Figures
Abstract
Background
Individuals with celiac disease (CD) are at increased risk of sepsis. The aim of this study was to examine whether CD influences survival in sepsis of bacterial origin.
Methods
Nationwide longitudinal registry-based study. Through data on small intestinal biopsies from Sweden’s 28 pathology departments, we identified 29,096 individuals with CD (villous atrophy, Marsh stage III). Each individual with CD was matched with five population-based controls. Among these, 5,470 had a record of sepsis according to the Swedish Patient Register (1,432 celiac individuals and 4,038 controls). Finally we retrieved data on mortality in sepsis patients through the Swedish Cause of Death Registry.
Results
CD was associated with a 19% increase in overall mortality after sepsis (95% confidence interval (CI) = 1.09–1.29), with the highest relative risk occurring in children (adjusted hazard ratio (aHR) = 1.62; 95%CI = 0.67–3.91). However, aHR for death from sepsis was lower (aHR = 1.10) and failed to reach statistical significance (95%CI = 0.72–1.69). CD did not influence survival within 28 days after sepsis (aHR = 0.98; 95%CI = 0.80–1.19).
Conclusions
Although individuals with CD seem to be at an increased risk of overall death after sepsis, that excess risk does not differ from the general excess mortality previously seen in celiac patients in Sweden. CD as such does not seem to influence short-term or sepsis-specific survival in individuals with sepsis and therefore is not an independent risk factor for poor prognosis in sepsis.
Citation: Röckert Tjernberg A, Bonnedahl J, Ludvigsson JF (2016) Does Celiac Disease Influence Survival in Sepsis? A Nationwide Longitudinal Study. PLoS ONE 11(4): e0154663. https://doi.org/10.1371/journal.pone.0154663
Editor: Karol Sestak, Tulane University, UNITED STATES
Received: February 10, 2016; Accepted: April 15, 2016; Published: April 28, 2016
Copyright: © 2016 Röckert Tjernberg et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: For legal reasons, data cannot be made publicly available. Researchers who are interested in obtaining the relevant data, despite the Swedish Secrecy Act, should contact the National Board of Health and Welfare (socialstyrelsen@socialstyrelsen.se).
Funding: This work was supported by the Swedish Research Council to JFL. ART and JB were funded by Kalmar County Council. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exists.
Abbreviations: CD, Celiac disease; aHR, adjusted Hazard ratio; CI, Confidence interval; VA, Villous atrophy; ICD, International classification of disease
Introduction
Celiac disease (CD) is a chronic immune-mediated disorder triggered by the ingestion of gluten in genetically susceptible individuals [1]. It occurs in approximately 1% of the western population and treatment consists of a life-long gluten-free diet [2]. CD has been associated with excess morbidity [3] as well as mortality [4–6]. In addition to an increased risk for malignancies [3] and different autoimmune diseases [7], individuals with CD are at an increased risk of severe infections [8] including tuberculosis [9], influenza [10] and pneumococcal infections [11]. In 2008 we reported an elevated risk for sepsis in patients with CD compared with a group of inpatient reference individuals (hazard ratio (HR) = 1.6; 95% confidence interval (CI) = 1.2–1.9) [12]. These findings supported earlier reports of an excess mortality from sepsis in CD [13].
Sepsis, which is characterized by a systemic inflammatory response with elevated or reduced body temperature, tachycardia, tachypnea and affected white blood cell count, is the body’s sometimes overwhelming reaction to an invasive infection. When organ dysfunction is added to these parameters, the condition is defined as severe sepsis [14, 15]. Incidence data of this life-threatening infection vary but studies report incidence numbers between 50–300 per 100,000 person-years, with a mortality ranging from 10–20% or even higher in severe sepsis and septic shock [14, 16].
