larger relative increase in burden, rising by 65·4% over the next 30 years. These projections reflect changes in underlying risk factors and demographic shifts.
Implications of all the available evidence
Given the expected surge in obstructive sleep apnoea prevalence, particularly among females, public health systems must prepare for substantial increases in diagnostic demand, clinical care, and long-term management. Strategic planning should focus not only on improving access to diagnosis and treatment but also on addressing modifiable risk factors such as obesity.
Acknowledgments
This study was funded by Resmed.
EB, MAB, AVB, LK, JA, KLS, and CMN are employees of Resmed. PAC reports having an appointment to an endowed academic Chair at the University of Sydney that was established from Resmed funding, has received research support from Resmed and SomnoMed, and is a consultant to Resmed, SomnoMed, Sunrise Medical, and Eli Lilly. J-LP is supported by the French National Research Agency in the framework of the Investissements d’Avenir program (ANR-15-IDEX-02) and the e-Health and Integrated Care and Trajectories Medicine and MIAI Artificial Intelligence chairs of excellence from the Grenoble Alpes University Foundation (ANR-19-P3IA-0003). J-LP also reports lecture fees or conference travel grants from Resmed, Philips, Jazz Pharmaceuticals, Agiradom, and Bioprojet. AM is funded by the National Institutes of Health and reports income related to medical education from LivaNova, Jazz Pharmaceuticals, ZOLL, and Eli Lilly. Resmed provided a philanthropic donation to University of California San Diego, San Diego, CA, USA, but AM has not received personal income from ResMed or medXcloud.
Footnotes
Declaration of interests
All other authors declare no competing interests.
Data sharing
The majority of the data included in this study are available in the public domain. Specific requests or questions should be submitted to the corresponding author for consideration.
References
- 1.Benjafield AV, Ayas NT, Eastwood PR, et al. Estimating the global burden of obstructive sleep apnea and the economic costs associated with untreated disease. Sleep 2019; 42: zsy233.30476248 [Google Scholar]
- 2.Peppard PE, Young T, Barnet JH, et al. Increased prevalence of sleep-disordered breathing in adults. Am J Epidemiol 2013; 177: 1006–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Jordan AS, McSharry DG, Malhotra A. Adult obstructive sleep apnoea. Lancet 2014; 383: 736–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Veasey SC, Rosen IM. Obstructive sleep apnea: diagnosis, mechanisms, and management. Clin Chest Med 2019; 40: 555–63. [Google Scholar]
- 5.Mattila S, Sillanpää S, Heikkinen K, et al. Economic burden of obstructive sleep apnea in the United States: a systematic review. J Sleep Res 2022; 31: e13680.35811092 [Google Scholar]
- 6.Frost & Sullivan. The economic burden of obstructive sleep apnea in the US American Academy of Sleep Medicine. 2016. https://aasm.org/resources/pdf/sleep-apnea-economic-crisis.pdf (accessed Feb 15, 2024). [Google Scholar]
- 7.Senaratna CV, Perret JL, Lowe AJ, et al. The prevalence of obstructive sleep apnea in the general population: a systematic review. Sleep 2017; 40: zsx069. [DOI] [PubMed] [Google Scholar]
- 8.Patil SP, Indu AA, Sean M, et al. Treatment of adult obstructive sleep apnea with positive airway pressure: an American Academy of Sleep Medicine clinical practice guideline. J Sleep Med 2019; 15: 335–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Esmaili A, Doroudgar S, Abbasinia M, et al. Impact of obesity on the development of obstructive sleep apnea: a review of the literature. J Obes Metab Syndr 2024; 33: 1–12.38281733 [Google Scholar]
- 10.Malhotra A, Grunstein RR, Fietze I, et al. Tirzepatide for the treatment of obstructive sleep apnea and obesity. New Engl J Med 2024; 391: 1193–205. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Arias E, Tejada-Vera B, Kochanek KD. Projections of the population of the United States by age, sex, race, and Hispanic origin: 2020 to 2060. Natl Vital Stat Rep 2022; 71: 1–39. [Google Scholar]
