The epidemiology of coeliac disease in South Wales: a 28-year perspective
- Correspondence to Dr Huw R Jenkins, , Department of Child Health, Children's Hospital of Wales, Heath Park, Cardiff CF14 4XW, UK;
- Received 29 September 2012
- Revised 21 February 2013
- Accepted 1 March 2013
- Published Online First 29 March 2013
Objectives The diagnosis of coeliac disease (CD) has increased in frequency, particularly since the accuracy of serological antibody testing has improved. Previous studies from South Wales have shown an increase in the frequency of diagnosis from 1983 to 2004 with a decrease in specific gastro-intestinal symptoms, as well as an increasing age at diagnosis.
Aims/methods We aimed to determine whether the frequency of diagnosis, the age at presentation and the clinical presentation of CD have changed between 2005 and 2011 compared with previously published data from 1983 to 2004. We reviewed all patients with CD presenting to the South Wales’ Regional Centre between 2005 and 2011 and compared the age and documented mode of presentation with previous data from the same area.
Results 163 cases of CD were diagnosed between 2005 and 2011 (23 cases/year) with the median age at diagnosis increasing to 14 years (range 0.8–16 years) compared with 50 cases (8/year) between 1999 and 2004 (median age at diagnosis 8 years), 25 cases (2.5/year) between 1990 and 1998 and 11 cases (1.5/year) between 1983 and 1989. 41% presented with specific gastro-intestinal symptoms, 23% with non-gastro-intestinal features and 36% were asymptomatic and diagnosed after serological screening of high-risk groups. Compared with the most recent previous study from the same population, the percentage of patients presenting with gastro-intestinal symptoms remain similar (42% vs 41%) but patients diagnosed after targeted screening had increased from 26% to 36%.
Conclusions The frequency of diagnosis of CD in this defined population has continued to rise, with an increase in the median age at diagnosis, and over 50% of patients exhibited few or no symptoms.
What is already known about this topic
Coeliac disease is common with a prevalence of 1 in 100.
There is a changing clinical presentation with a suggestion that children are presenting with fewer severe gastrointestinal symptoms and at an older age.
What this study adds
Coeliac disease may present in a non-specific and subtle way in many cases.
Age at diagnosis has increased within this region over a 30-year period.
The trend of increasing frequency of diagnosis continues within this region.
Coeliac disease (CD) is a disorder characterised by a gluten-sensitive enteropathy resulting in damage to the mucosa in the small intestine. It is now believed to be the most common genetically predetermined condition in humans with a childhood prevalence of 1% in most European countries and the USA,1 although most cases remain undiagnosed in childhood. It is clear that CD is a multi-system immunologically-mediated disorder involving the gastro-intestinal tract as one of many possible clinical manifestations.2The clinical presentation has altered over a number of years with an increasing number of children presenting with relatively minor gastro-intestinal symptoms and a higher proportion now being diagnosed following screening of asymptomatic individuals in high-risk groups. We have previously published epidemiological data on CD presenting in South Wales from 1983 to 2004,3–5 which have demonstrated an increase in frequency of diagnosis over this time and significant changes in the presentation of childhood CD, with a small proportion presenting with severe gastro-intestinal manifestations and almost 1 in 4 children diagnosed after targeted screening (a significant proportion being relatively asymptomatic).
The aims of our present study were to continue the 28 year study of children with CD diagnosed in South Wales and to determine whether the trend towards an older age at presentation and away from a primary gastro-intestinal presentation had been maintained.
The ages and presenting features of children (0–16 years) with CD diagnosed by the Regional Paediatric Gastroenterology Centre between 2005 and 2011 (inclusive) were recorded prospectively. Data were collected on the age at diagnosis, clinical symptoms at initial presentation and serological testing results; serological testing was undertaken by the tertiary paediatric gastroenterology centre on all patients within 1 month of the formal histological diagnosis, although, in addition, a small proportion of cases had initially been referred from secondary care paediatricians or primary care with positive serology. All serological tests were performed in the same regional immunology laboratory. All cases were biopsy-proven according to the European Society of Paediatric Gastroenterology, Hepatology and Nutrition (ESPGHAN) criteria from 1990.6 The case records of these children were also retrospectively crosschecked with the local CD Database maintained in the department, as well as the Histopathology/Regional Immunology Database and Dietetic Records to ensure full ascertainment of cases. This unit is the only one performing endoscopy in children in South East Wales and we were confident that there was full ascertainment of all cases. Over the study period, the childhood population age 16 years and under was relatively stable at 280 000 to 290 000 in the relevant geographical area.7
One hundred and sixty-three patients, residents of the South East Wales area, were diagnosed with CD between 2005 and 2011–68 (42%) male and 95 (58%) female, with the median age at diagnosis being 14 years (range 0.8–16 years) and mean age 9.1 years. The number of patients diagnosed per year was 23 between 2005 and 2011 (the precise numbers per year during this time being 17, 21, 27, 21, 27, 20 and 30) compared with 8/year between 1999 and 2004. The median age at diagnosis had risen to 14 years (range 0.8–16 years) compared with 8 years previously.