Several CD-related factors may predispose to severe infections, including sepsis. One such mechanism is hyposplenism [17] although the susceptibility to infectious diseases has also been attributed to the altered intestinal permeability seen in CD [18]. The incidence of sepsis seems to increase and considering the high mortality [16] it is important to identify risk factors affecting its prognosis. The outcome in sepsis is related to several determinants including different underlying conditions [19, 20]. While a few studies have evaluated the risk of death in infectious diseases in CD [4, 13] we are not aware of any study exclusively examining whether CD affects the prognosis in patients with sepsis.
The aim of this population-based study was therefore to examine whether CD influences survival in sepsis of bacterial origin. We hypothesised that individuals with sepsis and concomitant CD would have a poorer survival than individuals without CD.
Materials and Methods
Study participants
During 2006–2008 data on small intestinal biopsy reports were collected from all of Sweden’s pathology departments (n = 28). The biopsies had been performed in 1969–2008 and data were retrieved by local IT personnel and included information on personal identity number [21], date of biopsy, topography (duodenum or jejunum) and morphology in compliance with the Swedish SnoMed classification codes (see S1 file), a classification used by all 28 departments. Details on this data collection process have been published previously [22]. CD was defined has having villous atrophy (VA, equivalent to Marsh stage III). We did not require positive serology for the CD diagnosis; however, a previous validation has shown that in a random sample of study participants with VA some 88% were positive for CD serology at time of biopsy [22], a proportion consistent with data from other countries [23]
After removal of duplicates and erroneous data we identified 29,096 individuals with VA (identical to the dataset used in our previous study on CD and mortality [6]). Each individual was thereafter matched with five controls from the Swedish Total Population Register (n = 144,522). Matching criteria included sex, age, county and year of biopsy.
Sepsis
Through relevant international classifications of disease (ICD) codes (see S2 file), we were able to retrieve data on sepsis of bacterial origin from the Swedish Patient Register for all our study participants. The Patient Register started in 1964 and becoming nationwide in 1987 [24]. Initially, this register only contained inpatient care, but since 2001 the register includes also hospital-based outpatient care [24]. In this study sepsis was defined as having one of the ICD codes presented in the S2 file. Through linkage to the Patient Register we identified 5,470 individuals from the original data set who had received a sepsis diagnosis (CD: n = 1,432 (4.9%) and controls: n = 4,038 (2.8%)).
Mortality
Data on death date and causes of death were retrieved from the Swedish Total Population Register and the Swedish Cause of Death Register, respectively.
Demographic data
Data on socioeconomic status, educational level and country of birth were obtained from Statistics Sweden (the Swedish government agency responsible for producing official statistics). Socioeconomic status and education were divided a priori (for a detailed description, see Olén et al [25]).
Statistics
The risks of overall death and death from sepsis were estimated using Cox regression. Because the original study enrolled individuals at date of biopsy and matched controls that were alive at the corresponding date, we chose a model that allowed for staggered entry. Follow-up ended at death, emigration or end of the study period (Dec 31, 2009) whichever occurred first. HRs were adjusted for sex, age at diagnosis of sepsis, subtype of sepsis, calendar period at sepsis, socioeconomic status, level of education and country of birth. In a separate analysis we adjusted for type 1 diabetes mellitus and autoimmune thyroiditis. Comorbidities were identified from the Patient Register (see S3 file for relevant ICD codes). Individuals with missing data on socioeconomic status and educational level were fitted into a separate category. A priori sub-analyses included stratification for age at diagnosis of sepsis (0–19, 20–39, 40–59, 60–79, 80- years), calendar period (1969–1989, 1990–1999, 2000–2009) and sepsis according to bacterial origin (streptococcal (excluding pneumococcal), pneumococcal, Gram-negative, staphylococcal and non-specified sepsis; all defined according to relevant ICD codes). A separate analysis was also done to examine whether risk estimates changed if we restricted the study population to include only individuals who received a sepsis diagnosis after the CD diagnosis.
We used SPSS 22 (SPSS, Inc. Chicago, IL, USA) and the R survival package for all analyses. HRs with a 95% confidence interval (CI) not including one and p < 0.05 were considered significant.
Results
All participants in this study had a diagnosis of sepsis at some stage during life. Most study participants were diagnosed with sepsis at an older age (about 68% were 60 years or above) and the majority were diagnosed in the last calendar period (2000–2009). Approximately 48% of the study participants were females. Further information on characteristics of the study participants are presented in Table 1. Distribution of sepsis according to bacterial origin in the CD group versus the controls is summarised in Table 2.
All-cause mortality
During follow-up, 3,155 individuals died (841 celiac individuals and 2,314 controls). CD was associated with an increased risk of death from any cause after sepsis (adjusted hazard ratio (aHR) = 1.19; 95%CI = 1.09–1.29). The highest mortality rates was seen in children with CD although the risk estimate did not attain statistical significance (HR = 1.67; 95%CI = 0.67–3.91). Nor was the separate risk increase for adults (≥20 years) statistically significant (HR = 1.03; 95%CI = 0.95–1.12). Having a sepsis diagnosis early in the study period (1969–1989) was associated with the highest relative risk of death (aHR = 1.76; 95%CI = 1.38–2.26). When analysing subtypes of sepsis (i.e. stratified after bacterial origin) we noted an increased overall mortality among individuals with CD following streptococcal, pneumococcal and Gram-negative sepsis. (Table 3)
Restricting our study participants to those with sepsis occurring after CD diagnosis or study entry (controls) we found a slightly lower risk for all-cause mortality (aHR = 1.12; 95%CI = 1.02–1.23). The HRs for children and adults separately were also lower but still not significant (children: 1.31; 95%CI = 0.43–4.0), adults: 0.98; 95%CI = 0.89–1.08). The first two calendar periods (1969–1989 and 1990–1999) in the study were associated with higher risk for death than the last calendar period, however, in this restricted population only the second period had a statistically significant risk increase. (Calendar period 1: aHR = 1.26; 95%CI = 0.71–2.06 and calendar period 2: aHR = 1.25; 95%CI = 1.05–1.48). The risk estimates changed only marginally when we restricted the study population to individuals with sepsis after CD diagnosis (data not shown).
Mortality from sepsis
Of the 5470 individuals with a sepsis diagnosis, 147 had sepsis listed as their cause of death (CD: 33/1432 (4.9%); controls: 114/4038 (2.8%)). All deaths occurred in individuals diagnosed with CD in adulthood. Mortality from sepsis was higher in the CD group than in the non-CD group but the risk estimate failed to reach statistical significance (aHR = 1.10; 95%CI = 0.72–1.69). Risk estimates for death in sepsis according to the most common subtypes of sepsis are listed in Table 3. Because of insufficient power, we were unable to examine the risk for death according to subtype of sepsis. When broadening our definition “death from sepsis” to include not only cases where sepsis had been listed as the underlying cause of death but also cases where it was listed as contributing cause, we found an aHR of 1.15 (95% CI = 0.95–1.40).
Again, restricting our study participants to those who received a sepsis diagnosis after the CD diagnosis, the risk estimate decreased (aHR = 1.04; 95% = 0.67–1.62) and remained statistically non-significant. The relative risks presented in table 3 did not change substantially in this restricted study population
Discussion
In this nationwide study that included nearly 5500 individuals with sepsis, patients with concomitant CD had an increased risk for overall death. This result is consistent with several previous studies, which similarly have found excess mortality in individuals with CD compared with the general population [4–6]. However, the 19% increase in risk for overall death was due to an excess mortality beyond the first month after sepsis. We did not find any increased mortality within 28 days after sepsis; nor did we find an increased mortality from sepsis specifically. Thus CD seems unlikely to influence survival in sepsis.
