DOI: 10.1111/eci.12242

ORIGINAL ARTICLE Ischaemic heart disease in first-degree relatives to coeliac patients Louise Emilsson*,†, Stefan James‡ and Jonas F. Ludvigsson§,¶ * € €rmlands Nysa €ter, Va €rmland County, Sweden, †Department of Medicine, Orebro Primary care research unit, V
ardcentralen Va € University, Orebro, Sweden, ‡Department of Medical Sciences, Cardiology and Uppsala Clinical Research Center, Uppsala University, Uppsala, Sweden, §Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, € € Sweden, ¶Department of Pediatrics, Orebro University Hospital, Orebro, Sweden

ABSTRACT Objective Coeliac disease (CD) has been linked to an increased risk of ischaemic heart disease (IHD). We examined the risk of IHD in first-degree relatives and spouses to coeliac patients to ascertain the genetic contribution to IHD excess risk. Study design and setting Coeliac disease was defined as having a biopsy-verified villous atrophy (Marsh grade 3) in 1969–2008 (n = 29 096). Coeliac patients were matched to 144 522 controls. Through Swedish registers, we identified all first-degree relatives and spouses to coeliac patients and their controls, in total 87 622 unique coeliac relatives and 432 655 unique control relatives. Our main outcome measure was IHD defined according to relevant international classification of disease codes in the Swedish Inpatient Registry or in the Cause of Death Registry. Hazard ratios (HR) and confidence intervals (CI) were estimated through Cox regression adjusted for sex, age-group and calendar year at study entry of the relative. Result During a median follow-up of 10
8 years, 2880 coeliac relatives and 13 817 control relatives experienced IHD. First-degree relatives of coeliac patients were at increased risk of IHD (HR = 1
05; 95% CI = 1
00–1
09, P-value = 0
04), while spouses were at no increased risk (HR = 0
99; 95% CI = 0
87–1
12). The excess risk of IHD in coeliac first-degree relatives aged 40–59 years was 70/100 000 person-years. Conclusion First-degree relatives to coeliac patients seem to be at an increased risk of IHD but the excess risk is so small that it has little clinical relevance. Keywords Celiac, coeliac, cohort, ischaemic heart disease, risk factors, shared genetics. Eur J Clin Invest 2014; 44 (4): 359–364

Introduction

Methods

Many [1–3] but not all [4] studies suggest that patients with coeliac disease (CD) are at increased risk of ischaemic heart disease (IHD). The underlying causes of such an excess risk are unknown. We have earlier shown that coeliac patients developing myocardial infarction (MI) have a more favourable classical risk factor profile compared with controls developing MI [5], and hence, confounding by other well-known risk factors is unlikely to explain the association between CD and IHD. Other potential explanations for the excess risk of IHD in CD include systemic inflammation, shared genetics, the gluten-free diet and ascertainment bias. This study was performed to address the genetic impact on IHD risk in CD through evaluation of IHD frequency in relatives to coeliac patients compared to relatives to noncoeliac controls. This is the first study to address IHD risk in first-grade relatives of coeliac patients.

We calculated the risk of IHD in coeliac relatives vs. control relatives.

Collection of biopsy data We collected data on all computerized duodenal/jejunal biopsies between 1969 and 2008 from all Swedish pathology departments. Biopsies classified with villous atrophy equal to histopathology stage Marsh III[6] were considered as CD with date of first pathological biopsy equal to date of diagnosis and study entry. In total, 29 096 coeliac patients were identified. Small-intestinal biopsy is clinical routine in Sweden (96% of adult gastroenterologists and 100% of paediatricians perform a biopsy in at least 9/10 patients with suspected CD before diagnosis)[7], and more than 95% of individuals with Marsh III changes have CD in a Swedish setting [7].

European Journal of Clinical Investigation Vol 44

359

L. EMILSSON ET AL.

www.ejci-online.com

Reference individuals (controls) Using the Swedish Total Population Register, all coeliac patients were matched for sex, county, age and calendar year of birth, with up to five noncoeliac controls by the government agency Statistics Sweden [8]. In total, we identified 144 522 controls. All controls entered the study on the same date as their matched coeliac patient (the date of Marsh III biopsy). The original CD and controls cohorts have been described in detail elsewhere [9].

First-degree relatives and spouses Through the Swedish Multi Generation Register [10], we obtained data on all first-degree relatives (mother, father, sibling and offspring) to coeliac patients and controls, and through the Total Population Register [8], we obtained data on all registered spouses (defined through marriage) to CD and controls individuals (Fig. 1). Spouses were entered in the study to represent genetically different individuals sharing the same environment as the coeliac/index patients. All relatives and spouses entered the study at date of the corresponding index individuals’ study entry or at birth whichever occurred latest (Fig. 1).

Exposure Exposure was defined as being a first-degree relatives or spouse to a coeliac patient.

Outcome measure

First-degree relatives of CD patients

First-degree relatives of controls

Matched controls

Figure 1

360

Swedish total population register

Swedish total population register Spouses of CD patients

Spouses of controls

Follow up until: Death registered in Swedish Cause of death register. Emigration registered in the Swedish population register. Event registered in the Swedish patient registrer. End of follow-up on 31st of December 2008.

CD patients, defined by biopsy reports

Swedish Multi-generation register

The following ICD codes were used to define MI: ICD-8: 410, ICD-9: 410 and ICD-10: I21-22, AP: ICD-8: 411, ICD-9: 411B, 413, 414A and ICD-10: I20 and unstable AP (ICD-8: 411, ICD-9: 411B and ICD-10: I20
0). Data on incident disease were obtained from the Swedish Patient Registry [11], and data on ischaemic

Linkage of registries. CD, coeliac disease.

deaths were obtained from the national Cause of Death Registry [12]. All ischaemic outcomes (IHD death, myocardial infarction, unstable angina and stable angina) were included in the composite main outcome measure called any IHD, but were also analysed separately.

Statistical analyses We used Cox regression to estimate hazard ratios (HR) adjusted for sex, age-group and calendar year. Follow-up started at study entry of the relatives and ended at IHD outcome, death, emigration or on 31st of December 2008 whichever occurred first (Fig. 1). Individuals with an event of any IHD prior to the study entry were excluded. We analysed strata of relatives separately (mother, father, sibling, offspring and partner) and together (excluding spouses). In a secondary analysis, we examined our main outcome measure (any IHD) in first-degree relatives using an internal stratification when firstdegree relatives to each coeliac patient were only compared with the first-degree relatives of the matched controls of that very coeliac patient. This increases homogeneity and ensures an even better matching but at the expense of slightly lower statistical power due to fewer participants (some strata may lack control relatives and are then excluded from the analysis). We also examined the different IHD outcomes separately (MI death, MI, unstable angina and stable angina) as well as absolute and excess risk for different age categories, and we adjusted for CD in the first-degree relatives. In addition, we used a case-control approach and examined the risk of IHD prior to CD in first-degree relatives and spouses using conditional logistic regression. The risk of earlier IHD was presented as odds ratios (ORs) adjusted for sex, county, age and calendar year of birth. Finally, we performed a post hoc analysis stratifying on early (during the first 2 years) and late (more than 2 years after study entry) IHD events as well as analysing our main outcome measure adjusted for number of hospitalizations in the relatives imputed as a continuous variable. Statistical significance was defined as 95% confidence intervals (CIs) for risk estimates not including 1
0. We used SPSS version 21.0 (SPSS Inc, Chicago, IL, USA) for all analyses.

Results Background data In total, we had data on 87 622 unique coeliac relatives followed for 977 536 person-years and 432 655 control relatives followed for 4 819 431 person-years (Table 1). The majority of the relatives were siblings or offspring, and 61
7% (coeliac relatives) vs. 61
9% (control relatives) were male (Table 1). In total, 2880 ischaemic events occurred in coeliac relatives and 13 087 events in control relatives. Ischaemic events were most common in fathers (Table 1).

ª 2014 Stichting European Society for Clinical Investigation Journal Foundation

IHD IN CELIAC RELATIVES

Table 1 Characteristics of coeliac patient relatives and reference individual relatives

Number, total

Coeliac relative, N (%)

Reference individual relative, N (%)

87 622 (100)*

432 655 (100)*

17 328 (18
8)

84 377 (18
9)

9417 (10
2)

47 374 (10
6)

Relation Mother Father Sibling

31 682 (34
4)

148 072 (33
2)

Offspring

27 126 (29
4)

133 445 (29
9)

6665 (7
2)

32 572 (7
3)

Male

56 936 (61
7)

276 025 (61
9)

Female

35 282 (38
3)

169 815 (38
1)

