Background Giardia lamblia is a common cause of gastroenteritis worldwide, but there is limited knowledge about the long-term complications.
Objective To estimate the relative risk of irritable bowel syndrome (IBS) and chronic fatigue 3 years after acute giardiasis.
Design Controlled historic cohort study with 3 years' follow-up. Data collected by mailed questionnaire.
Setting Waterborne outbreak of giardiasis in the city of Bergen, Norway.
Participants 817 patients exposed to Giardia lamblia infection verified by detection of cysts in stool samples and 1128 matched controls.
Main outcome measures IBS and chronic fatigue.
Results The prevalence of IBS in the exposed group was 46.1%, compared with 14.0% in the control group, and the adjusted RR=3.4 (95% CI 2.9 to 3.8). Chronic fatigue was reported by 46.1% of the exposed group and 12.0% of the controls, the adjusted RR was 4.0 (95% CI 3.5 to 4.5). IBS and chronic fatigue were associated and the RR for the exposed group of having a combination of the two outcomes was 6.8 (95% CI 5.3 to 8.5). The RR was also increased for having just one of the two syndromes, 1.8 for IBS (95% CI 1.4 to 2.3) and 2.2 for chronic fatigue (95% CI 1.7 to 2.8).
Conclusions Infection with Giardia lamblia in a non-endemic area was associated with a high prevalence of IBS and chronic fatigue 3 years after acute illness, and the risk was significantly higher than in the control group. This shows that the potential consequences of giardiasis are more serious than previously known. Further studies are needed, especially in areas where giardiasis is endemic.
- Giardia lamblia
- functional gastrointestinal disorders
- irritable bowel syndrome
- chronic fatigue disorder
- parasitic diseases
- infectious disease
Statistics from Altmetric.com
- Giardia lamblia
- functional gastrointestinal disorders
- irritable bowel syndrome
- chronic fatigue disorder
- parasitic diseases
- infectious disease
Significance of this study
What is already known about this subject?
Giardia lamblia is an established cause of waterborne gastroenteritis, both in developed and developing countries.
Irritable bowel syndrome is a common short- and long-term complication after bacterial gastroenteritis.
Knowledge about the clinical consequences after Giardia infection is limited.
What are the new findings?
Acute giardiasis is associated with an increased risk of irritable bowel syndrome and chronic fatigue 3 years after the infection.
There is an association between irritable bowel syndrome and chronic fatigue after giardiasis.
Gender is not a strong risk factor for irritable bowel syndrome after giardiasis.
How might it impact on clinical practice in the foreseeable future?
It is important to consider Giardia lamblia as a cause of infectious gastroenteritis.
Knowledge about long-term complications from giardiasis will aid counselling and treatment of particular patients.
Giardia lamblia (Giardia duodenalis/Giardia intestinalis) is a protozoan found on every continent and a common cause of acute and chronic gastroenteritis. The parasite is mainly spread through drinking water, but transmission from person to person is also reported. It is endemic in many tropical and subtropical areas, especially where sanitary conditions are poor, whereas in Europe and North America Giardia is a frequently identified pathogen in waterborne outbreaks of disease.1
In acute and longstanding giardiasis, malabsorption and weight loss are common features,1 and in children chronic or repeated infections may lead to stunting and impaired cognitive development.2 However, after eradication of the parasite the predominant notion has been that Giardia does not cause long-term complications, in contrast to several other infections that may trigger post-infectious disorders. Following bacterial gastroenteritis, abdominal symptoms are commonly described, most often irritable bowel syndrome (IBS),3 and other infections are associated with a risk of prolonged fatigue after the acute phase.4 IBS and chronic fatigue are most often investigated separately, but when studied together there is some association and an overlap in prevalence has then been demonstrated.5 6
The objective of this study was to investigate the prevalence of IBS and chronic fatigue in a group of patients 3 years after they had had acute giardiasis. The diagnoses were made using validated questionnaires, and prevalence was compared with that of a control group.
Bergen is a city of 250 000 inhabitants on the western coast of Norway. A reservoir providing drinking water to 48 000 registered subscribers in the central part of the city was contaminated with Giardia cysts for several weeks in the autumn of 2004. The number of infected individuals is unknown, but based on metronidazole prescription data it has been estimated that 2500 people were treated for giardiasis during the outbreak.7
In the period from October 2004 to December 2005 the parasitology laboratory at Haukeland University Hospital, the only one of its kind serving the Bergen area, detected Giardia in stool samples from 1253 people assumed to have been infected during the outbreak. The samples were collected as part of regular clinical practice and surveillance of the outbreak. An unregistered number of individuals with positive stool samples were excluded from the database before this study was designed because the information accompanying the samples made it clear that the people had recently visited or moved to Bergen from areas where giardiasis is endemic. In the follow-up period after the outbreak one patient died. The remaining 1252 individuals (61% female, mean age 34.6 years) were included in this study and classified as ‘exposed’, the exposure being acute giardiasis verified by detection of Giardia in stool samples during the outbreak. For each exposed individual Statistics Norway randomly sampled two people of the same age and gender from the entire population in Bergen, thus establishing a 2:1 matched control cohort. The matching was done this way to balance the high number of females and young people among the exposed group.
