Skip to main content

Research · 2026

COPD in the US and China: who, where, and how many

The same disease looks very different in each market. This brief brings together the published epidemiology for both countries — patient counts, who has the disease, where they live, and how the numbers are changing — and states clearly which figures come from surveys and which are estimates.

Headline figures for both markets

China

~100M

patients by spirometry — 13.7% of adults aged 40 and over [1]

China

1.29M

deaths a year — the third leading cause of death [3]

USA

11–24M

11.1M diagnosed adults; about 24M with airflow obstruction [16][20]

USA

145,643

deaths in 2024 — the fifth leading cause of death [49]

Explore the data

Who, where, and how many.

Pick a country, then click any bar or slice to filter the other three charts to that group. Charts show published survey data until a filter is applied; from then on, they are labeled as estimates.

Country

Adults aged 40 and over · 95.0M patients · 13.7% prevalence on the 2020 census

Select a bar or slice to filter the other charts.

By age

Published data

Patients in each age band, with the prevalence rate underneath.

  • An additional ~8M patients aged 20–39 fall outside this basis.
  • Percentages are always the published rate for that group. Only the patient counts change when you filter.

By sex

Published data

How the patient population splits.

95.0Mpatients

By region

Published data

The ten largest estimated patient pools, launch market first.

  • Top 10 of 13 surveyed regions shown, ranked by estimated patients.
  • Selected provinces and states only — these are the surveyed high- and low-burden ends, not the full national picture.

Prevalent cases over time

Published data

GBD 2021 cycle, 1990 to 2021, with a modeled projection to 2050 — millions of people living with COPD.

0M22M44M66M88M110M199020052021203020402050≈58M51M99.9M · Spirometry survey
  • Solid line: GBD 2021, the latest published China estimate (GBD 2023 is released, but China tables are not yet public). Dashed line with band: modeled projection — cases scaled by the UN-projected 40-and-over population (WPP 2024) under the two published 2050 prevalence paths (age-standardized 2,004 vs 2,793 per 100k); the band is their disagreement.

Definitions

Two valid counts per country.

Both markets circulate two very different patient counts. Neither is wrong; they measure different things, and mixing them in one calculation produces misleading results.

MarketEstimateBasisWhat it measures
China99.9M8.6% of adults 20+ · 13.7% of adults 40+Airflow obstruction on post-bronchodilator spirometry, national field survey 2012–15 (n=50,991). Over half are GOLD I and mostly without symptoms. This is the basis of the official “nearly 100 million” (近1亿) framing. [1]
China50.6M3,556 per 100,000GBD-modeled symptomatic, severity-weighted cases — 23.7% of the world total, with about 4.4M new cases a year. [3]
USA11.1–11.7MBRFSS 2023 · NHIS 2022Self-reported, physician-diagnosed COPD among adults 18 and over. The 11.7M figure is the crude "ever told" prevalence. [16]
USA~24MSpirometry (NHANES)Adults with measurable airflow obstruction. Fewer than half carry a diagnosis; in one 30,874-person cohort, 81% of persistent obstruction was undiagnosed. [20]

For market sizing and materials shared with Chinese partners, the official figure of nearly 100M is the standard reference. For a serviceable-market denominator — patients who perceive a need — the GBD figure of ~50M is the more appropriate one. Figures from different instruments should never be combined in a single calculation.

The care gap

A guideline-recommended program that few patients reach.

Pulmonary rehabilitation is a guideline-recommended part of COPD care in both countries. Access to it is the constraint.

2.7%

of Medicare COPD hospitalizations in the US start pulmonary rehabilitation within twelve months [23]

1,776

US counties with no hospital-based rehabilitation program at all [24]

27%

of Chinese hospitals have ever delivered pulmonary rehabilitation [14]

~10%

patient awareness in China; only 12% have ever had spirometry, against a Healthy China 2030 target of 30% [13][1]

How Aera addresses this

Designed to address each gap.

The epidemiology defines the problem precisely: an enormous, aging, rural-skewed patient population; a guideline-recommended program that reaches 2.7% of eligible US patients and is offered by 27% of Chinese hospitals; a diagnosis funnel that loses half to nine-tenths of patients before care begins; and a comorbidity burden that makes every missed warning sign consequential. Aera is designed to address each of these.