This study included patients from 1969 and the highest HR for overall death was observed among patients diagnosed with sepsis in the first part of the study period, between 1969 and 1989. This group was accordingly diagnosed with CD early in the study period when the use of serology was less common, the villous atrophy at diagnosis potentially more advanced than today and diagnosed cases were limited to symptomatic CD. Earlier research suggests that CD patients with malabsorption symptoms have a higher mortality than CD patients with minor symptoms [26]; such symptoms were more common in the first part of the study period [27]. Furthermore, research indicates that individuals with CD have lower levels of vitamin D than the general population [28] and vitamin D deficiency has also been linked to excess mortality [29]. Also, low levels of vitamin B have been linked to both untreated and treated CD [30]. In this study we did not have access to individual-based data on vitamin B-levels in the study participants. Although most research does not support a strong association between vitamin B-deficiencies and mortality in sepsis [31, 32], we cannot rule out that nutritional deficiencies may have influenced the sepsis mortality in some of our subanalyses. Even though the clinical presentation of CD has changed during the study period [27] we choose to include all celiac patients with sepsis to increase the statistical power. Restricting our data to the last calendar period, the HR for overall death in sepsis in individuals with CD compared with controls was 1.06 (96%CI = 0.94–1.18). When interpreting our results it is of importance to remember that this paper is a null paper and the potentially more severe disease diagnosed in the first part of our study period has not influenced our overall risk estimates more than marginally. Likewise, we abstained from examining sepsis survival in different pediatric age groups since this would lead to multiple comparisons and a risk of false positive significances.
Naturally, co-morbidities other than CD can influence the outcome in sepsis [19]. In this study we chose to adjust for type 1 diabetes mellitus, which shares several genetic traits with CD and accordingly, is more common in the CD population [33] and autoimmune thyroiditis (also associated with CD) [34]. As in celiac patients, patients with diabetes have an increased risk of acquiring sepsis compared with the general population [35]. However, studies are contradictory regarding whether this autoimmune condition affects the prognosis [36, 37]. In a study from a Taiwanese university hospital comparing 242 diabetic (both type 1 and type 2) patients with bacteraemia and 597 non-diabetic patients with bacteraemia, 30-day mortality rates were similar [37], perhaps suggesting that certain forms of underlying comorbidity are not as important for the outcome in sepsis as is rapid and correct treatment of the sepsis per se. However, an increasing number of pre-existing comorbidities have been linked to a higher sepsis mortality [36, 38].
Although this is a large study, the number of events in each subtype of sepsis was limited; in fact, for some bacteria we were unable to perform sub-analyses. The three bacterial species most strongly associated with an increased mortality in the CD group were streptococcal, pneumococcal and Gram-negative sepsis. However, because of the small numbers of cases in each subtype, we urge caution when interpreting these results. Because of the high prevalence of hyposplenism in the CD population [17] we had expected an excess mortality from pneumococcal sepsis in CD. Studies have shown that individuals with CD seem to have an increased risk for pneumococcal infections [11, 12]. Pneumococci (and Haemophilus influenzae and meningococci) are encapsulated bacteria and the spleen is an important part of the immune system’s defence against such bacteria [39]. Based on our results CD seems to influence the incidence of pneumococcal disease but not its prognosis. Likewise, data has shown that individuals with CD are more prone to acquire sepsis, but again, our results suggest that the outcome of the infection is unaffected by CD. Given the high mortality rate in sepsis, it is possible that the contribution of a concomitant CD diagnosis is too small to have a statistically detectable impact on death rates.
Several studies have evaluated the risk for infections in individuals with CD [11, 12] but to our knowledge this is the first study to focus on survival in sepsis according to CD status. The major strengths of our study are its large size, long follow-up and use of histopathology data to ascertain CD [22]. We used biopsy data from all pathology departments in Sweden to identify a representative cohort with CD patients. Biopsies showing VA was the gold standard for CD diagnosis in Sweden throughout the whole study period. Histopathology examination was carried out by a large number of pathologists (28 pathology departments between 1969 and 2008). It was not possible to carry out an agreement rate for diagnosis between all pathologists, but when representatives from the different pathology departments carried out a blinded test, 90% (95%CI = 87–94%) of biopsies with VA (according to the national steering group of small intestinal pathology) were correctly classified [22]. In most western countries, a substantial proportion of celiac patients remains undiagnosed (and are hence false-negative). However, that proportion is unlikely to constitute more than 1–2% of the control group. That 1–2% of the controls suffer from CD is however not likely to have influenced the base-line survival rate used for comparison in our study (for instance, a survival of 5 years in a false-positive control and 10 years in truly CD-negative controls would yield an average survival of 9.95 years if 1/100 controls and CD).
However the study also has some limitations. We do not have data on dietary adherence. However, since there was no statistically significant difference in risk estimates between the groups that knowledge is of less importance in this paper. Likewise we miss data on medication. Recent data suggests that that olmesartan can cause villous atrophy. Of note, olmesartan is not used in Sweden, and other ACE-inhibitors have not been linked to villous atrophy [40].
While the specificity of a sepsis diagnosis among patients in intensive care units in Sweden is extremely high (>99%) the sensitivity is lower (51%) [41]. A non-optimal sensitivity however, is unlikely to have influenced the risk estimates in our study. It is possible that the diagnostic accuracy for sepsis is better in intensive care units than in other departments, but validation studies of the Swedish Patient Register have shown that in general 85–95% of the diagnoses are correct [24]. Likewise, misclassification of sepsis is unlikely to depend on CD status.
We are not aware of any recent validation of the Swedish Cause of Death Register, but missingness is rare (1.1% in 2013) [42]. A Swedish study from 1995 showed that in a number of cases there was a discrepancy between death certificates and hospital discharge conditions indicating certification errors [43]. However, sepsis was not evaluated and it is unlikely that incorrect cause of death certificates have influenced our mortality rates more than marginally.
In conclusion, this study found an increased overall mortality in patients with sepsis and concomitant CD compared with sepsis patients without CD. Still CD did not influence survival in sepsis per se.
Supporting Information
S1 File. Comparison of small intestinal histopathology classifications.
https://doi.org/10.1371/journal.pone.0154663.s001
(DOCX)
S2 File. International classification codes for sepsis used in this paper.
https://doi.org/10.1371/journal.pone.0154663.s002
(DOCX)
S3 File. International classification of disease (ICD) codes for comorbidities used in this paper.
https://doi.org/10.1371/journal.pone.0154663.s003
(DOCX)
Author Contributions
Conceived and designed the experiments: ART JB JFL. Performed the experiments: ART JB JFL. Analyzed the data: ART JFL. Contributed reagents/materials/analysis tools: JFL. Wrote the paper: ART JB JFL.
References
- 1. Ludvigsson JF, Leffler DA, Bai JC, Biagi F, Fasano A, Green PH, et al. The Oslo definitions for coeliac disease and related terms. Gut. 2013;62(1):43–52. Epub 2012/02/22. gutjnl-2011-301346 [pii] pmid:22345659; PubMed Central PMCID: PMC3440559.
- 2. Ludvigsson JF, Green PH. Clinical management of coeliac disease. J Intern Med. 2011;269(6):560–71. Epub 2011/03/30. pmid:21443532.
- 3. West J, Logan RF, Smith CJ, Hubbard RB, Card TR. Malignancy and mortality in people with coeliac disease: population based cohort study. BMJ. 2004;329(7468):716–9. Epub 2004/07/23. pmid:15269095; PubMed Central PMCID: PMC518895.
- 4. Grainge MJ, West J, Card TR, Holmes GK. Causes of death in people with celiac disease spanning the pre- and post-serology era: a population-based cohort study from Derby, UK. Am J Gastroenterol. 2011;106(5):933–9. Epub 2011/01/20. pmid:21245833.
- 5. Solaymani-Dodaran M, West J, Logan RF. Long-term mortality in people with celiac disease diagnosed in childhood compared with adulthood: a population-based cohort study. Am J Gastroenterol. 2007;102(4):864–70. Epub 2007/02/28. pmid:17324126.
- 6. Ludvigsson JF, Montgomery SM, Ekbom A, Brandt L, Granath F. Small-intestinal histopathology and mortality risk in celiac disease. JAMA. 2009;302(11):1171–8. Epub 2009/09/17. 302/11/1171 [pii] pmid:19755695.
- 7. Iqbal T, Zaidi MA, Wells GA, Karsh J. Celiac disease arthropathy and autoimmunity study. Journal of gastroenterology and hepatology. 2013;28(1):99–105. Epub 2012/09/20. pmid:22988822.
- 8. Walters JR, Bamford KB, Ghosh S. Coeliac disease and the risk of infections. Gut. 2008;57(8):1034–5. Epub 2008/04/11. pmid:18400879.
- 9. Ludvigsson JF, Sanders DS, Maeurer M, Jonsson J, Grunewald J, Wahlstrom J. Risk of tuberculosis in a large sample of patients with coeliac disease—a nationwide cohort study. Aliment Pharmacol Ther. 2011;33(6):689–96. Epub 2011/01/22. pmid:21251029.
- 10. Marild K, Fredlund H, Ludvigsson JF. Increased risk of hospital admission for influenza in patients with celiac disease: a nationwide cohort study in Sweden. Am J Gastroenterol. 2010;105(11):2465–73. Epub 2010/09/09. ajg2010352 [pii] pmid:20823839.
- 11. Thomas HJ, Wotton CJ, Yeates D, Ahmad T, Jewell DP, Goldacre MJ. Pneumococcal infection in patients with coeliac disease. Eur J Gastroenterol Hepatol. 2008;20(7):624–8. Epub 2008/08/06. 00042737-200807000-00006 [pii]. pmid:18679063.
- 12. Ludvigsson JF, Olen O, Bell M, Ekbom A, Montgomery SM. Coeliac disease and risk of sepsis. Gut. 2008;57(8):1074–80. Epub 2008/02/14. gut.2007.133868 [pii] pmid:18270242.
- 13. Peters U, Askling J, Gridley G, Ekbom A, Linet M. Causes of death in patients with celiac disease in a population-based Swedish cohort. Arch Intern Med. 2003;163(13):1566–72. Epub 2003/07/16. pmid:12860579.
- 14. Jawad I, Luksic I, Rafnsson SB. Assessing available information on the burden of sepsis: global estimates of incidence, prevalence and mortality. Journal of global health. 2012;2(1):10404. Epub 2012/12/01. pmid:23198133; PubMed Central PMCID: PMC3484761.
- 15. Dellinger RP, Levy MM, Rhodes A, Annane D, Gerlach H, Opal SM, et al. Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock, 2012. Intensive care medicine. 2013;39(2):165–228. Epub 2013/01/31. pmid:23361625.
- 16. Martin GS. Sepsis, severe sepsis and septic shock: changes in incidence, pathogens and outcomes. Expert review of anti-infective therapy. 2012;10(6):701–6. Epub 2012/06/28. pmid:22734959; PubMed Central PMCID: PMC3488423.
- 17. Di Sabatino A, Rosado MM, Cazzola P, Riboni R, Biagi F, Carsetti R, et al. Splenic hypofunction and the spectrum of autoimmune and malignant complications in celiac disease. Clin Gastroenterol Hepatol. 2006;4(2):179–86. Epub 2006/02/14. S1542-3565(05)00982-1 [pii]. pmid:16469678.
- 18. Sapone A, Lammers KM, Casolaro V, Cammarota M, Giuliano MT, De Rosa M, et al. Divergence of gut permeability and mucosal immune gene expression in two gluten-associated conditions: celiac disease and gluten sensitivity. BMC medicine. 2011;9:23. Epub 2011/03/12. pmid:21392369; PubMed Central PMCID: PMC3065425.