Questionnaires were sent to participants by regular mail in October 2007, and non-responders were mailed again after 1 month. In an effort to reduce possible bias caused by a low response rate among controls this group was expanded by adding two more controls for each exposed individual when none of the first two controls had responded. As a result the questionnaire was mailed to 1094 additional controls in May 2008.
Primary outcome variables were IBS and chronic fatigue. This was based on clinical observations by members of the research group and data from a previous study after the outbreak.8 Secondary outcomes were more serious IBS and fatigue, and total fatigue scores.
IBS was defined according to the Rome III criteria: recurrent abdominal pain or discomfort at least 3 days a month in the past 3 months and associated with at least two of three criteria related to defecation (onset associated with a change in frequency or form of stool, or improvement with defecation).9 Further, we designed two categories of more serious IBS. ‘Frequent IBS’ was defined as IBS with pain or discomfort more than 1 day a week. ‘Severe IBS’ was defined as IBS limiting or restricting daily activities at least ‘often’.
The questions in the Rome III Diagnostic Questionnaire10 necessary to diagnose IBS were translated from the original English version to Norwegian by the researchers. The Norwegian text was then translated back to English by a professional translator having English as her native language, and based on this translation minor adjustments were made to the Norwegian version.
A question about subjective milk intolerance was included, and participants reporting this were also asked when the symptoms started.
Fatigue was measured by the validated Fatigue Questionnaire (FQ)11 which consists of 11 questions dealing with different aspects of fatigue. The same version of the FQ has previously been used in a study on fatigue in the general population in Norway.12 To each question there are four possible answers (‘less than normal’, ‘as normal’, ‘more than normal’, ‘much more than normal’) that are scored (0, 1, 2, 3) and added to give a total fatigue score (range 0–33). Responses are also dichotomised (0 and 1 into 0, 2 and 3 into 1), and added to give a dichotomised score where four or more defines ‘chronic fatigue’, provided that fatigue has been present for at least the past 6 months. We also designed two categories of more serious fatigue. ‘Severe fatigue’ was defined as the combination of chronic fatigue and a total fatigue score of ≥23, and ‘consistent fatigue’ as chronic fatigue combined with fatigue present at least 75% of the time. The FQ was accepted if seven or more of the 11 questions were answered, and then any unanswered question was assigned the same value as the mean of all responses to that particular question.
We included the demographic variables gender, age (categorised in 20-year groups), marital status (four categories), level of education (three categories), main occupation (eight categories, reduced to four in the analyses) and whether the person had been a student during the outbreak. All these variables were considered potential confounders, and gender was also considered a possible effect modifier.
Analyses and statistical methods
Non-responders were excluded from all analyses, except non-responder analyses. People with missing data were excluded from the analysis involving that particular variable, unless data were still sufficient to unequivocally define the outcome according to set criteria.
Association in 2×k tables was tested by Pearson's χ2 test. Results are reported as relative risks (RRs) with 95% CIs. IBS and chronic fatigue were analysed separately with respect to the risk factors and possible interactions using multiple logistic regression producing adjusted OR with 95% CI,13 which were converted to RRs and corresponding CIs by the method of Zhang and Yu.14 Effect modification was tested by the Breslow–Day test for homogeneity of ORs after stratification. Confounding was evaluated by use of the Mantel–Haenszel common OR and multiple logistic regression analyses, and variables found to be confounders were then controlled for by multiple logistic regression. We calculated the attributable fraction among the exposed (AFE)—that is, the proportion of IBS and chronic fatigue in the exposed group that can be attributed to the Giardia infection, as a percentage given by the formula AFE = (1−1/RR)×100.15 The CIs for AFE was calculated from the CIs for RRs.
Means of continuous variables were compared between groups using Gosset's unpaired t test.16 Internal consistency of the FQ was tested by calculating Cronbach's α for all items.
Level of statistical significance was set at 0.05, and all tests were two-sided. All analyses were done using SPSS for Windows V.15.0.
The response rates were 65.3% (817/1252) among the exposed group and 31.5% (1133/3598) among controls (table 1). The questionnaires from five controls were rejected because they had been given the diagnosis of giardiasis by a medical doctor during the outbreak. Responders were older than non-responders (mean age 36.1 vs 32.9 years, p<0.001), and a higher proportion among them were female (65.7% vs 55.8%, p<0.001). Characteristics of responders in the exposed and control groups are shown in table 2.
Of the responders 1875 (96.4%, 770 exposed and 1105 controls) completed the IBS-part of the questionnaire. The prevalence of IBS was 46.1% (355/770) among the exposed group and 14.0% (155/1105) among controls, giving an adjusted RR of 3.4 for IBS in the group who had been sick with giardiasis 3 years earlier (table 3). The fraction of IBS attributable to giardiasis (AFE) in the exposed group was 69.7% (95% CI 65.5% to 73.0%). The RRs of the secondary outcomes ‘severe’ or ‘frequent’ IBS were 5.1 (95% CI 3.6 to 7.2) and 6.2 (95% CI 4.5 to 8.3), respectively, for the exposed group.
The question about subjective milk intolerance was answered by 1800 responders, data were missing from 93 of the exposed group and 52 of the controls. Those with previous giardiasis reported intermediate or severe milk intolerance more often than the controls (15.2% (110/724) vs 6.6% (71/1076), p<0.001), but this was not related to IBS status. Among patients with IBS, the same proportion in both groups reported milk intolerance, 24.3% (78/321) in the exposed group and 24.2% (36/149) in the control group. The majority (65%) of those with subjective milk intolerance could not remember when this problem started, and meaningful analyses on the debut of symptoms could not be performed.
Among the responders 1912 (98.3%, 794 exposed and 1118 controls), answered the FQ. The prevalence of chronic fatigue was 46.1% (366/794) in the exposed group and 12.0% (134/1118) in the control group, giving an adjusted RR of 4.0 for chronic fatigue 3 years after acute giardiasis (table 3). The fraction of chronic fatigue attributable to giardiasis (AFE) in the exposed group was 75.0% (95% CI 71.4% to 77.8%). For the secondary outcomes, ‘severe’ and ‘consistent’ chronic fatigue, the RRs were 7.4 (95% CI 4.9 to 10.9) and 3.6 (95% CI 2.6 to 4.7), respectively.
The mean total fatigue score among exposed people was 16.1 and among controls 12.2 (p<0.001). When comparing only those with chronic fatigue, the exposed still had a slightly higher mean total fatigue score than controls (20.8 vs 19.0, p<0.001). Cronbach's α for the FQ was 0.93 in this study.
IBS and chronic fatigue were associated in both groups. In the exposed group 62.6% of those with IBS had chronic fatigue, compared with 30.6% of those without IBS (p<0.001). In the control group the corresponding figures were 32.0% and 9.0% (p<0.001). There was a high prevalence (28.6%) of combined IBS and chronic fatigue in the exposed group, but also of only one of the two conditions (17.1% IBS and 16.6% chronic fatigue). Compared with the controls the exposed had a higher RR of the combination IBS/chronic fatigue (RR=6.8; 95% CI 5.3 to 8.5), IBS without chronic fatigue (RR=1.8; 95% CI 1.4 to 2.3), and also of chronic fatigue without IBS (RR=2.2; 95% CI 1.7 to 2.8) (table 3).
Gender was not an effect modifier for IBS or for chronic fatigue or any of the secondary outcomes, when comparing the exposed and control groups (data not shown). Prevalence of IBS was higher among female than male subjects in both the exposed group (48.9% vs 40.8%, p=0.032) and the control group (15.9% vs 10.4%, p=0.012), but for chronic fatigue there was no significant gender difference in the prevalence within each group (in the exposed group 47.2% female subjects vs 43.9% male subjects (p=0.37) and in the control group 12.8% vs 10.4% (p=0.23)).
The key finding in this study was a high prevalence of IBS (46%) and chronic fatigue (46%) in exposed patients 3 years after acute giardiasis. Compared with the control group the RR was 3.4 for IBS and 4.0 for chronic fatigue. The two conditions were strongly associated, especially in the exposed group which had an RR of 6.8 for the combination of the conditions, compared with the control group.
