How common is celiac disease? The figures people quote, and what each one counts

Short answer

The figures people quote come from different groups, case definitions and years. Read each one for who was counted and when:

  1. About 1 percent of people worldwide and about 2 million people in the United States: the estimate NIDDK reports1.
  2. 0.71 percent (1 in 141) of people aged 6 and older in the United States in 2009-2010, in one national survey2, 3.
  3. 7.5 percent of first-degree relatives of people with celiac disease, pooled across studies in a 2015 meta-analysis4.
  4. About 3 percent of people with the DQ2 or DQ8 gene types develop celiac disease, and NIDDK says about 30 percent of people carry them5.

Published by Gluten Verdict · updated · 12 sources, each linked below

Reviewed by Mark Peter, Registered dietitian (RD), on

Queue of shoppers seen from behind at a bakery counter with a yellow tote

Key points

  • NIDDK reports an expert estimate of about 1 percent of people worldwide; the page carrying it was last reviewed in October 20201, 6.
  • In one US national survey, 29 of the 35 people found to have celiac disease were unaware of their diagnosis7.
  • Relatives’ figures differ by study: 1 in 22 in a 2003 US study8, 7.5 percent pooled in 20154, and 11 percent by antibody test in a 2024 meta-analysis9.
  • Stored blood from 1948-1954 showed undiagnosed celiac disease at 0.2 percent; two recent Minnesota cohorts showed 0.9 and 0.8 percent10, 11.
  • Repeat US surveys over 2009-2014 found no significant change in overall prevalence12.

Why there is no single number

A prevalence figure is the share of a group that has a condition, so it depends on which group was tested and how a case was defined. The figures in the table below count three different kinds of group: everyone, relatives of a person with celiac disease, and carriers of the DQ2 or DQ8 gene types.

Figures people quote, with the group each one counts (every source read on 2026-11-17)
FigureWho was countedWhere it comes from
About 1 percent (about 1 in 100)People around the worldThe estimate NIDDK reports1. A 2018 meta-analysis pooled 1.4 percent by antibody test and 0.7 percent by biopsy13
0.71 percent (1 in 141)People aged 6 and older in the United States, 2009-2010National Health and Nutrition Examination Survey, Rubio-Tapia et al., 20122, 3
1 in 133Not-at-risk individuals in a 2003 US study (4,126 of 13,145 people screened)Fasano et al., 200314, 8
1 in 22First-degree relatives of people with biopsy-proven celiac disease, USFasano et al., 20038, 14
7.5 percent (and 2.3 percent in second-degree relatives)First-degree relatives of people with celiac disease, pooled across studiesSingh et al., 20154
11 percent by antibody test; 7 percent confirmed by biopsyFirst-degree relatives, pooled across studiesKarimzadhagh et al., 20249
About 3 percentPeople with DQ2 or DQ8 who develop celiac disease (about 30 percent of people have one of these)NIDDK5

A first-degree relative is a sister, brother, mother, father, son or daughter of the person with celiac disease15. NIDDK says you are more likely to develop celiac disease if someone in your family has it16, and the relatives’ figures differ with the relationship: the 2015 meta-analysis found the highest pooled prevalence in siblings (8.9 percent), then offspring (7.9 percent) and parents (3.0 percent)17.

The 3 percent figure is a different kind of number. NIDDK reports that people without DQ2 or DQ8 are very unlikely to develop celiac disease, that about 30 percent of people have DQ2 or DQ8, and that only about 3 percent of those people develop it5. It answers “of carriers, how many?”, not “of everyone, how many?”.

What the NIDDK figure counts

NIDDK, part of the US National Institutes of Health, writes that many people who have celiac disease have not been diagnosed, and that experts estimate about 2 million people in the United States, and about 1 percent of people around the world, have it1. Because the estimate is stated for people who have celiac disease, it is not limited to those with a diagnosis.

The first reference on that page is a paper on the National Health and Nutrition Examination Surveys from 2009 to 201418, and the page was last reviewed in October 20206. Treat it as an expert estimate, not a count made this year.

How many cases are undiagnosed?

The 0.71 percent figure comes from serum samples of 7,798 people aged 6 and older in the 2009-2010 survey, tested for tissue transglutaminase antibodies and, when abnormal, endomysial antibodies2. A case was either a double-positive antibody result or a reported diagnosis by a health-care professional together with a gluten-free diet19.

The authors found celiac disease in 35 participants, 29 of whom were unaware of their diagnosis7. They put the US prevalence at 0.71 percent (1 in 141) and wrote that most cases were undiagnosed3. Among non-Hispanic whites the figure was 1 percent, and the authors describe celiac disease as rare among minority groups20.

The same survey series later found overall prevalence did not change significantly between 2009-2010 (0.7 percent), 2011-2012 (0.8 percent) and 2013-2014 (0.7 percent), while undiagnosed celiac disease fell from 0.6 percent in 2009-2010 to 0.3 percent in 2013-201412. The authors wrote that the proportion of cases that is hidden considerably decreased21.

Has celiac disease become more common?

Four sets of studies speak to this. Each covers a different place and period, so read them side by side rather than as one trend line.

