Investigations of the morbidity, mortality and effect of treatment in coeliac disease and dermatitis herpetiformis
Principal Investigator: Dr Nina Lewis
Institution: University of Nottingham
Project completion: 2009
Grant awarded: £200K clinical training fellowship
From reading the literature, people with coeliac disease appeared to be at a reduced risk of vascular disease and breast cancer. However it was not known whether people with undetected or non classic coeliac disease had the same level of risks from the disease or benefits from the treatment of a gluten-free diet as those with classic coeliac disease. This project determined whether treatment with a gluten-free diet alters the risk factors for vascular disease and breast cancer in patients with coeliac disease and examined the mortality risks associated with undetected coeliac disease.
Several projects were completed within this clinical research fellowship:
Reproductive and hormonal risk factors for breast cancer among women with coeliac disease: a population based questionnaire study.
Several studies have observed that coeliac disease is associated with more than a 50% reduced risk of breast cancer though the reasons for this reduction are unclear. This project concluded that reduced risk of breast cancer in women with coeliac disease may be related to differences in menstrual and reproductive factors.
The project determined the conventional breast cancer risk factors reported by women with coeliac disease in comparison to the general population.
9000 female Members of Coeliac UK, diagnosed with coeliac disease and 1193 women with coeliac disease who had attended either Nottingham University Hospital, Royal Hallamshire Hospital or the Derbyshire Royal Infirmary between January 2000 – December 2006 were invited to complete a questionnaire relating to known risk factors for breast cancer. Data from British birth cohort studies were used as a general population comparison.
7416 women with coeliac disease, average age 58.7 years, completed the questionnaire. The average age at adult diagnosis was 44.8 years with 6% of women being diagnosed in childhood. The average age for the start of menstruation was 13.4 years and adult height 162.8 cm, neither differed from that observed in the general population. The average age at first term pregnancy was 25.8 years and did not differ from the general population, though a greater proportion of women in the coeliac cohort had their first term pregnancy by the age of 30 years (73.6% v 66.7%). Average age at menopause for the group of women with coeliac disease was 47.7 years and appeared to be younger than in the general population. Of those who were 50 years or younger at the time of responding to the questionnaire, 17.6% reported having had the menopause.
Reduced risk of breast cancer in women with coeliac disease may be related to differences in menstrual and reproductive factors.
Cholesterol profile in people with newly diagnosed coeliac disease: a comparison with the general population and changes following treatment.
Previous studies have suggested that untreated coeliac disease is associated with lower total cholesterol compared to the general population while the effect of treatment with a gluten-free diet on the cholesterol profile of clinically apparent coeliac disease was unknown.
This study concluded that any increase in risk of heart disease or stroke in people with coeliac disease is unlikely due to an adverse cholesterol profile either before diagnosis or after treatment with a gluten-free diet.
The cholesterol profile at diagnosis and again after following 12 months treatment with a gluten-free diet was measured in 100 adults with coeliac disease attending the Royal Hallamshire Hospital, Sheffield, UK. The results were compared with Health Survey for England figures.
The average age at diagnosis was 51 years and the average total cholesterol was 4·84 mmol/l in adults newly diagnosed with coeliac disease. At diagnosis of coeliac disease, men had 21% lower and women had 9% lower average total cholesterol in comparison to the general population.
There was no change in average total cholesterol following treatment with a gluten-free diet. However, there was a small but statistically significant increase of 0·12 mmol/l in the average HDL-cholesterol.
Total cholesterol was lower at diagnosis in coeliac patients than in the general population and did not increase with 1 year of a gluten-free diet, while HDL-cholesterol increased following treatment.
Any increase in risk of heart disease or stroke in people with coeliac disease is unlikely due to an adverse cholesterol profile either before diagnosis or after treatment with a gluten-free diet.
Lewis NR, Sanders DS, Logan RFA. et al. (2009) Cholesterol profile in people with newly diagnosed coeliac disease: a comparison with the general population and changes following treatment. Br. J. Nutr. 10, 509-13
Risk of morbidity in contemporary coeliac disease
It is important that disorders associated with coeliac disease are recognised so that patients are properly managed.
It has been known for years that many disorders coexist with coeliac disease and a number of complications may arise, but which of these occur other than by chance and the magnitude of the risks have been less clear. This is because studies have often been poorly designed and underpowered to provide reliable information.