- 19. Dhainaut JF, Claessens YE, Janes J, Nelson DR. Underlying disorders and their impact on the host response to infection. Clin Infect Dis. 2005;41 Suppl 7:S481–9. Epub 2005/10/21. pmid:16237651.
- 20. Christensen JS, Jensen TG, Kolmos HJ, Pedersen C, Lassen A. Bacteremia with Streptococcus pneumoniae: sepsis and other risk factors for 30-day mortality—a hospital-based cohort study. European journal of clinical microbiology & infectious diseases: official publication of the European Society of Clinical Microbiology. 2012;31(10):2719–25. Epub 2012/05/15. pmid:22581362.
- 21. Ludvigsson JF, Otterblad-Olausson P, Pettersson BU, Ekbom A. The Swedish personal identity number: possibilities and pitfalls in healthcare and medical research. Eur J Epidemiol. 2009;24(11):659–67. Epub 2009/06/09. pmid:19504049; PubMed Central PMCID: PMC2773709.
- 22. Ludvigsson JF, Brandt L, Montgomery SM, Granath F, Ekbom A. Validation study of villous atrophy and small intestinal inflammation in Swedish biopsy registers. BMC Gastroenterol. 2009;9:19. Epub 2009/03/17. 1471-230X-9-19 [pii] pmid:19284576; PubMed Central PMCID: PMC2664822.
- 23. Ludvigsson JF, Rubio-Tapia A, van Dyke CT, Melton LJ 3rd, Zinsmeister AR, Lahr BD, et al. Increasing incidence of celiac disease in a North American population. Am J Gastroenterol. 2013;108(5):818–24. Epub 2013/03/21. pmid:23511460; PubMed Central PMCID: PMC3686116.
- 24. Ludvigsson JF, Andersson E, Ekbom A, Feychting M, Kim JL, Reuterwall C, et al. External review and validation of the Swedish national inpatient register. BMC Public Health. 2011;11:450. Epub 2011/06/11. 1471-2458-11-450 [pii] pmid:21658213; PubMed Central PMCID: PMC3142234.
- 25. Olen O, Bihagen E, Rasmussen F, Ludvigsson JF. Socioeconomic position and education in patients with coeliac disease. Dig Liver Dis. 2012;44(6):471–6. Epub 2012/02/22. S1590-8658(12)00024-2 [pii] pmid:22341742.
- 26. Corrao G, Corazza GR, Bagnardi V, Brusco G, Ciacci C, Cottone M, et al. Mortality in patients with coeliac disease and their relatives: a cohort study. Lancet. 2001;358(9279):356–61. Epub 2001/08/15. pmid:11502314.
- 27. Rampertab SD, Pooran N, Brar P, Singh P, Green PH. Trends in the presentation of celiac disease. The American journal of medicine. 2006;119(4):355 e9–14. Epub 2006/03/28. pmid:16564784.
- 28. Zanchi C, Di Leo G, Ronfani L, Martelossi S, Not T, Ventura A. Bone metabolism in celiac disease. J Pediatr. 2008;153(2):262–5. Epub 2008/06/07. S0022-3476(08)00186-8 [pii] pmid:18534236.
- 29. Schottker B, Jorde R, Peasey A, Thorand B, Jansen EH, Groot L, et al. Vitamin D and mortality: meta-analysis of individual participant data from a large consortium of cohort studies from Europe and the United States. BMJ. 2014;348:g3656. Epub 2014/06/19. pmid:24938302; PubMed Central PMCID: PMC4061380.
- 30. Theethira TG, Dennis M, Leffler DA. Nutritional consequences of celiac disease and the gluten-free diet. Expert review of gastroenterology & hepatology. 2014;8(2):123–9. pmid:24417260.