–1939

13 186 (14
3)

64 115 (14
4)

1940–1963

33 028 (35
8)

162 594 (36
5)

1964–1986

23 140 (25
1)

110 536 (24
8)

1987–2010

22 861 (24
8)

108 548 (24
3)

–1989

11 720 (12
7)

57 102 (12
8)

1990–1999

36 157 (39
2)

175 986 (39
5)

2000–2010

44 341 (48
1)

212 752 (47
7)

Spouse Gender

Age-group (birthdate)

Calendar year study entry

excess risk was only significant in siblings. When we analysed first-degree relatives with internal stratification, the HR diminished to 1
04 and lost statistical significance (Table 2). Coeliac spouses were at no increased risk of IHD compared with control spouses (controls HR = 0
99; Table 2).

MI death, incident MI, unstable angina and stable angina Coeliac first-degree relatives were at no increased risk of MI death, but were more often admitted to hospital with MI (HR = 1
05; 95% CI = 1
00–1
10). This latter risk increase remained statistically significant also when we examined the association using internal stratification (Table 2). With few exceptions, the HRs were above one in coeliac first-degree relatives (exceptions: unstable angina in offspring (HR = 0
91) and MI death in mothers (HR = 1
00), whereas the nonsignificant HRs for spouses were as follows: MI inpatient 0
99, unstable angina 0
80 and stable angina 1
09 (there were no events of MI death in spouses) (Table 2). Adjustment for CD in the firstdegree relatives did not change the risk estimate (HR = 1
05; 95% CI = 1
00–1
09).

Absolute and excess risks The absolute risks of any IHD in coeliac first-degree relatives were 42, 435 and 1440 per 100 000 person-years for the age-groups 0–39, 40–59 and 60 and above respectively. This corresponded to an excess risk of 20, 70 and 36 per 100 000 person-years for the different age-groups. For spouses, the absolute risk was similar and excess risks were small (Table 3).

Any IHD prior study entry

Events (any IHD)

The conditional logistic regression found an increased risk of earlier IHD in coeliac first-degree relatives (OR = 1
04; 95% CI = 1
01–1
06). The corresponding OR for spouses was 1
01 (95% CI = 0
90–1
12).

Mother

1035 (6
0)

5171 (6
1)

Father

761 (8
1)

3551 (7
5)

Sibling

688 (2
1)

2731 (1
8)

Offspring

233 (0
8)

1093 (0
8)

Post hoc analyses

Partner

302 (4
5)

1535 (4
7)

Ischaemic heart disease events during the first 2 years from study entry in all first-grade relatives showed HR = 1
12 (95% CI = 1
01–1
25), whereas the HR for IHD events more than 2 years after study entry was 1
03 (95% CI = 0
98–1
08). The HR for IHD events adjusted for number of hospitalizations was 1
04 (95% CI = 1
00–1
08, P-value 0
04).

Total unique events

2880

13 087

Median

10
8

10
8

Range

0–42

0–42

Follow-up time (years)

*The sum is greater than the total number as you can be both mother and sibling.

Any IHD Coeliac first-degree relatives were at an increased risk of IHD. The estimate was statistically significant by a narrow margin (HR 1
05; 95% CI = 1
00–1
09, P = 0
04; Table 2). The IHD

Discussion To our knowledge, this is the first study to examine the risk of IHD in coeliac relatives. Most earlier studies on IHD in coeliac patients have found a positive association [1–3,9], but exceptions exist [4]. In 2011, we reported a 1
19-fold increased risk of IHD in Swedish patients with biopsy-verified CD[2]. That

European Journal of Clinical Investigation Vol 44

361

L. EMILSSON ET AL.

www.ejci-online.com

Table 2 Results MI death, HR (CI)

MI inpatient, HR (CI)

Unstable AP inpatient, HR (CI)

Stable AP inpatient, HR (CI)

Any IHD, HR (CI)

All first-grade relatives

1
02 (0
93–1
12)

1
05 (1
00–1
10)*

1
09 (0
99–1
20)

1
07 (1
01–1
13)

1
05 (1
00–1
09)*

All first-grade relatives (analysed in coeliac/ index patient strata)

1
08 (0
97–1
19)

1
06 (1
01–1
11)*

1
06 (0
96–1
18)

1
05 (0
99–1
12)

1
04 (0
99–1
09)

Mother

1
00 (0
87–1
15)