Recent epidemiological studies have attempted to minimize these sources of error and provide more reliable information. Researchers have access to large databases such as the General Practice Research Database (GPRD) in the UK and the Swedish In-Patient Registry, and these have allowed more precise estimates of the risks of comorbidities to be determined. These more reliable studies have been used within this review and the findings are summarised:
Autoimmune diseases constitute clinically important associations with coeliac disease, of which Type 1 diabetes mellitus and thyroid disorders are the most important. Several liver disorders, including primary biliary cirrhosis and primary sclerosing cholangitis, are also associated.
The frequency of malignant complications of coeliac disease is much lower than earlier studies have indicated.
The increase in fracture risk in coeliac disease is only modest and although neurological and psychiatric conditions affect coeliac patients, no disorder specifically associated with coeliac disease has been identified. Reproductive problems have been over exaggerated.
It is important to acknowledge that even the more recent, more reliable studies are not without their drawbacks, and may still introduce an element of bias. Only large studies will reveal whether some of the rare comorbidities are associated with coeliac disease rather than occurring just by chance.
An added difficulty is that so many coeliac patients are undiagnosed that prevalence figures will only ever be an approximation for a given number of diagnosed cases at a specific time. Nevertheless, studies using the GPRD and the Swedish In-Patient Registry have advanced understanding and will help to inform opinion on the questions of whether and which conditions merit screening.
Lewis NR, Holmes GK. (2010) Risk of morbidity in contemporary celiac disease. Expert Rev, Gastroenterol Hepatol.Dec; 4(6): 767-80
No increase in risk of fracture, malignancy or mortality in dermatitis herpetiformis: a cohort study.
Dermatitis herpetiformis forms part of the same spectrum of gluten-sensitive conditions as coeliac disease yet may have different risks of morbidity and mortality.
This project quantified the risks of fracture, malignancy and mortality in people with dermatitis herpetiformis compared with the general population using hazard ratios.
Using the General Practice Research Database (GPRD), 846 people with dermatitis herpetiformis were identified along with 4225 age-, gender- and GP practice matched controls.
Comparing people with dermatitis herpetiformis to the general population, the overall hazard ratio for any fracture was 1.1 (95% CI: 0.77–1.52). The overall hazard ratio for any malignancy was 1.0 (95% CI: 0.73–1.49); there was no increased risk of gastrointestinal cancers (HR: 1.6; 95% CI: 0.67–3.67) or lymphoproliferative cancers (HR: 1.6; 95% CI: 0.44–6.06). A reduction in risk of breast cancer was not statistically significant (HR: 0.19; 95% CI: 0.03–1.39). The hazard ratio for all-cause mortality was 0.93 (95% CI: 0.70–1.23).
Unlike the fivefold increased risk seen in coeliac disease, there was no found increased risk of lymphoproliferative cancer and no increase in fracture, malignancy or mortality in people with dermatitis herpetiformis compared with the general population. It is not clear whether differences in degree of intestinal inflammation or other reasons account for this.
Lewis NR, Logan RFA, Hubbard RB et al. (2008) No increase in risk of fracture, malignancy or mortality in dermatitis herpetiformis: a cohort study. Aliment. Pharmacol. Ther. 27:1140-47
In addition to the main projects and focus of the clinical research fellowship, other research activities were undertaken:
Is the diagnosis of coeliac disease associated with socio-economic status? A population-based study
The rate of diagnosis of coeliac disease in developed countries has increased dramatically since the introduction of serological tests. Little was known about the socioeconomic distribution of coeliac disease; there was some evidence that it was less common in more deprived social groups.
This project quantified the incidence of new diagnoses of coeliac disease by socio-economic status in a large, population based cohort.
The postcode of residence at diagnosis in 837 adult cases of coeliac disease between January 2000 – December 2006 at Nottingham University Hospital and Royal Hallamshire Hospital was used to determine the Index of Multiple Deprivation 2007 (IMD07) score. Incidence rates were calculated using the total adult population for Sheffield and Nottingham derived from UK 2001 National Census categorised into corresponding IMD07 groups.
The rate of new diagnoses of coeliac disease was twice as high in people from affluent areas compared with that in people living in poorer areas. This striking difference could be due to variation in environmental exposures such as breastfeeding practices or could be accounted for by differences in uptake and utilisation of health services.