The problem, from the dataWhat Aera does
Rehabilitation deserts. 1,776 US counties have no program and more than 14M people live over an hour from one; 73% of Chinese hospitals have never delivered rehabilitation. [24][14]Delivers an eight-week, guideline-informed program at home on an ordinary smartphone — structured exercise with video coaching, breathing training with paced audio, and progress tracking — without travel to a facility or a waiting list.
Clinicians cannot see between visits. Symptom shifts, adherence and decline are invisible until the next appointment. [40]A 60-second daily check-in feeds a risk engine that watches symptom shifts, oxygen trend and rescue-inhaler use against defined thresholds, and escalates along a defined ladder — from self-care guidance, to “contact your care team today,” to emergency instructions.
Patients do not know they have it. About 10% awareness in China, half undiagnosed in the US, and only 12% of Chinese patients have ever had spirometry. [13][20][1]Makes education central to the program. Each instruction explains the reason behind it, supported by short daily lessons, knowledge checks, and a structured path. This also gives a newly screened patient a place to start, and it supports the Healthy China 2030 awareness target.
The cohort is 60 to 89, rural, and lower income. Cases peak at 70–74 in China; US prevalence below the poverty line is four times the top bracket. [5][17]Designed for older adults first: large text, one action per screen, voice input, illustrations instead of jargon, multiple languages, and family sharing so adult children — who often coordinate care — can follow along. It runs on the phone the patient already owns.
COPD rarely travels alone. Depression runs at 34.5%, about 16% have undiagnosed reduced heart function, and comorbidities account for 51% of COPD healthcare costs. [34][31][33]Whole-person monitoring is built in: validated PHQ-9 and GAD-7 mood screening with escalation to a crisis line (the 988 Suicide & Crisis Lifeline in the U.S.), cardiac-signal rules in the risk engine, panic-episode support with guided breathing recovery, and adverse-event reporting — all on one clinician timeline.
The program has to be financially sustainable within existing systems, in two very different markets.The clinician portal organizes the days between visits: risk-tiered patient panels, longitudinal assessments, secure messaging with photo triage, drafted encounter notes, and engagement data. In the U.S., it supports the documentation practices need for existing Remote Therapeutic Monitoring codes (CPT 98975–81); in China, it gives a hospital a procurement-ready pulmonary rehabilitation (肺康复) service line.

Addressable market

Where the patients are.

The same epidemiology, read as a funnel. Each step narrows the population to the patients a program like this can reach today, and names the channel that reaches them. Only the top line of each funnel is a published count; every step below it is published rates carried onto a smaller pool.

Modeled estimates

中国 China

Adults aged 40 and over

  1. TAM

    All patients

    95.0M

    13.7% prevalence applied to the 693.8M adults aged 40 and over in the 2020 census.

    Channel: The full clinical need the program addresses.

    Published, derived

    [1]

  2. Aware or diagnosed today

    ≈9.5M

    About 10% of Chinese patients know they have the disease.

    Channel: Hospital screening programs plus the Healthy China 2030 awareness target of 30% or more would lift this to ≈28.5M, a favorable policy trend.

    Modeled estimate

    [13]

  3. Aware and rehab-priority severity

    ≈4.2M

    GOLD II or worse is 44% of patients, applied to the aware pool.

    Channel: Pulmonology-department referral — where the procurement conversation happens.

    Modeled estimate

    [2][13]

  4. SOM

    Launch market: Zhejiang / Hangzhou

    ≈400k

    Zhejiang is surveyed at 12.34% prevalence; the same ~10% awareness rate is applied to the province and to Hangzhou.

    Channel: Starting with tier-3 hospital partnerships in Hangzhou.

    Zhejiang patients
    4.0M
    Zhejiang aware
    ≈400k
    Hangzhou patients
    ≈760k
    Hangzhou aware
    ≈76k
    Modeled estimate

    [41][44][13]

USA

Adults aged 18 and over

  1. TAM

    Airflow obstruction

    ≈24M

    Adults with measurable airflow obstruction on spirometry. About half carry no diagnosis.

    Channel: The full clinical need — reachable only once a diagnosis exists.

    Published data

    [20]

  2. Diagnosed

    11.1M

    Self-reported, physician-diagnosed COPD among adults 18 and over, BRFSS 2023.

    Channel: Pulmonology and primary-care panels — patients who already have a COPD diagnosis code.

    Published data

    [16]

  3. Medicare age (reimbursement channel)

    ≈4.8–6.1M

    The 65-and-over share of diagnosed patients, computed from the age bands: 43% of the 11.1M diagnosed is 4.8M, while the bands themselves sum to 6.1M on the 14.2M BRFSS 2021 anchor.

    Channel: The Remote Therapeutic Monitoring reimbursement channel (CPT 98975–81).

    Modeled estimate

    [16][27]

  4. SOM

    Launch market: California

    ≈0.5M

    1.22M diagnosed patients at 4.0% prevalence; ≈0.5M of them are 65 and over, from the national 65-and-over share applied to the state count.

    Channel: Post-hospitalization referral is the most immediate entry point; nationally, only 2.7% of Medicare COPD hospitalizations are followed by pulmonary rehabilitation within 12 months.

    California diagnosed
    1.22M
    California 65 and over
    ≈0.5M
    Modeled estimate

    [18][23]

Bar widths are on a square-root scale. Linear would render the launch-market bar at under half a percent of the top bar — invisible — and a log scale would flatten the ten-fold drop the funnel exists to show. The counts carry the magnitudes; the bars carry the shape.

Assumptions

  • The awareness rate is applied uniformly across geographies. Awareness almost certainly differs between a tier-3 city and a rural county; no source publishes it at that resolution.
  • The severity distribution is applied independently of awareness. The correlation between the two is unknown, and this is likely conservative — diagnosed patients skew more severe than the undiagnosed population.
  • Medicare age is used as a proxy for reimbursement eligibility. Being 65 or over is neither necessary nor sufficient for RTM billing; it is the closest published cut.
  • No adoption or penetration rate is assumed anywhere. These are population ceilings, not forecasts.

None of this is a revenue forecast; it is the population arithmetic behind where Aera starts.

Methodology

Methodology and modeled estimates.

Every figure in the unfiltered charts is a published marginal — a number that appears in one of the cited sources, or a clearly flagged interpolation between two published bands. The sources report age, sex and geography separately. None of them publishes the joint distribution, so there is no surveyed figure for, say, women aged 70 and over in Sichuan.

The dashboard fills that gap with an explicit model. For a selected segment, the patient count is the region's published patient count multiplied by the national age share and the national sex share — an independence assumption between the three dimensions. That assumption is false in detail: Sichuan skews older and more male than the national average, Utah skews younger. It is a reasonable first-order approximation and nothing more, which is why every chart that uses it carries an amber badge.

Counts are published prevalence applied to a census denominator, then rescaled so each dimension sums to the published same-instrument national total. China anchors at 95.0M — 13.7% of the 693.8M adults aged 40 and over in the 2020 census — with a scale factor of 0.889 applied because the band prevalences were surveyed in 2012–15 and the age structure is 2020. The US anchors at the BRFSS 2021 diagnosed total of 14.2M, with a scale factor of 0.824 taking 2021 band percentages onto 2023 civilian population estimates. Province and state counts are that region’s prevalence on its own census denominator.

  • Unfiltered charts show published data and are badged "Published data".
  • Any chart recomputed for a selected segment is badged "Modeled estimate".
  • Prevalence percentages are always the published rate for that group and are never re-estimated for a cross-filtered segment. Only the counts move.
  • The US figures are anchored to BRFSS 2021 throughout — bands, sex split and total on one instrument. The narrower NHIS diagnosed count of 11.7M measures something slightly different and is shown in the definitions table above.
  • The region chart lists surveyed high- and low-burden provinces and states, not all of them — the bars do not sum to the national total.
  • Trend series under a filter are the national series scaled by that segment's share of the national patient total. That is a proportional attribution, not a measured segment series.

Sources

Sources.

Numbered as cited above. Vintage matters here: the freshest Chinese field survey is 2014–15, the freshest US federal prevalence is NHIS 2023, and the US rehabilitation-utilization figure comes from 2012 Medicare claims, which remains the standard citation. References stay in their original language.