Stored blood from 1948 to 1954

Researchers tested stored blood from 9,133 young adults at Warren Air Force Base, collected between 1948 and 1954, and blood from 12,768 people in two recent cohorts from Olmsted County, Minnesota; all were tested for tissue transglutaminase antibodies and, if abnormal, endomysial antibodies22. Undiagnosed celiac disease was found in 14 of the 9,133 (0.2 percent)10, against 0.9 percent and 0.8 percent in the two recent cohorts, which the authors give as 4.5-fold and 4-fold greater11. They concluded that the prevalence of undiagnosed celiac disease seems to have increased dramatically in the United States during the past 50 years, as of their 2009 paper23.

The old samples come from one group of young adults at one air base, so the comparison speaks for that group and the Minnesota cohorts22.

Finland, 1978-80 and 2000-01

Researchers in Finland measured total prevalence in two population samples of adults, 8,000 people in 1978-80 and 8,028 in 2000-01, counting both clinically diagnosed patients and previously unrecognized cases found by endomysial antibodies24. They found 1.05 percent in 1978-80 and 1.99 percent in 2000-0125. The authors concluded that prevalence seems to have doubled and that the increase cannot be attributed to the better detection rate; they called the environmental factors behind it issues for further studies26.

Incidence: the 2020 meta-analysis

A 2020 systematic review and meta-analysis of incidence found that the pooled average annual percent change was an increase of 7.5 percent (95% CI 5.8 to 9.3)27. Its authors say incidence has been rising in the latter half of the 20th century and into the 21st throughout the Western world, and that population-based studies in Africa, Asia and Latin America are still needed28.

Incidence is reported as new cases per 100,000 person-years, as in the review’s child-specific figure of 21.329. It is not a share of the population, so do not set it beside “1 percent”. The 7.5 percent here is an average annual percent change; it has nothing to do with the 7.5 percent of relatives in the table above.

The recent US survey series

Over 2009-2014, the US surveys found overall prevalence stable21. That window is much shorter than the decades covered by the stored-blood and Finnish comparisons.

How the figure differs by sex, age, region and race or ethnicity

NIDDK says that in the United States celiac disease is more common among white Americans than among other racial or ethnic groups, and that a diagnosis is more common in females than in males30.

The 2018 global meta-analysis reported regional values in its abstract, along with higher prevalence in females than males (0.6 vs 0.4 percent) and in children than adults (0.9 vs 0.5 percent)31:

Pooled prevalence by region, as listed in the abstract of Singh et al., 2018
RegionPrevalence
South America0.4 percent31
Africa and North America0.5 percent31
Asia0.6 percent31
Europe and Oceania0.8 percent31

The abstract gives these regional values without saying next to them which test definition each uses, so check the full paper before setting them against the 1.4 percent and 0.7 percent global figures.

Within the United States, a study of 22,277 participants aged 6 and older in the 2009-2014 surveys32 found celiac disease more common at latitudes of 35°-39° North (odds ratio 3.2; 95% CI 1.4 to 7.1) and at 40° North or more (odds ratio 5.4; 95% CI 2.6 to 11.3) than below 35° North, independent of race or ethnicity, socioeconomic status and body mass index33.

How to read a prevalence figure you meet online

  1. Find the group. Everyone, relatives, symptomatic patients, gene carriers or one country? The table above shows several groups for figures that all sound alike.
  2. Find the definition of a case. Antibody blood tests and biopsy confirmation give different numbers: the 2018 meta-analysis pooled 1.4 percent by antibody test and 0.7 percent by biopsy13.
  3. Find the year and the place. A 2009-2010 US survey, blood from 1948-1954 and Finnish samples from 1978-80 and 2000-01 describe different people at different times2, 22, 24.
  4. Tell prevalence from incidence. One is the share of a group that has the condition; the other is new cases per 100,000 person-years.
  5. Open the paper. Every study in this guide is listed under Sources with its PubMed link; the abstract is the place to look for the group, the dates and how a case was defined.

Gluten avoidance is a separate count

The same US surveys also counted people without celiac disease who avoid gluten34. That group rose from 0.5 percent in 2009-2010 to 1.0 percent in 2011-2012 and 1.7 percent in 2013-201435. Across the 2009-2014 surveys, one analysis found 0.7 percent of participants had celiac disease and 1.1 percent avoided gluten without it32, 36.

Those are shares of people on a gluten-free diet without celiac disease, so they say nothing about how many people have the disease36. NIDDK separates celiac disease from gluten sensitivity, which it says does not damage the small intestine37. For that comparison, see celiac disease versus gluten intolerance.

Questions people ask

Is celiac disease common?

NIDDK reports an expert estimate that about 1 percent of people around the world have it1. One US national survey found 0.71 percent of people aged 6 and older3.

How many people in the United States have celiac disease?

NIDDK reports that experts estimate about 2 million1. The national survey figure for 2009-2010 was 0.71 percent, or 1 in 1413.

Why do some pages say 1 in 133 and others 1 in 141?

They come from different studies. 1 in 133 is the figure for not-at-risk individuals in a 2003 US study8; 1 in 141 comes from the 2009-2010 national survey3.

What share of relatives of a person with celiac disease have it?

Studies report different figures: 1 in 22 for first-degree relatives in a 2003 US study8, 7.5 percent pooled in 20154, and 11 percent by antibody test and 7 percent by biopsy in 20249.

Do most people with the DQ2 or DQ8 genes develop celiac disease?

NIDDK reports that about 30 percent of people have DQ2 or DQ8 and that only about 3 percent of those people develop celiac disease5.

More in Gluten basics and myths

Sources