- 12.Eckert DJ, White DP. Mechanisms of upper airway obstruction in obstructive sleep apnea. J Appl Physiol 2008; 105: 1–17.18450975 [Google Scholar]
- 13.Franklin KA, Lindberg E, Larkin EK, et al. Obstructive sleep apnoea and hypertension: a systematic review and meta-analysis. J Hypertens 2015; 33: 1–10.25426569 [Google Scholar]
- 14.Lin HJ, Huang YS, Hsu CY, et al. The effect of gender on the development of obstructive sleep apnea: a population-based study. Sleep 2008; 31: 775–81. [Google Scholar]
- 15.Bank World. Population estimations and projections. Data catalog. 2024. https://data.worldbank.org (accessed Feb 15, 2024). [Google Scholar]
- 16.Osterman MJ, Hamilton BE, Martin JA, Driscoll AK, Valenzuela CP. Births: final data for 2020. https://stacks.cdc.gov/view/cdc/112078 (accessed Feb 15, 2024). [PubMed] [Google Scholar]
- 17.Peppard PE, Young T, Barnet JH, et al. Increased prevalence of sleep-disordered breathing in adults. Am J Epidemiol 2013; 177: 1006–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.WHO. Mortality database. World Health Organization, 2024. [Google Scholar]
- 19.Marshall NS, Wong KK, Cullen SR, et al. Sleep apnea and mortality: a meta-analysis. Sleep 2008; 31: 1076–82. [Google Scholar]
- 20.Marshall NS, Knuiman MW, Wang J, et al. Sleep apnea and long-term cardiovascular mortality: a population-based study. Sleep 2014; 37: 527–33. [Google Scholar]
- 21.Visaria V, Muralidharan A, Kapoor N. The role of obesity in the mortality risk associated with sleep apnea. J Clin Sleep Med 2023; 19: 657–66. [Google Scholar]
- 22.Epstein LJ, Kristo D, Strollo PJ, et al. (2009). Clinical guideline for the evaluation, management and long-term care of obstructive sleep apnea in adults. J Clin Sleep Med, 5: 263–76. [PMC free article] [PubMed] [Google Scholar]
- 23.Young T, Peppard PE, Gottlieb DJ. Epidemiology of obstructive sleep apnea: a population health perspective. Am J Respir Crit Care Med 2002; 165: 1217–39. [DOI] [PubMed] [Google Scholar]
- 24.Malhotra A, Heilmann CR, Banerjee KK, Dunn JP, Bunck MC, Bednarik J. Weight reduction and the impact on apnea-hypopnea index: a systematic meta-analysis. Sleep Med 2024; 121: 26–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Cooper AJ, Gupta SR, Moustafa AF, Chao AM. Sex/gender differences in obesity prevalence, comorbidities, and treatment. Curr Obesity Rep 2021; 2: 1–9. [DOI] [PubMed] [Google Scholar]
- 26.Weiss T, Carr RD, Pal S, et al. Real-world adherence and discontinuation of glucagon-like peptide-1 receptor agonists therapy in type 2 diabetes mellitus patients in the United States. Patient Pref Adher 2020; 27: 2337–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Lee DS, Lee H. Adherence and persistence rates of major antidiabetic medications: a review. Diabetol Metab Syndr 2022; 14: 12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Iqbal J, Wu HX, Hu N, et al. Effect of glucagon-like peptide-1 receptor agonists on body weight in adults with obesity without diabetes mellitus—a systematic review and meta-analysis of randomized control trials. Obesity Rev 2022; 23: e13435. [DOI] [PubMed] [Google Scholar]
- 29.Malhotra A, Szalewicz P, McGrath S, et al. GLP-1 receptor agonists in sleep apnea and metabolic disorders. Lancet Respir Med 2024; 12: 202–09. [Google Scholar]
- 30.Lim ZW, Wang ID, Wang P, et al. Obstructive sleep apnea increases risk of female infertility: a 14-year nationwide population-based study. PLoS One 2021; 16: e0260842. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Krishnan V, Collop NA. 14 gender differences in obstructive sleep apnea. Obstr Sleep Apnea 2007; 15: 247. [Google Scholar]
- 32.Ng M, Fleming T, Robinson M, et al. Global obesity prevalence and trends. Lancet 2024; 403: 1519–30.38554723 [Google Scholar]
- 33.Rubino F, Glickman M, Iannelli A, et al. Obesity: epidemiology, comorbidities, and societal burden. Lancet 2025; 386: 1656–73. [Google Scholar]
- 34.Krishan J, Patel AD, Malhotra A. The effects of smoking on sleep-disordered breathing and cardiovascular risks. Chest 2014; 146: 684–92. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.