  1. [1]Wang C, Xu J, Yang L, et al. China Pulmonary Health (CPH) study. Lancet 2018;391:1706–17. thelancet.com
  2. [2]Fang L, Gao P, Bao H, et al. Nationwide COPD prevalence (China CDC surveillance). Lancet Respir Med 2018;6:421–30. pubmed 29650407
  3. [3]Dong F, Su R, Ren Y, Yang T. GBD 2021 China COPD burden. CMJ Pulm Crit Care Med 2025. PMC12266252
  4. [4]The burden of COPD in China and its provinces, GBD 2019. Front Public Health 2022. frontiersin.org
  5. [5]Liu M, et al. GBD 2021 China COPD with projections to 2050. Front Med 2025; companion projection paper. frontiersin.orgPMC12568032
  6. [5b]Spatial analysis of COPD prevalence, 2014–15 national survey. PMC7025678
  7. [6]Luo Y, et al. Meta-analysis of Chinese prevalence studies 2014–2024. Asia Pac J Public Health 2025;37:467–81. sagepub.com
  8. [7]Fujian provincial survey with cross-province comparison table (Sichuan 25.4%, Chongqing 18.7%, Jiangsu 11.9%, Jiangxi 10.6%, Hebei 9.6%, Inner Mongolia 9.3%). Int J COPD 2022; Anhui: BMC Pulm Med 2019; Shanxi. PMC9491879springer.comsciencedirect.com
  9. [8]Non-smoking COPD in China: systematic review and meta-analysis. BMC Public Health 2024. springer.com
  10. [9]Kashi (Xinjiang) prevalence survey 2018–19. Int J Gen Med 2021; Guizhou (ages 20+): Tob Induc Dis 2025. PMC7981135PMC12966984
  11. [10]Tibet high-altitude prevalence 2015–16. Front Med 2020; severity at ≥3,000 m: Int J COPD 2021. PMC7744817PMC8408343
  12. [11]Yunnan high-risk population screening. Int J COPD 2024. PMC11227295
  13. [12]Household solid-fuel use trends 2014–2020. China CDC Weekly 2024. PMC11534580
  14. [13]Awareness: 中国居民慢阻肺监测 via 羊城晚报 2026-06-21; ScienceNet expert interview 2024-01-13; China Daily 近1亿 2024-10-09. ycwb.comsciencenet.cnchinadaily.com.cn
  15. [14]COPD prevention and pulmonary-rehabilitation availability review (27.4% figure). PMC8264680
  16. [15]Physician pulmonary-rehabilitation awareness/training survey. J Thorac Dis 2021. PMC8411159
  17. [16]American Lung Association, COPD Trends Brief (prevalence, mortality, burden; NHIS/BRFSS series). lung.org
  18. [17]NCHS Data Brief No. 529: COPD in Adults, United States, 2023 (published May 2025). cdc.gov
  19. [17b]State COPD death-rate range (CDC WONDER–derived; EPA ROE indicator); sex-specific state mortality. PMC9098243
  20. [18]American Lung Association, COPD in Your State (BRFSS 2023 state table). lung.org
  21. [18b]MMWR 72(46) 2023: COPD prevalence trends and age/race/state detail, BRFSS 2011–2021. PMC10684355
  22. [19]CDC PLACES model-based state estimates (2026 aggregation via RHT Compass — secondary source; KY/PA gaps). rhtcompass.com
  23. [20]Undiagnosed share: USPSTF evidence summary; ATS under/overdiagnosis review; NHLBI COPD National Health Issue report. pubmed 18316746atsjournals.orgnhlbi.nih.gov
  24. [21]Rural–urban: MMWR QuickStats 2023; MMWR 71(18) 2022 rural burden. cdc.govcdc.gov (PDF)
  25. [22]Women's COPD susceptibility. PMC8761805
  26. [23]Spitzer et al. Pulmonary-rehabilitation participation after COPD hospitalization among Medicare beneficiaries. Ann ATS 2019. PMC6344454
  27. [24]Moscovice et al. Pulmonary-rehabilitation geographic access. CHEST 2019; drive-time accessibility; Medicare access. pubmed 30978331PMC10844999PMC10112415
  28. [25]Exacerbation volumes: StatPearls COPD. NBK559281
  29. [26]COPD cost projections through 2029. CHEST 2023. pubmed 38042365
  30. [27]CDC Chronic Disease Indicators API (BRFSS COPD yearly values, dataset hksd-2xuw; queried 2026-08-30); MMWR 64(11) 2015; MMWR 67(7) 2018. data.cdc.govcdc.govPMC5858043
  31. [28]US CLRD deaths (ICD-10 J40–J47), yearly: NCHS leading-causes series 1999–2017 (API bi63-dtpu); NCHS Mortality in the United States 2018; 2019–2023 leading-causes summary; ALA COPD mortality series. data.cdc.govcdc.govPMC11874302lung.org
  32. [29]Chan KY, et al. China COPD prevalence synthesis (GHERG). J Glob Health 2017;7:020704; GBD 2013 China figures as quoted therein. jogh.org
  33. [30]Guo B, et al. COPD burden and 2030 projections, China/US/India. Int J COPD 2024. PMC10929568