- 31. Tsantes A, Tsangaris I, Nikolopoulos G, Bagos P, Kopterides P, Antonakos G, et al. The effect of homocysteine on the clinical outcomes of ventilated patients with severe sepsis. Minerva Anestesiol. 2010;76(10):787–94. pmid:20935614.
- 32. Ploder M, Kurz K, Spittler A, Neurauter G, Roth E, Fuchs D. Early increase of plasma homocysteine in sepsis patients with poor outcome. Mol Med. 2010;16(11–12):498–504. pmid:20386870; PubMed Central PMCID: PMCPMC2972395.
- 33. Cohn A, Sofia AM, Kupfer SS. Type 1 diabetes and celiac disease: clinical overlap and new insights into disease pathogenesis. Current diabetes reports. 2014;14(8):517. Epub 2014/06/22. pmid:24952108; PubMed Central PMCID: PMC4156157.
- 34. Ventura A, Magazzu G, Greco L. Duration of exposure to gluten and risk for autoimmune disorders in patients with celiac disease. SIGEP Study Group for Autoimmune Disorders in Celiac Disease. Gastroenterology. 1999;117(2):297–303. Epub 1999/07/27. pmid:10419909.
- 35. Thomsen RW, Riis AH, Kjeldsen S, Schonheyder HC. Impact of diabetes and poor glycaemic control on risk of bacteraemia with haemolytic streptococci groups A, B, and G. The Journal of infection. 2011;63(1):8–16. Epub 2011/06/15. pmid:21663970.
- 36. Ayala-Ramirez OH, Dominguez-Berjon MF, Esteban-Vasallo MD. Trends in hospitalizations of patients with sepsis and factors associated with inpatient mortality in the Region of Madrid, 2003–2011. European journal of clinical microbiology & infectious diseases: official publication of the European Society of Clinical Microbiology. 2014;33(3):411–21. Epub 2013/10/01. pmid:24078023.
- 37. Tsai CL, Lee CC, Ma MH, Fang CC, Chen SY, Chen WJ, et al. Impact of diabetes on mortality among patients with community-acquired bacteremia. The Journal of infection. 2007;55(1):27–33. Epub 2007/03/14. pmid:17349696.
- 38. Oltean S, Tatulescu D, Bondor C, Slavcovici A, Cismaru C, Lupse M, et al. Charlson's weighted index of comorbidities is useful in assessing the risk of death in septic patients. Journal of critical care. 2012;27(4):370–5. Epub 2011/10/29. pmid:22033051.
- 39. Di Sabatino A, Carsetti R, Corazza GR. Post-splenectomy and hyposplenic states. Lancet. 2011;378(9785):86–97. Epub 2011/04/09. pmid:21474172.
- 40. Marild K, Lebwohl B, Green PH, Murray JA, Ludvigsson JF. Blockers of Angiotensin Other Than Olmesartan in Patients With Villous Atrophy: A Nationwide Case-Control Study. Mayo Clinic proceedings. 2015;90(6):730–7. Epub 2015/06/06. pmid:26046408; PubMed Central PMCID: PMC4469984.
- 41. Gedeborg R, Furebring M, Michaelsson K. Diagnosis-dependent misclassification of infections using administrative data variably affected incidence and mortality estimates in ICU patients. Journal of clinical epidemiology. 2007;60(2):155–62. Epub 2007/01/09. pmid:17208121.
- 42.
Welfare TSNBoHa. Causes of Death 2013. In: Welfare TSNBoHa, editor. The Swedish National Board of Health and Welfare: The Swedish National Board of Health and Welfare; 2013. p. 8–9.
- 43. Johansson LA, Bjorkenstam C, Westerling R. Unexplained differences between hospital and mortality data indicated mistakes in death certification: an investigation of 1,094 deaths in Sweden during 1995. Journal of clinical epidemiology. 2009;62(11):1202–9. Epub 2009/04/15. pmid:19364635.