1
02 (0
95–1
09)

1
10 (0
95–1
28)

1
02 (0
92–1
12)

1
01 (0
94–1
08)

Father

1
11 (0
96–1
27)

1
08 (0
99–1
18)

1
10 (0
88–1
36)

1
10 (0
97–1
24)

1
08 (0
99–1
16)

Sibling

1
06 (0
80–1
42)

1
08 (0
99–1
18)

1
07 (0
90–1
28)

1
09 (0
98–1
22)

1
09 (1
00–1
19)*

Offspring

1
23 (0
80–1
90)

1
05 (0
90–1
22)

0
91 (0
67–1
27)

1
07 (0
89–1
30)

1
07 (0
93–1
24)

Spouse

No events

0
99 (0
87–1
12)

0
80 (0
61–1
07)

1
09 (0
94–1
27)

0
99 (0
87–1
12)

AP, Angina pectoris; CD, Coeliac Disease; MI, Myocardial Infarction; HR, Hazard Ratio. *Results in bold italics were statistically significant (P < 0
05).

Table 3 Absolute and excess risk

Age-category

Observed events

Expected events

Absolute risk per 100 000 person-years

Person-years at risk

HR; 95% CI

Excess risk per 100 000 person-years

First-degree relatives 0–39

271

251

644 345

42

1
08; 0
95–1
23

20

40–59

1 075

1005

246 988

435

1
07; 1
00–1
14 (P = 0
06)

70

60–

1 241

1205

86 203

1440

1
03; 0
97–1
10

36

0–39

23

15

29 657

77

1
49; 0
94–2
37

8

40–59

155

168

39 948

388

0
92; 0
77–1
09

–13

60–

115

114

12 687

906

1
01; 0
82–1
23

1

Spouses

HR, hazard ratio; CI, confidence interval.

study found a positive association both with MI and angina pectoris. Risk estimates were however highest in the first year of follow-up (HR = 1
77) signalling that either acute inflammation or ascertainment bias may have influenced the risk estimates. While acute inflammation is unlikely to explain the positive association in our study, we cannot fully rule out that surveillance bias in first-degree relatives has contributed to the excess risk seen in the current study, and the risk estimate being higher for the first 2 years after study entry supports this explanation. Against this argues the complete lack of excess risk in spouses that share the environment of the coeliac patient and is often involved in the health care of the patient. Not even in the first year of follow-up did we note an increased risk of IHD in spouses.

362

The impact of shared genetics for IHD risk in CD has been suggested earlier [13] but this is the first study we know of addressing this issue. The most important genetic changes in CD are found on the HLA class II region on the chromosome 6p21
3[14], accounting for about 35% of the genetic risk in CD [15] although many more risk loci are identified, but each of them only contributing a small portion of the heritability in CD [16]. There has been a wide range of studies covering genetic contribution in IHD. During last year, a GWAS study identified a loci associated with increased risk of coronary artery disease within the MCH loci on 6p21
3[17]. The authors statistically imputed the MHC class I molecules without finding the explanation of the association and therefore concluded that further studies were needed to establish the causative alleles. It

ª 2014 Stichting European Society for Clinical Investigation Journal Foundation

IHD IN CELIAC RELATIVES

can be speculated that the proximity of the associated loci or the MCH class II alleles associated with CD themselves confer an increased risk of coronary heart disease and atherosclerosis, and MCH class II variants have also been identified in smaller studies of single-nucleotide polymorphisms earlier [18]. Shared genetics offers a plausible biological explanation of the increased risk found independent of temporality (HR = 1
05 for IHD following relative CD and OR = 1
04 for CD following IHD in conditional logistic regression) although the excess risk is small. Another feature of the genetic impact in CD is that a number of first-degree relatives may themselves have CD[19]. This is however unlikely to influence our risk estimates more than marginally considering that 9/10 first-degree relatives do not have CD and that adjustment for CD did not change the estimate. The results in our study are difficult to interpret as the risk increase identified is minimal but statistically significant for the main analysis. HRs for subtypes of ischaemia are with two exceptions (IHD death in mothers and unstable angina in offspring) positive (above one) and several of them are of borderline significance (Table 2). For spouses on the other hand, the main outcome is perfectly neutral, HR = 0
99, and the subanalyses are both positive and negative. This data coherence suggests that there might be a small genetic contribution in the increased risk of IHD in CD. Also, the risk of prior IHD in firstdegree relatives of coeliac patients was significantly increased, whereas there was no excess risk among spouses. On the other hand, the increased risk in first-degree relatives might equally likely be due to surveillance bias. The finding of no increased risk in spouses also indicates that the environment is not responsible for the increased risk of IHD seen in coeliac patients.

Strengths and limitations One obvious strength of our study is the population-based setting. During the study period, some 2880 coeliac first-degree relatives had an IHD event. Corresponding numbers for control relatives are about fourfold (Tables 1 and 3). Furthermore, data were collected prospectively, and this minimizes the risk of selection bias. We used biopsy registers to collect information about CD. Biopsy registers have a high sensitivity for diagnosed CD as 96– 100% of all physicians diagnosing CD perform a biopsy before diagnosis [7]. When we manually examined more than 1500 biopsy reports with villous atrophy (Marsh 3) or inflammation (Marsh 2), other diagnoses than CD were very rare; the most frequent being inflammatory bowel disease found in 0
3% of reports with villous atrophy [7]. Earlier validation has found a high positive predictive value for MI in the Swedish Patient Registry [11,20]. We chose to

identify patients with IHD both in the Patient Registry and in the Cause of Death Registry. The Patient Registry began in 1964 and the Cause of Death Registry in the 18th century, however computerized in 1961. Through linkage with the Personal Identity Number [21], virtually complete follow-up of subjects can be guaranteed. The Swedish healthcare system is publicly funded which means that patients with IHD have high access to health care independently of socioeconomic characteristics. A limitation in this study is our lack of data on other IHD risk factors such as smoking, hypertension, cholesterol levels or body mass index (BMI). Earlier data on an inverse relationship between CD and smoking [22,23], hypertension [24], hypercholesterolaemia [25] and obesity [26] do however suggest that confounding by these factors is unlikely to explain a positive association with IHD in coeliac relatives, provided that such background variables are less common in the relatives as well. While excess risks were small in this study, should they be confirmed in other studies, they have some implications. IHD is one of the commonest causes of death and morbidity in the Western world [27]. In this study, the excess risk of IHD in coeliac patients aged 40–59 years was 70/100 000 person-years, and 36/100 000 person-years in those aged 60+ years, so even small excess risks of IHD have clinical implications.

Conclusion We conclude that being a first-degree relative to a coeliac patient is not associated with any clinically relevant increased risk of IHD, but that a genetic contribution to the increased risk of IHD in CD cannot be ruled out.

Source of funding LE was supported by a grant from Va¨rmland County. JFL was supported by a grant from The Swedish Society of Medicine, ¨ rebro Society of Medicine the Swedish Research Council, the O and the Swedish Celiac Society. None of the funders had any role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the manuscript. Conflicts of interest No potential conflicts of interest.

Data sharing No additional data.

Details of ethics approval This project (2006/633-31/3) was approved by the Research Ethics Committee of the Karolinska Institute, Sweden on June 14, 2006.

European Journal of Clinical Investigation Vol 44

363

L. EMILSSON ET AL.

Address Primary care research unit, V
ardcentralen V€armlands Nys€ ater, Nystr€ oms v€ag 5, 66195 V€armlands Nys€ater, Sweden (L. Emilsson); Department of Medical Sciences, Cardiology and Uppsala Clinical Research Center, Uppsala University, Uppsala, Sweden (S. James); Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, Sweden (J.F. Ludvigsson). Correspondence to: Louise Emilsson, V
ardcentralen V€ armlands Nys€ater, Nystr€ oms v€ag 5, 66195 V€armlands Nys€ ater, Sweden. Tel.: +46 (0) 533-81013; fax: +46 (0) 533-30214; e-mail: [email protected] Received 1 October 2013; accepted 14 January 2014 References 1 Viljamaa M, Kaukinen K, Pukkala E, Hervonen K, Reunala T, Collin P. Malignancies and mortality in patients with coeliac disease and dermatitis herpetiformis: 30-year population-based study. Dig Liver Dis 2006;38:374–80. 2 Ludvigsson JF, James S, Askling J, Stenestrand U, Ingelsson E. Nationwide cohort study of risk of ischemic heart disease in patients with celiac disease. Circulation 2011;123:483–90. 3 Wei L, Spiers E, Reynolds N, Walsh S, Fahey T, MacDonald TM. The association between coeliac disease and cardiovascular disease. Aliment Pharmacol Ther 2008;27:514–9. 4 West J, Logan RF, Card TR, Smith C, Hubbard R. Risk of vascular disease in adults with diagnosed coeliac disease: a population-based study. Aliment Pharmacol Ther 2004;20:73–9. 5 Emilsson L, Carlsson R, Holmqvist M, James S, Ludvigsson JF. The characterisation and risk factors of ischaemic heart disease in patients with coeliac disease. Aliment Pharmacol Ther 2013;37:905–14. 6 Marsh MN. Gluten, major histocompatibility complex, and the small intestine. A molecular and immunobiologic approach to the spectrum of gluten sensitivity (‘celiac sprue’). Gastroenterology 1992;102:330–54. 7 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. 8 Johannesson I. The Total Population Register of Statistics Sweden. New € Possibilities and Better Quality. Orebro: Statistics Sweden; 2002. 9 Ludvigsson JF, Montgomery SM, Ekbom A, Brandt L, Granath F. Small-intestinal histopathology and mortality risk in celiac disease. JAMA 2009;302:1171–8. 10 Ekbom A. The Swedish Multi-generation Register. Methods Mol Biol 2011;675:215–20. 11 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.

364

www.ejci-online.com

12 Johansson LA, Westerling R. Comparing Swedish hospital discharge records with death certificates: implications for mortality statistics. Int J Epidemiol 2000;29:495–502. 13 Lagro J, Duijts L, Jaddoe VW. Letter by Lagro et al. regarding article, “Nationwide cohort study of risk of ischemic heart disease in patients with celiac disease”. Circulation 2011;124:e422; author reply e3. 14 Lavant EH, Carlson J. HLA DR-DQ genotyping by capillary electrophoresis for risk assessment for celiac disease. Methods Mol Biol 2013;919:297–307. 15 Dubois PC, Trynka G, Franke L, Hunt KA, Romanos J, Curtotti A et al. Multiple common variants for celiac disease influencing immune gene expression. Nat Genet 2010;42:295–302. 16 Trynka G, Wijmenga C, van Heel DA. A genetic perspective on coeliac disease. Trends Mol Med 2010;16:537–50. 17 Davies RW, Wells GA, Stewart AF, Erdmann J, Shah SH, Ferguson JF et al. A genome-wide association study for coronary artery disease identifies a novel susceptibility locus in the major histocompatibility complex. Circ Cardiovasc Genet 2012 Apr 1;5:217– 25. 18 Hansson GK, Hermansson A. The immune system in atherosclerosis. Nat Immunol 2011;12:204–12. 19 Rubio-Tapia A, Van Dyke CT, Lahr BD, Zinsmeister AR, ElYoussef M, Moore SB et al. Predictors of family risk for celiac disease: a population-based study. Clin Gastroenterol Hepatol 2008;6:983–7. 20 Elo SL, Karlberg IH. Validity and utilization of epidemiological data: a study of ischaemic heart disease and coronary risk factors in a local population. Public Health 2009;123:52–7. 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;24:659–67. 22 Snook JA, Dwyer L, Lee-Elliott C, Khan S, Wheeler DW, Nicholas DS. Adult coeliac disease and cigarette smoking. Gut 1996;39:60–2. 23 Vazquez H, Smecuol E, Flores D, Mazure R, Pedreira S, Niveloni S et al. Relation between cigarette smoking and celiac disease: evidence from a case-control study. Am J Gastroenterol 2001;96:798– 802. 24 West J, Logan RF, Hill PG, Lloyd A, Lewis S, Hubbard R et al. Seroprevalence, correlates, and characteristics of undetected coeliac disease in England. Gut 2003;52:960–5. 25 Lewis NR, Sanders DS, Logan RF, Fleming KM, Hubbard RB, West J. Cholesterol profile in people with newly diagnosed coeliac disease: a comparison with the general population and changes following treatment. Br J Nutr 2009;102:509–13. 26 Olen O, Montgomery SM, Marcus C, Ekbom A, Ludvigsson JF. Coeliac disease and body mass index: a study of two Swedish general population-based registers. Scand J Gastroenterol 2009;44:1198–206. 27 Kim AS, Johnston SC. Global variation in the relative burden of stroke and ischemic heart disease. Circulation 2011;124:314–23.

ª 2014 Stichting European Society for Clinical Investigation Journal Foundation