  34. [31]Chen W, et al. Cardiovascular comorbidity in COPD: systematic review and meta-analysis. Lancet Respir Med 2015;3:631–39; undiagnosed cardiac dysfunction meta: ERJ Open Res 2023. thelancet.comPMC10680032
  35. [32]Schnell K, et al. Comorbidities in US COPD, NHANES 1999–2008. BMC Pulm Med 2012;12:26. PMC3461433
  36. [33]Huang K, et al. Comorbidity burden in Chinese COPD (national survey). Lancet Reg Health West Pac 2024 (cost-share and readmission figures cited therein). PMC10865049
  37. [34]Depression in COPD, global meta 2000–2022. J Affect Disord 2023; anxiety in older COPD meta. Geriatr Nurs 2025. sciencedirect.comsciencedirect.com
  38. [35]Chen Y-W, et al. Osteoporosis in COPD meta-analysis. Chest 2019;156:1092–1110. pubmed 31352034
  39. [36]Zhao G, et al. Lung cancer in COPD meta-analysis. Front Oncol 2022; 4–6× risk. PMC9523743PMC6831920
  40. [37]Benz E, et al. Sarcopenia in COPD. Eur Respir Rev 2019;28:190049; Marengoni A, et al. Frailty in COPD. Chest 2018;154:21–40. ersnet.orgpubmed 29477493
  41. [38]COPD–OSA overlap meta. Sleep Med 2025; GERD and exacerbations: BMC Pulm Med 2020;20:2. sciencedirect.comPMC6950869
  42. [39]Vanfleteren LE, et al. Clusters of comorbidities in COPD. AJRCCM 2013;187:728–35; Medicare comorbidity analysis. Int J COPD 2017. pubmed 23392440PMC5327909
  43. [40]Divo M, et al. COTE comorbidity index. AJRCCM 2012;186:155–61; EXACOS-CV US (exacerbation → death/CV multipliers). Int J COPD 2024; Pumar MI, et al. depression/anxiety outcome risks. J Thorac Dis 2014. pubmed 22561964PMC10802125amegroups.org
  44. [41]Zhejiang provincial COPD prevalence survey, five sites including Tonglu, Hangzhou (12.34%, adults 40+). 中国公共卫生 2023. zgggws.com
  45. [42]Hangzhou CDC: COPD mortality trend, age-standardized 44.55 per 100,000, annual percent change −9.52%. 疾病监测 2020;35(5). doi.org
  46. [43]China 2020 census (第七次全国人口普查), Zhejiang provincial tabulations and age structure. tjj.zj.gov.cncitypopulation.de
  47. [44]Hangzhou 2023 statistical communiqué — resident population 12.52M, 19.6% aged 60 and over. hangzhou.gov.cn
  48. [45]America's Health Rankings, COPD prevalence by state — California 4.0%. americashealthrankings.org
  49. [46]CDC PLACES county-level model-based estimates, 2023 release (dataset swc5-untb) — California county range 3.5–6.7%. data.cdc.gov
  50. [47]NCHS Vital Statistics Rapid Release provisional mortality (dataset 489q-934x) — California CLRD deaths 11,982 in 2022; age-adjusted rate 25.7 per 100,000 in 2023 against a US rate of 33.4. data.cdc.gov
  51. [48]US Census Bureau Vintage 2024 population estimates, civilian population by age and state — the denominators behind the US band reconciliation. census.gov
  52. [49]NCHS Data Brief: Mortality in the United States, 2024 — CLRD deaths 145,643, fifth leading cause. cdc.gov (NBK620752)
  53. [50]NCHS VSRR Report No. 44 (July 2026): provisional US mortality for 2025 — CLRD 148,408. cdc.gov (vsrr044)
  54. [51]UN World Population Prospects 2024, medium variant — the population denominators behind the 2030/2040/2050 projections. population.un.org
  55. [52]Boers E, et al. Global burden of COPD through 2050. JAMA Network Open 2023 — global cases 480M (2020) to 592M (2050). jamanetwork.com

About the modeled estimates

With no filter applied, every chart on this page shows published data taken directly from the numbered sources below.

Cross-filtered views — any chart after a selection is applied — are computed estimates. They are reconciled from published marginal distributions, which report age, sex and geography separately, together with census population denominators, under an assumption of independence between those dimensions. No source publishes the joint breakdown.

Modeled values are provided for market-analysis orientation only. They are not epidemiological findings and should not be cited as published statistics.

See Methodology for the derivation and every underlying source.

Bringing the program to where the patients are.

If you work in a health system, a hospital, or a rehabilitation service, we would like to hear from you.

Request a demoFor health systems