Forty-one per cent presented with specific classic symptoms of malabsorption, diarrhoea, steatorrhoea, weight loss, iron deficient anaemia or growth failure as defined by the Oslo criteria published recently.8 Twenty-three per cent presented with non-classic CD features (no signs or symptoms of malabsorption) and 36% were asymptomatic and diagnosed after screening of high-risk groups (family history of CD in first-degree relatives (12%), screening of children with type 1 diabetes mellitus (18%), autoimmune thyroid disease (2%), autoimmune hepatitis (1%), Down syndrome (2%) and Turner syndrome (1%)). The majority of patients diagnosed after screening were in the older age group with those referred with type 1 diabetes mellitus having a median age of 14 years (range 6–16) and those with a positive family history 13 years (range 8–16).
All patients underwent serological screening with anti-tissue transglutaminase (TTG) antibody testing (IgA and IgG if IgA deficient) and all had upper gastro-intestinal endoscopy and multiple biopsies from the second and third part of the duodenum, as well as the duodenal cap. Fifty-seven patients were biopsied within 1 month of referral with all patients undergoing a biopsy within 3 months of presentation.
Compared with previous studies from the same population area (since 1983; table 1)—the median age at diagnosis had increased again and the number of patients diagnosed between 2005 and 2010 had risen dramatically from 8/year to 23/year, with the percentage of patients presenting with gastro-intestinal symptoms remaining similar to the study published in 2006 (42% vs 41%).3 This rise is primarily due to an increase in patients with recurrent abdominal pain and constipation (>50%) compared to previous studies from this region, which demonstrated more classic symptoms (diarrhoea, weight loss and vomiting). Those diagnosed through targeted screening rose from 26% to 36%.
It is clear that the incidence of childhood CD has increased dramatically over the 28-year time period with a current total of 265 patients diagnosed being under 16 years of age, within a stable childhood population of 285 000—the prevalence is almost one case in 1000 in this region.
It is clear that the majority of patients with CD now present with relatively mild symptoms and indeed over a third of patients diagnosed in this population have few if any symptoms and are diagnosed as a consequence of targeted screening of high-risk groups. If the expected prevalence of CD is 1 in 100, it is clear that a large proportion (over 90%) of cases remain undiagnosed in this population as is the case in other parts of the UK.9 The rate of diagnosis, however, has increased in this region by 50% over the last 7 years, which is encouraging. It is possible that the recent National Institute for Health and Clinical Excellence (NICE) guidance10 on CD has prompted an increase in awareness of the condition, although in this population the increase in frequency of diagnosis started before this time. Other publications have highlighted the importance of thinking about the diagnosis in children without the classic and severe gastro-intestinal presentations,2 ,11 and we hope this will increase the rate of diagnosis across the UK. This at least suggests that there is an increased awareness of the possible diagnosis in children who are in high-risk categories and indeed those who have extra-intestinal symptoms, and monosymptomatic presentation with oral ulceration or recurrent abdominal pain. The referral base and childhood population over the 28-year period is relatively unchanged and we are confident of full ascertainment of all patients diagnosed over this period. As the population genetics is unlikely to have altered during this time and there is no evidence of significant alteration in environmental risk factors, this dramatic increase in the frequency of diagnosis is likely to be due to an increasing awareness of the condition and its changing and more subtle presentation, resulting in more appropriate referral patterns. However, with a likely prevalence rate of 1% in the population, it is clear that many cases are still undiagnosed in childhood and clinicians must be vigilant in considering the diagnosis in children who present with relatively non-specific symptoms and signs.
Contributors LAW, HRJ—design of study, data analysis, drafting and reviewing article.
Competing interests None.
Patient consent Obtained.
Ethics approval Ethics Committee University Hospital of Wales.
Provenance and peer review Not commissioned; externally peer reviewed.