Long-Term Mortality in People With Celiac Disease Diagnosed in Childhood Compared With Adulthood: A Population-Based Cohort Study

Masoud Solaymani-Dodaran, M.D., Ph.D., M.P.H.; Joe West, M.B., Ph.D., M.R.C.P.; Richard F.A. Logan, M.B., M.Sc., F.R.C.P.


Am J Gastroenterol. 2007;102(4):864-870. 

In This Article



Our study cohort was originally set up in 1979 through an attempt to identify all cases of celiac disease diagnosed in the Lothian region of Scotland. Details of the recruitment process have been published previously.[2,9] In brief, the following sources were used to identify celiac cases: the records of gastrointestinal units of all the hospitals in the region including the Edinburgh Royal Infirmary, Western General Hospital, and Royal Hospital for Sick children up to December 1981; The Scottish Hospital Inpatient Statistics for the years 1961-1977; all the existing regional histopathology records in the three adult and one pediatric pathology laboratories between 1958 and 1980; a postal survey of all general practitioners in the Lothian region in 1979; and finally the local branch of the Celiac Society.

Histological Verification

All the original study participants had to have had at least one abnormal small bowel biopsy and two pathologists reviewed all the specimens at the time of recruitment. For an original celiac case to be included in the current study the findings from histopathology review had to be typical of celiac disease,[9] which resulted in exclusion of cases with the diagnosis of possible celiac disease.


For our current study the follow-up started from the date of clinical diagnosis of celiac disease or January 1, 1970, whichever was later. This was to avoid any survival bias in our analysis. Information about the date of clinical diagnosis was obtained from hospital records. The follow-up ended upon death of the individual, loss to follow-up, or December 31, 2004, whichever came first. The vital status for 95% of the 625 eligible cases was verified at the end of the follow-up period. For those who had emigrated or been lost to follow-up, the date of emigration or the last documented date when they were known to be alive was used as the date that follow-up ended in these individuals.

Identification of Deaths

Members of the cohort were flagged in the National Health Service Central Register (General Register Office for Scotland, Edinburgh) and copies of death certificates were received upon death, which enabled us to determine the causes of death. Underlying cause of death as recorded on the death certificate was used for this purpose. We examined deaths resulting from all causes as well as deaths because of all malignancies, gastrointestinal, lung, and breast cancers, lymphoma and lymphosarcoma, cerebrovascular diseases, ischemic heart diseases, and deaths as a result of accidents, suicide, and violence. ICD (International Classification of Disease) codes were used to define these specific causes (see Appendix).

Population Mortality Rates

We calculated population cause-specific mortality rates for the Lothian region with a population of 780,000 in 2002. The population number has changed very little in the region for the last three decades. The number of deaths and the mid-year population estimates for the region were obtained from the General Registrar office of Scotland for the period of 1974-2002. The rates were calculated for each calendar year in both men and women in 5-yr age categories. The average mortality rates for the periods of 1974-1976 and 2000-2002 were used for 1970-1973 and 2002-2004 calendar years, respectively.

Statistical Analysis

Initially we examined mortality utilizing all follow-up within the study. We then divided follow-up into the first 5 yr following diagnosis of celiac disease and subsequent follow-up. To further examine time trends we divided follow-up into five time periods: the first year, 1-4, 5-14, 15-24, and over 25 yr after the diagnosis of celiac disease. We calculated the person-days contribution of study subjects to each 5-yr age category in each calendar year and each period after diagnosis. The expected numbers of deaths were calculated using population mortality rates for the Lothian region. Standardized mortality ratios (SMRs) (the number of observed deaths divided by the number of expected deaths based on population mortality rates) and their 95% confidence intervals (CIs) were estimated. The Stata 7 (StataCorp LP, College Station, TX) and SPSS, version 11.5 (SPSS Inc., Chicago, IL) statistical packages were used for data manipulation and analysis.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: