You are here

  • Home
  • Archive
  • Volume 64, Issue 2
  • Risk of irritable bowel syndrome in first-degree, second-degree and thirddegree relatives of affected individuals: a nationwide family study in Sweden

Article Text

Article menu
Neurogastroenterology
Original article
Risk of irritable bowel syndrome in first-degree, second-degree and thirddegree relatives of affected individuals: a nationwide family study in Sweden
  1. Rasmus Waehrens1,
  2. Henrik Ohlsson1,
  3. Jan Sundquist1,2,
  4. Kristina Sundquist1,2,
  5. Bengt Zöller1
  1. 1Center for Primary Health Care Research, Lund University/Region Skåne, Malmö, Sweden
  2. 2Stanford Prevention Research Center, Stanford University School of Medicine, Palo Alto, California, USA
  1. Correspondence to Dr Bengt Zöller, Center for Primary Health Care Research, CRC, Building 28, Floor 11, Jan Waldenströms gata 35, Skåne University Hospital, Malmö S-205 02, Sweden; bengt.zoller{at}med.lu.se

Abstract

Objectives IBS aggregates in families, but the familial risk of IBS has only been determined in first-degree relatives and spouses. This nationwide study aimed to determine the familial risk of IBS in first-degree, second-degree, and third-degree relatives and spouses of affected individuals in order to estimate the relative influences of genes and shared family environment.

Methods We performed a case-cohort study. The Swedish Multigeneration Register was linked to the Hospital Discharge Register for the period 1987–2010 and the Swedish Outpatient Care Register for 2001–2010. ORs for IBS were calculated for relatives of individuals who had been diagnosed with IBS compared with relatives of individuals unaffected by IBS as the reference group. ORs were also determined for IBS cases diagnosed in primary healthcare in four Swedish counties (2001–2007).

Results The ORs for IBS were 1.75 in siblings (95% CI 1.63 to 1.89), 1.82 in offspring (1.67 to 1.97), 1.90 in parents (1.76 to 2.05), 1.10 in maternal half-siblings (0.88 to 1.39), 1.78 in paternal half-siblings (1.48 to 2.15), 1.27 in nieces/nephews (1.18 to 1.38), 1.11 in cousins (1.04 to 1.18), and 1.51 in spouses (1.24 to 1.84) of probands diagnosed with IBS. The OR for probands diagnosed in primary healthcare was 1.82 in siblings (1.52 to 2.18), and 1.82 in offspring (1.49 to 2.21).

Conclusions The increased IBS risk among first-degree relatives and also second-degree and third-degree relatives indicates a genetic component of the familial clustering of IBS. However, a non-genetic contribution is also suggested by the increased risk among spouses.

https://doi.org/10.1136/gutjnl-2013-305705

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Significance of this study

    What is already known on this subject?

    • IBS is known to aggregate in families.

    • Familial aggregation may be due to shared genetic or environmental factors.

    What are the new findings?

    • IBS aggregates in Swedish families, and a non-genetic familial contribution is suggested by the increased risk among spouses.

    • A genetic contribution to the familial aggregation of IBS in Sweden is suggested by the increased familial risks among first-degree relatives and also second-degree and third-degree relatives.

    • This the largest register-based family study of IBS, and the first nationwide one.

    How might it impact on clinical practice in the foreseeable future?

    • Family history of IBS is a potential useful predictor for IBS.

    • Genetic studies in order to identify IBS-associated genetic variants might be worthwhile.

    Introduction

    IBS is a chronic functional bowel disorder characterised by abdominal pain or discomfort. It is relieved by defecation, and its onset coincides with a change in defecation frequency or stool consistency.1 ,2 It is one of the most common gastrointestinal conditions.3 Although a number of disease mechanisms have been suggested, the pathophysiology of IBS is still poorly understood.1 ,2 IBS has been shown to aggregate in families.4–7 This may be due to shared genes or shared family environmental exposures.8 ,9 Twin and adoptee studies can help to disentangle genetic and environmental influences.9 Twin studies support the concept that IBS has genetic and environmental contributions.10–14 Conflicting data from twin studies exist. For instance, a study by Mohammed at al reported that there was no significant difference in casewise concordance rates between the MZ (monozygotic) and DZ (dizygotic) twins (28% vs 27%), suggesting that genetic factors are of little or no influence on IBS, and that the predominant influences are environmental.12 By contrast, Levy et al found that the concordance for IBS was significantly greater in MZ (17.2%) than in DZ (8.4%) twins (p=0.0030), supporting a genetic contribution to IBS.11 However, Levy et al also found that the proportion of dizygotic twins with IBS who have mothers with IBS (15.2%) was greater than the proportion of dizygotic twins with IBS who have co-twins with IBS (6.7%, p<0.001).11 Levy et al concluded that although heredity contributes to the development of IBS, social learning has an equal or greater influence.11 By contrast, the twin studies by Morris–Yates et al10 and Lembo et al14 suggested that a substantial proportion of the liability for functional bowel disorders and IBS may be under genetic control. Bengtson et al13 found support for a genetic contribution to IBS but also for a significant influence of restricted fetal growth on the development of IBS later in life. Another approach to estimate the familial non-genetic contribution is to study spouses, who share adult environments but not genetic factors.9 However, Saito et al found no increase in IBS risk among spouses.7 While family studies suggest a genetic contribution, definitive disease-causing genes remain to be determined for IBS; in spite of that many case-control studies have been carried out.15

    A further possibility to study the influence of genetic and non-genetic familial factors is to study first-degree, second-degree and third-degree relatives.16 First-degree relatives share 50% of their genes, in addition to environmental exposures common to their family. Second-degree relatives (half-siblings and uncles/aunts/nieces/nephews) share 25% of their genes, and third-degree relatives (eg, first cousins) share 12.5% of their genes. An increased disease risk in second-degree and third-degree relatives of affected individuals supports the interpretation that genetic factors influence familial aggregation, since individuals outside the nuclear family are less likely to share the same environmental exposure.16 ,17 Half-siblings are special in terms of family environmental exposures. According to national census data, 83% of maternal half-siblings in Sweden were registered as living in the same household, compared with only 3% of paternal half-siblings.18 To our knowledge, no nationwide studies have examined the familial aggregation in first-degree, second-degree and third-degree relatives of patients with IBS.

    Our aim was to estimate the familial risk of IBS in first-degree, second-degree and third-degree relatives in a nationwide family study in order to estimate the relative influences of genetics and shared environment on risk of IBS. The study was based on the Swedish Hospital Discharge Register and the Swedish Outpatient Care Register. We also determined familial risks when cases (or controls) were diagnosed in a primary healthcare database from four Swedish counties,19 while controls (or cases) were diagnosed in the Hospital Discharge Register and Outpatient Care Register. The findings of the present study could be potentially useful for IBS prediction in families, or as a basis for future molecular biological studies.

    Methods

    To assess IBS among individuals in Sweden, comprehensive registers and nationwide healthcare data from five sources were linked.17 ,20–24 Linkage was also made to a primary healthcare database from four Swedish counties.19 This linkage was based on the unique individual Swedish 10-digit personal ID numbers assigned at birth or immigration to all Swedish residents for life, information on which is nearly 100% complete. These numbers were replaced with serial numbers to preserve anonymity. We used data from six sources:

    1. The Swedish Multigeneration Register, which contains information on family relationships. The register contains information on index persons registered in Sweden between 1 January 1961 and 31 December 2008 and born between 1 January 1932 and 31 Decemcber 2008.

    2. The Total Population Register, which contains annual data on education and marital status from 1990 to 2010. The total population registry holds data on sociodemographic factors including data on education and marital status.

    3. The Swedish Hospital Discharge Register, which contains all hospital diagnoses for all people in Sweden from 1987 to 2010. Every record has the main discharge diagnosis.

    4. The Outpatient Care Register, which contains information on diagnoses from all specialist outpatient clinics in Sweden from 2001 to 2010.

    5. The Swedish Cause of Death Register, which contains data on date and cause of death from 1987 to 2010.

    6. The primary healthcare database covering 71 primary healthcare centres in the counties of Stockholm (n=687 310), Värmland (n=145 943), Gotland (n=84 898), and Uppsala (n=12 790).19 The primary healthcare database contains individual-level data from totally 919 954 individuals, who visited their general practitioner in the period 2001–2007.

    Variable definition

    Cases of IBS in the Swedish Hospital Discharge Register and Outpatient Care Register were identified by the following ICD (International Classification of Diseases) codes: ICD-9 564B (IBS); and ICD-10 K58 (IBS). The validity in the Hospital Discharge Register is generally 85–95%.21 ,24 The present study is not representative of all Swedish IBS patients, and may introduce a selection bias as the diagnosis of IBS is based on healthcare seeking. Therefore, we linked the nationwide Registers to a primary healthcare database in order to determine whether the calculated familial ORs were different or not when different data sources were used.19 We also determined comorbidities known to be associated with IBS in order to evaluate whether patients with IBS diagnosis in the present study had the same comorbidity patterns as previously described.1 ,2 ,3 ,4 ,25 We also exclude IBS patients with gastrointestinal differential diagnosis, that is, coeliac disease, inflammatory bowel disease and colorectal cancer. ICD codes are presented in online supplementary table S1.

    Sample

    The analyses were based on a database (see above) containing information on the entire Swedish population, including relationships.17 In the database we double-entered all sibling pairs, all maternal half-sibling pairs, all paternal half-sibling pairs, all cousin pairs, and all spouse pairs. We also single-entered all parent–offspring pairs, all offspring–parent pairs, and all aunt/uncle–niece/nephew pairs. We also required that the proband and the relative were alive after 1986. We selected pairs where at least one member of the pair (which we defined as the proband) was diagnosed with IBS. In total, 56 813unique individuals were diagnosed with IBS during the period 1987–2010. Totally, 11% of the cases were found in the Hospital Discharge Register (1987–2010) and 89% in the Outpatient Care Register (2001–2010). Potential differential diagnoses (ie, coeliac disease, IBD, and colorectal cancer) to IBS were checked for. Among IBS probands, 2.2% suffered from coeliac disease. The corresponding figures were 6.4% for IBD and 0.6% for colorectal cancer. Thus, 9% of IBS probands diagnosed in the Hospital discharge register and outpatient care register may, instead of IBS, have their symptoms explained by coeliac disease, IBD or colorectal cancer. Probands and relatives with IBS diagnosis as well as a concomitant diagnosis of coeliac disease, IBD, or colorectal cancer were therefore excluded. A total of 51 952 IBS individuals remained after exclusion of those individuals with coeliac disease, IBD, or colorectal cancer.

    Statistics

    The statistical methods used have previously been described.17 We used a case-cohort approach in order to investigate our research question. We conducted eight main analyses; proband–sibling, proband–offspring, proband–parent, proband–maternal half-sibling, proband–paternal half-sibling, proband–niece/nephew, proband–cousin, and proband–spouse. In all analyses, we studied all IBS proband–relative pairs (one affected proband and one proband relative) that could be matched to five control pairs (one control without IBS and one control relative) from the Swedish population. The proband–relative in the proband–relative pair and the control–relative in the control–relative pair may or may not be affected by IBS. For example, in the proband–sibling analysis, we selected all sibling pairs where at least one sibling was diagnosed with IBS, and matched each such pair to five control pairs. The control pairs were chosen randomly from individuals who lived in Sweden at the time of the proband's diagnosis of IBS. Furthermore, both individuals in the control pair also had to have lived in Sweden sometime during the period 1987–2010. Control pairs were matched based on birth year, sex, country of birth, and level of education (the year before the date of diagnosis). The matching was done on the proband and on the entire pair. The matching was conservative as the control in the control pair was allowed to develop IBS during the follow-up time (but not before or at the time for inclusion). For the proband–spouse analysis, we only had information on marital status from 1990 onwards. This limited our study period to 1991–2010 as we defined marital status the year before IBS diagnosis. In this analysis the control individual also had to be registered as being married the same year as the case proband. For descriptive statistics, see table 1.

    View this table:
    Table 1

    Descriptive statics for the probands, that is, age, sex, and education

    In order to test the trend that there was a higher risk of IBS in relatives who were more closely related, we included all types of proband–relative pairs in one dataset. Each pair was assigned their genetic resemblance (ie, 0.5 for sibling-pairs, parent–offspring pairs, child–parent pairs, and 0.25 for half-sibling pairs, niece/nephew pairs, 0.125 for cousin pairs and 0 for spouse pairs). We conducted the same analysis as described previously, but we also included an interaction term between the genetic resemblance and IBS in relatives. The hypothesis was that if this interaction term was not significant, there existed no trend of higher risk in IBS in relatives who were more closely related.

    In order to investigate possible sex differences in the familial clustering of IBS, we stratified the data by sex and included an interaction term between sex of proband–relative and IBS (1 for males and 0 for females).

    In an additional model, in order to investigate the potential effect of possible age in the familial clustering of IBS, we included an interaction between the defined predictor variable (eg, IBS in sibling) and age at diagnosis of IBS in the proband (centred at the mean value). The matching does not interfere with the interaction analyses. We also tested for interaction between IBS in relative and age difference between the relatives in a pair. Interaction was also tested between IBS in relative and the difference (in years) in diagnosis of IBS between proband and proband relative. We examined linear effects for these differences.26

    We also performed a sensitivity analysis where we only investigated probands that were registered twice for IBS in the registers. This analysis was performed in the same way as the analyses explained above.

    Analyses were conducted using conditional logistic regression. As an example, in the proband–sibling analysis, IBS in sibling (yes/no) was used as the independent variable. As a particular proband could be included several times, we adjusted for non-independence by using a robust sandwich estimator. In all analyses, less than 2% of the proband pairs could not be matched to five controls and were excluded from the analysis. Approximately 70% of these excluded pairs were born outside Sweden. We present ORs and corresponding 95% CIs. All calculations were performed using SAS V.9.3.

    Results

    Descriptive statistics and comorbidities

    A total of 51 952 individuals were diagnosed with IBS (excluding individuals with a concomitant coeliac disease, IBD, and colorectal cancer) during the study period (1987–2010). Table 1 presents the descriptive statistics for proband siblings, that is, age, sex and educational attainments. Table 2 presents other functional gastrointestinal disorders (for a detailed description, see below) and comorbidities of all 60 489 sibling pairs included in the study. Other functional gastrointestinal disorders and comorbidities were more common in probands than controls, and in probands than proband–relatives. Other functional gastrointestinal disorders and comorbidities were also more common in proband–relatives than control–relatives.

    View this table:
    Table 2

    Descriptive statistics of other functional gastrointestinal disorders and comorbidities for the 60 489 sibling pairs included in the analysis

    Of the 60 489 siblings, 6806 (11%) had an ICD diagnosis of functional constipation (ICD-9 564A or ICD-10 K590) and/or functional diarrhoea (ICD-9 564F or ICD-10 K591). However, to fulfil the diagnostic criteria for functional constipation or functional diarrhoea, there should be insufficient criteria for IBS according to Rome II and Rome III criteria, that is, an IBS-diagnosis cannot exist together with these diagnoses. However, these patients could have a possible IBS, and if these two diagnoses were made within a short time interval, the patients are more likely to have the same diagnosis at both occasions. We checked the mean time difference between IBS and functional constipation and functional diarrhoea. The mean time difference between diagnosis of IBS and functional constipation was 42 days with a SD of 1971 days (median −57 days and IQR −715 and 699 days). The time difference in diagnosis between IBS and functional diarrhoea was 3 days with a SD of 1271 days (median −48 days and IQR −308 and 150 days). The distribution of the time difference in days between IBS and functional constipation or functional diarrhoea is shown in online supplementary figure S1 and S2.

    Familial risks for IBS

    We found familial clustering of IBS, with significant ORs for all first-degree biological relationships (table 3). The OR of IBS for full siblings (50% genetic similarity) was 1.75 (95% CI 1.63 to 1.89), which was higher than that for maternal half-siblings who share 25% genetic similarity (OR 1.10, 95% CI 0.88 to 1.39) and have similar family environmental exposures. Nieces/nephews of proband cases had an OR of 1.27 for IBS. Nieces/nephews share 25% genetic similarity with the proband but do not usually share family environmental exposures. The OR was also significantly increased for paternal half-siblings. The OR for maternal half-siblings (who have 25% genetic similarity and often share family environmental exposures) was lower than the OR for paternal half-siblings (who have 25% genetic similarity, but rarely share family environmental exposures). The OR for cousins, who have 12.5% genetic similarity but who usually do not share family environmental exposures to any major degree, was also increased. Spouses share adult family environmental exposures but are genetically unrelated. The OR for IBS in the spouse analysis was 1.51 (95% CI 1.24 to 1.84) (table 3).

    View this table:
    Table 3

    ORs of IBS in relatives of probands diagnosed with IBS in Sweden between 1987 and 2010 compared to relatives of matched controls

    In order to test the trend that there was a higher risk of IBS in relatives who were more closely related, we included all types of proband-relative pairs in one dataset. We conducted the same analysis as described previously but we also included an interaction term between the genetic resemblance and IBS in the relative. The interaction term was strong and highly significant (OR 3.36 95% CI 2.78 to 4.05; p<0.0001), that is, there was a strong association between genetic resemblance and familial ORs of IBS.

    As the diagnosis of IBS may or may not be correct in the 6806 individuals with a diagnosis of functional constipation and/or functional diarrhoea we calculated familial risks for the remaining 53 683 siblings (89%) without a diagnosis of functional constipation and/or functional diarrhoea.The familial OR was 1.70 (95% CI 1.55 to 1.85), which is similar to the OR 1.75 (95% CI 1.63 to 1.89) in table 3, which was calculated without exclusion of cases with a possible IBS diagnosis (11%).

    Interaction between IBS, sex and age

    In table 4, the sex specific familial risks odds for IBS are presented. There were no major differences between sexes. The importance of age and sex was investigated with the insertion of interactions terms in the models. There were no significant interactions between sex and IBS (data not shown in table). Regarding age, the familial clustering was significantly stronger at younger ages but only among cousins (see online supplementary table S2), that is, the OR for IBS was higher for young individuals than for older individuals among cousins. Otherwise there were no significant interactions between age and IBS.

    View this table:
    Table 4

    Odds ratios of IBS of all probands who were diagnosed with IBS during 1987–2010 in Sweden compared to relatives to matched controls stratified on gender of proband.

    Interaction between age difference or time of diagnosis and IBS

    To test further for shared environmental factors, the interaction between age difference between relatives in a pair and IBS in relative was determined. There was no interaction between age difference between probands and proband–relatives and IBS except among paternal half-siblings (left part of online supplementary table S3). Moreover, there was no interaction between year difference in time of diagnosis between proband and proband–relative (right part of online supplementary table S3) and IBS in relative.

    Sensitivity analysis

    In a sensitivity analysis, we only included individuals registered with IBS in the registers at least twice. The ORs for cases with at least two IBS diagnoses were 2.57 (95% CI 2.06 to 3.19) for siblings, 2.66 (2.06 to 3.43) for offspring, 3.63 (2.84 to 4.64) for parents, 1.24 (1.02 to 1.52) for cousins, and 1.17 (0.56 to 2.45) for spouses. Thus, the ORs were even higher for biological relatives, but lower for spouses compared to those presented in table 3. Since only 13 136 individuals had the same diagnosis at least twice there were not enough cases to calculate risks for half-siblings.

    Additional analysis in the Outpatient Care Register

    We wanted to test whether the familial odds were affected by the use of different data sources, and used the Outpatient Care Register for this purpose. The probands diagnosis of IBS was determined in the hospital discharge register and the Outpatient Care Register, but the relative diagnosis of IBS was determined in the Outpatient Care Register. The OR for siblings was 1.72 (95% CI 1.59 to 1.87) and 1.79 (95% CI 1.64 to 1.96) for offspring, which is similar to the results in table 3.

    Additional analysis in the primary healthcare database

    When the probands were diagnosed in the Outpatient Care Register and the hospital discharge register, and the relatives were diagnosed in the primary healthcare database, the ORs for siblings were 1.90 (95% CI 1.58 to 2.28) and 2.09 (95% CI 1.75 to 2.50) for offspring. When the probands were diagnosed in the primary healthcare database and the relatives in the Outpatient Care Register and the hospital discharge register, the ORs for siblings were 1.82 (95% CI 1.52 to 2.18), and for offspring 1.82 (95% CI 1.49 to 2.21).

    Discussion

    This is the first total population study to provide robust estimates of familial IBS odds in relatives at varying genetic and environmental distances from each other (full siblings, offspring, parents, maternal and paternal half-siblings, niece/nephews, cousins and spouses). The odds of IBS was significantly increased in the first-degree, second-degree and third-degree relatives of individuals diagnosed with IBS, with the odds of IBS tending to be higher in more closely related relatives. The present large nationwide register-based follow-up study confirms previous studies of familial aggregation of IBS among first-degree relatives.4–7 ,10–14 However, previous family studies have been twin-studies, case-control studies of first-degree relatives, and based on contact to probands and relatives.4–7 ,10–14 The present study has also a much larger sample size than previous studies. The results of the present study suggest a genetic, but also a non-genetic, contribution to the familial aggregation of IBS. This is in line with previous twin studies that also suggest genetic and environmental contribution to IBS.10–14 The design of the present nationwide study is different and underlines a potential biological and genetic contribution to familial aggregation of IBS. For instance, second-degree and third-degree relatives usually do not share household (except for maternal siblings) and the increased familial ORs among these relatives suggest a genetic contribution. A genetic influence was further suggested by the finding that paternal half-siblings (3% of whom share households) had a significantly increased OR (1.78 95% CI 1.48 to 2.15). Moreover, a genetic influence on the odds of IBS was also suggested by the higher OR for full siblings (OR 1.75, 95% CI 1.63 to 1.89) than for maternal half-siblings (OR 1.10, 95% CI 0.88 to 1.39), because they share similar environmental exposures.

    Spouses are genetically unrelated, but share adult environments and demographic characteristics except for sex.27 ,28 Cardiovascular health-related behaviours, such as smoking, exercise and alcohol consumption correlate much more strongly among spouses than among siblings or between parents and offspring.27 ,28 The increased odds of IBS among spouses in our study also suggests that familial adult non-genetic factors may contribute to the increased familial risk in the Swedish population. However, assortative mating could also contribute to the association among spouses. Thus, individuals with IBS may be more likely to marry individuals with similar health problems. Moreover, spouses of individuals with IBS may also be more likely to seek medical attention. A previous study of spouses found no increased risk of IBS.7 The cause of this discrepancy between that study and the present study is unclear, but may be related to factors such as different study populations, diagnostic criteria and study sizes. Spouses share the same family environmental exposures, including diets, daily activities such as physical activity, and positive and negative life events. Spousal concordance has been shown in studies on food choice.29 Shared environmental factors like diet could, therefore, contribute to the observed spousal odds of IBS. Several diseases have previously been reported to be shared among spouses, such as asthma, depression, hypertension, hyperlipidaemia and peptic ulcer.30 Infection is another example of possible shared environmental familial factors. An increased risk of developing IBS after acute gastroenteritis has been verified in a meta-analysis.31 Bacterial gastroenteritis and helminth and protozoan infections have been associated with IBS.32 Moreover, a recent study by Zanini et al showed that even viral gastroenteritis increased the risk of IBS.33 Other examples of shared familial adult environmental risk factors are adult life trauma, low socioeconomic status, chemical exposures and allergenic antigens.28

    Strengths and weaknesses in relation to other studies

    The present study has several limitations. It included specialist-treated inpatients (11%) and outpatients (89%) in Sweden. Many IBS patients are only treated by family physicians or may not even be diagnosed. Thus, the present study is not representative of all Swedish IBS patients, and may introduce a selection bias as the diagnosis of IBS is based on healthcare seeking. However, we linked the nationwide hospital discharge register and the Outpatient Care Register to a primary healthcare database from four Swedish counties, and found similar familial OR for probands diagnosed in the primary healthcare database. However, the majority of IBS patients in the population may never be seen in primary care, and the results of this study may, therefore, also be a reflection of healthcare-seeking behaviour in families. This could potentially underestimate the familial clustering of IBS and, ideally, all relatives should be interviewed. Additionally, many controls are not without IBS, they are only without IBS diagnosis. This could result in an overestimation of familial odds. However, the estimated familial odds in the present study are somewhat lower than previously published.5–7 A further weakness is the lack of diagnostic information. We do not know which diagnostic criteria were used. The validity of an IBS diagnosis in registers relative to the existing diagnostic criteria (Rome I, II or III) has not been evaluated, so we do not know the proportion of subjects who have received an IBS diagnosis from their physician, and who actually fulfils the diagnostic criteria for IBS. Additionally, 6086 (11%) of IBS patients also had a diagnosis of functional constipation or functional diarrhoea. An IBS-diagnosis should not exist together with these diagnoses according to Rome II and Rome III criteria. However, exclusion of these possible IBS cases did not change the familial ORs for siblings (1.70 vs 1.75). Moreover, the diagnostic criteria have changed over time. Nonetheless, the fact that the patients were identified in the Hospital Discharge Register, which has 85–95% validity, and the Outpatient Care Register (hospital-based specialised outpatient care) makes it more likely that the diagnoses are correct.21 ,24 For instance, diagnoses in the Hospital Discharge Register were shown to be correct in 86% of patients with coeliac disease and 74% of patients with Crohn's disease and ulcerative colitis.24 It is possible that the diagnostic accuracy is lower for functional diagnoses. Moreover, only 9% of IBS probands in the present study developed coeliac disease, inflammatory bowel disease, or colorectal cancer during follow-up and these IBS patients were all excluded from the study. A general practice-based database in the UK has been extensively validated. The positive predictive value for an IBS diagnosis (ICD 8th revision code 5641) in the UK database was 77%.34 ,35 The validity of the IBS diagnosis made by hospital specialists are not expected to be lower. Moreover, the sex and age distribution and associated comorbidities are similar to those in other studies of IBS.1–4 ,25 This indirectly suggests that the used ICD codes for IBS mostly identifies IBS patients. Moreover, the familial risks for IBS was similar among primary healthcare patients and patients from the Hospital Discharge Register for the period and Swedish Outpatient Care Register, which also is reassuring. Additionally, the large number of comparisons is a point worthy consideration. While some advocate correcting for multiple comparisons, others suggest that in observational studies this should not be recommended.36

    The present study also has many strengths. Its design eliminates the risk of recall bias, which is an important problem in case-control studies. The study is also the largest family study of IBS and the only nationwide study of IBS. The nationwide design and the use of several well studied and high-quality national registers is also a very important strength.17 ,20–24 The Swedish personal ID numbers (replaced by serial numbers) are a valuable tool for linking medical registers, and allow for almost 100% coverage of the Swedish healthcare system.23 For instance, in 2001, serial numbers and main diagnoses were missing for only 0.4 and 0.9% of hospitalisations, respectively.22

    In conclusion, the present nationwide study confirms that IBS aggregates in families, and suggests that genetic factors contribute to IBS risk among Swedish families, although disease-causing variants remain to be found in the Swedish population. However, a non-genetic contribution to the observed familial aggregation is also suggested by the clustering among spouses. Future studies could examine which shared familial factors are important for the development of IBS.

    Acknowledgments

    The authors wish to thank the CPF's Science Editor Stephen Gilliver for his useful comments on the text. The nationwide registers used in the present study are maintained by Statistics Sweden and the National Board of Health and Welfare.

    References

    1. ↵
      1. Khan S,
      2. Chang L
      . Diagnosis and management of IBS. Nat Rev Gastroenterol Hepatol 2010;7:565–81.
      OpenUrlCrossRefPubMed
    2. ↵
      1. Longstreth GF,
      2. Thompson WG,
      3. Chey WD,
      4. et al
      . Functional bowel disorders. Gastroenterology 2006;130:1480–91.
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      1. Saito YA,
      2. Schoenfeld P,
      3. Locke GR III.
      . The epidemiology of irritable bowel syndrome in North America: a systematic review. Am J Gastroenterol 2002;97:1910–15.
      OpenUrlPubMedWeb of Science
    4. ↵
      1. Whorwell PJ,
      2. McCallum M,
      3. Creed FH,
      4. et al
      . Non-colonic features of irritable bowel syndrome. Gut 1986;27:37–40.
      OpenUrlAbstract/FREE Full Text
    5. ↵
      1. Kalantar JS,
      2. Locke GR III.,
      3. Zinsmeister AR,
      4. et al
      . Familial aggregation of irritable bowel syndrome: a prospective study. Gut 2003;52:1703–7.
      OpenUrlAbstract/FREE Full Text
    6. ↵
      1. Saito YA,
      2. Zimmerman JM,
      3. Harmsen WS,
      4. et al
      . Irritable bowel syndrome aggregates strongly in families: a family-based case-control study. Neurogastroenterol Motil 2008;20:790–7.
      OpenUrlCrossRefPubMedWeb of Science
    7. ↵
      1. Saito YA,
      2. Petersen GM,
      3. Larson JJ,
      4. et al
      . Familial aggregation of irritable bowel syndrome: a family case-control study. Am J Gastroenterol 2010;105:833–41.
      OpenUrlCrossRefPubMed
    8. ↵
      1. Burton PR,
      2. Tobin MD,
      3. Hopper JL
      . Key concepts in genetic epidemiology. Lancet 2005;366:941–51.
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      1. Risch N
      . The genetic epidemiology of cancer: interpreting family and twin studies and their implications for molecular genetic approaches. Cancer Epidemiol Biomarkers Prev 2001;10:733–41.
      OpenUrlAbstract/FREE Full Text
    10. ↵
      1. Morris-Yates A,
      2. Talley NJ,
      3. Boyce PM,
      4. et al
      . Evidence of a genetic contribution to functional bowel disorder. Am J Gastroenterol 1998;93:1311–17.
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
      1. Levy RL,
      2. Jones KR,
      3. Whitehead WE,
      4. et al
      . Irritable bowel syndrome in twins: heredity and social learning both contribute to etiology. Gastroenterology 2001;121:799–804.
      OpenUrlCrossRefPubMedWeb of Science
    12. ↵
      1. Mohammed I,
      2. Cherkas LF,
      3. Riley SA,
      4. et al
      . Genetic influences in irritable bowel syndrome: a twin study. Am J Gastroenterol 2005;100:1340–4.
      OpenUrlCrossRefPubMedWeb of Science
    13. ↵
      1. Bengtson MB,
      2. Ronning T,
      3. Vatn MH,
      4. et al
      . Irritable bowel syndrome in twins: genes and environment. Gut 2006;55:1754–9.
      OpenUrlAbstract/FREE Full Text
    14. ↵
      1. Lembo A,
      2. Zaman M,
      3. Jones M,
      4. et al
      . Influence of genetics on irritable bowel syndrome, gastro-oesophageal reflux and dyspepsia: a twin study. Aliment Pharmacol Ther 2007;25:1343–50.
      OpenUrlCrossRefPubMedWeb of Science
    15. ↵
      1. Camilleri M
      . Genetics and irritable bowel syndrome: from genomics to intermediate phenotype and pharmacogenetics. Dig Dis Sci 2009;54:2318–24.
      OpenUrlCrossRefPubMed
    16. ↵
      1. Robertson NP,
      2. Fraser M,
      3. Deans J,
      4. et al
      . Age-adjusted recurrence risks for relatives of patients with multiple sclerosis. Brain 1996;119:449–55.
      OpenUrlAbstract/FREE Full Text
    17. ↵
      1. Zöller B,
      2. Ohlsson H,
      3. Sundquist J,
      4. et al
      . Familial risk of venous thromboembolism in first-, second- and third-degree relatives: a nationwide family study in Sweden. Thromb Haemost 2013;109:458–63.
      OpenUrlCrossRefPubMed
    18. ↵
      1. Frisell T,
      2. Pawitan Y,
      3. Långström N,
      4. et al
      . Heritability, assortative mating and gender differences in violent crime: results from a total population sample using twin, adoption, and sibling models. Behav Genet 2012;42:3–18.
      OpenUrlCrossRefPubMed
    19. ↵
      1. Waehrens R,
      2. Ohlsson H,
      3. Sundquist J,
      4. et al
      . Low prevalence of irritable bowel syndrome in general practice in four Swedish counties. Scand J Prim Health Care 2013;31:132–7.
      OpenUrlCrossRefPubMed
    20. ↵
      Statistics Sweden. [The Swedish Multigeneration Register (1960–1990)] Registret över totalbefolkningen/RTB. In Swedish. Stockholm: Statistics Sweden, 2005.
    21. ↵
      The National Board of Health and Welfare. [Validity of the diagnoses from the Swedish In-Care Register 1987 and 1995]. In Swedish. Stockholm: Epidemiologiskt Centrum, the National Board of Health and Welfare, 2000.
    22. ↵
      1. Rosen M,
      2. Hakulinen T
      . Use of disease registers. In: Ahrens W, Pigeot I, eds. Handbook of epidemiology. Berlin: Springer-Verlag, 2005;231–51.
    23. ↵
      1. Ludvigsson JF,
      2. Otterblad-Olausson P,
      3. Pettersson BU,
      4. et al
      . The Swedish personal identity number: possibilities and pitfalls in healthcare and medical research. Eur J Epidemiol 2009;24:659–67.
      OpenUrlCrossRefPubMedWeb of Science
    24. ↵
      1. Ludvigsson JF,
      2. Andersson E,
      3. Ekbom A,
      4. et al
      . External review and validation of the Swedish national inpatient register. BMC Public Health 2011;11:450.
      OpenUrlCrossRefPubMed
    25. ↵
      1. Whitehead WE,
      2. Palsson OS,
      3. Levy RR,
      4. et al
      . Comorbidity in irritable bowel syndrome. Am J Gastroenterol 2007;102:2767–76.
      OpenUrlCrossRefPubMed
    26. ↵
      1. Breslow NE,
      2. Day NE
      . Statistical methods in cancer research. Volume I—The analysis of case-control studies. IARC Sci Publ 1980;32:5–338.
      OpenUrlPubMed
    27. ↵
      1. Thomas DC
      . Statistical methods in genetic epidemiology. New York: Oxford University Press, 2004.
    28. ↵
      1. Lawlor DA,
      2. Mishra GD
      , eds. Family matters. Deigning, analyzing and understanding family based studies in life course epidemiology. New York: Oxford University Press, 2009.
    29. ↵
      1. Pachucki MA,
      2. Jacques PF,
      3. Christakis NA
      . Social network concordance in food choice among spouses, friends, and siblings. Am J Public Health 2011;101:2170–7.
      OpenUrlCrossRefPubMedWeb of Science
    30. ↵
      1. Hippisley-Cox J,
      2. Coupland C,
      3. Pringle M,
      4. et al
      . Married couples’ risk of same disease: cross sectional study. BMJ 2002;325:636.
      OpenUrlAbstract/FREE Full Text
    31. ↵
      1. Dai C,
      2. Jiang M
      . The incidence and risk factors of post-infectious irritable bowel syndrome: a meta-analysis. Hepatogastroenterology 2012;59:67–72.
      OpenUrlPubMed
    32. ↵
      1. Mearin F
      . Editorial: From the acute infection to the chronic disorder “Don't worry it's just a viral gastroenteritis”. Am J Gastroenterol 2012;107:900–1.
      OpenUrlCrossRefPubMed
    33. ↵
      1. Zanini B,
      2. Ricci C,
      3. Bandera F,
      4. et al
      . Incidence of post-infectious irritable bowel syndrome and functional intestinal disorders following a water-borne viral gastroenteritis outbreak. Am J Gastroenterol 2012;107:891–9.
      OpenUrlCrossRefPubMed
    34. ↵
      1. Khan NF,
      2. Harrison SE,
      3. Rose PW
      . Validity of diagnostic coding within the General Practice Research Database: a systematic review. Br J Gen Pract 2010;60:e128–36.
      OpenUrlAbstract/FREE Full Text
    35. ↵
      1. Ruigómez A,
      2. García Rodríguez LA,
      3. Johansson S,
      4. et al
      . Is hormone replacement therapy associated with an increased risk of irritable bowel syndrome? Maturitas 2003;44:133–40.
      OpenUrlCrossRefPubMed
    36. ↵
      1. Rothman KJ
      . No adjustments are needed for multiple comparisons. Epidemiology 1990;1:43–6.
      OpenUrlCrossRefPubMed

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

      Files in this Data Supplement:

    Footnotes

    • Contributors RW, HO, JS, KS, and BZ were involved in study design and execution, and finalising the paper. RW drafted the manuscript. All authors critically revised the paper and all authors approved the final draft submitted. All authors had full access to all the data (including statistical reports and tables) and take responsibility for the integrity of the data and the accuracy of their analysis.

    • Funding This work was supported by grants awarded to Dr Bengt Zöller from the Swedish Heart-Lung Foundation and Region Skåne (REGSKANE-124611), to KS from the Swedish Research Council, and to JS from the Swedish Council for Working Life and Social Research (2007–1754) and King Gustaf V and Queen Victoria's Foundation of Freemasons, as well as by ALF funding from Region Skåne awarded to Bengt Zöller, Jan Sundquist, and Kristina Sundquist.

    • Competing interests None.

    • Ethics approval The study was approved by the Ethics Committee of Lund University, Sweden (approval number 409/2008 Lund with complementary approvals dated September 1, 2009, and January 22, 2010) and by the Ethics Committee of the Karolinska Institute, Huddinge, Sweden (reference number 12/2000, 2000-03-06 and 2002-11-18), and was performed in compliance with the Helsinki Declaration.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Data sharing statement The nationwide registers used in the present study are maintained by Statistics Sweden and the National Board of Health and Welfare.