Outbreaks offer good opportunities to investigate clinical aspects of infectious diseases owing to a clear definition of exposure and a high number of patients, but there are also limitations as research is challenged by lack of time to plan and conduct a study, and by increased demands on health services due to the number of affected patients. Valuable background data about the participants before the outbreak will often be scarce. In our cohort, information about previous IBS, chronic fatigue and related health problems would be of special interest, but was not available. In three other studies from this outbreak, premorbid abdominal symptoms were registered. The exposed group in this study was previously asked whether they had gastrointestinal problems before the outbreak, and the answers to that question were not associated with symptoms 2 years later.8 Another study found that in a group of patients referred to Haukeland University Hospital, 15% reported minor abdominal symptoms before the outbreak.17 However, in a smaller study, primary care visits by patients to the general practitioner for abdominal problems during the past 2 years before the outbreak were associated with persistent symptoms.18 If previous complaints made patients more inclined to seek healthcare this would bias our results in the direction of stronger association between giardiasis and late symptoms.
In this study, participants were included based on an investigation of stool samples submitted as part of regular clinical practice. The composition of the study group is therefore influenced both by healthcare seeking behaviour by individuals and by doctors' decision to ask for stool samples. Both situations might potentially introduce a bias, and partly explain the female majority (61%) among the exposed group. Another explanation for the gender difference might be that females drink more water. It was shown that a daily intake of more than five glasses of tap water was associated with a higher risk of giardiasis during the outbreak.7
The response rate of 65% in the exposed group is acceptable, and high compared with other studies based on mailed questionnaires19 or consultations,20 but could still cause the results to be biased in either direction. Some might not respond owing to ill health, while others who did not have any complaints might feel the questionnaire did not concern them. Given the high prevalence of symptoms in the exposed group, it is not likely that the true prevalence is higher than our finding.
The low response rate (31.5%) among the controls gives reason to be aware of possible selection bias in this group. However, the prevalence of 14% for IBS and 12% for chronic fatigue among controls in this study is comparable to previously reported findings in population-based studies in Norway. In a public health survey including 4622 people aged 30–75 years from another part of Norway, 388 (8.4%) had IBS according to Rome II criteria.21 The difference in prevalence between the control group in our study and this reference population correlates with findings in studies comparing prevalence based on Rome II and Rome III criteria.22–24 A study using the FQ on a sample from the general population in Norway reported a prevalence of 10.5% chronic fatigue in the age group 30–49 years.12 This suggests that a potential bias caused by the low response rate among controls has not led to an overestimated RR of IBS and chronic fatigue.
There is a possibility that those having had acute giardiasis will be more aware of symptoms and more likely to find them abnormal than the controls, and thus report more complaints. Another cause of potential bias is that the municipality claimed full responsibility for the outbreak and decided to give compensation to those affected. This might possibly have led to exaggeration of symptoms, but the benefit of doing so would be limited as compensation was restricted to any verified economical loss, and 3 years after the outbreak these matters would be settled for most of the people affected. Even though higher awareness may explain some of the prevalence in the exposed group, we do not find it plausible that this is the only reason for the very high prevalence of both IBS and chronic fatigue.
Ideally, examination of stool samples should be performed to rule out chronic giardiasis before IBS is diagnosed,25 but the design of this study did not make this possible. However, other studies have shown very low prevalence of chronic giardiasis or asymptomatic carrier status after this outbreak.26–28
Both IBS and fatigue may be triggered by infections. IBS is described after bacterial gastroenteritis (Salmonella, Shigella and Campylobacter),3 and a persistently high risk has been demonstrated up to 8 years after the initial infection.29 Fatigue as part of chronic fatigue syndrome is linked to several infections (Epstein–Barr virus, parvovirus B19, enteroviruses, Coxiella burnetii and others).30 Post-infectious complications after giardiasis are rarely described, the only exception being lactose intolerance, which has been poorly characterised.1 Lactose intolerance does not seem to play a role in the development of post-infectious IBS after bacterial gastroenteritis,31 nor does subjective milk intolerance predict lactose intolerance in patients with IBS.32 In this study the majority of patients with IBS did not report subjective milk intolerance, and there was no correlation between previous Giardia infection and subjective milk intolerance in patients with IBS. We therefore conclude that lactose intolerance is not a mechanism for post-giardiasis IBS.
Previous studies exploring the clinical course of giardiasis in different patient populations after the Bergen outbreak have shown a high prevalence of prolonged gastrointestinal symptoms8 27 and tiredness8 and have investigated the development of IBS.33 None of these studies included a control group and therefore interpretation is difficult since both IBS and fatigue are common in the general population, also among individuals where there is no clear history of a preceding infection.12 21 34 This study confirms that among those who fell sick with giardiasis in Bergen there was a high risk of acquiring post-infectious complications. When employing the stricter criteria, which defined more serious IBS and fatigue that would affect activities of daily living, the RR in the exposed group remained high. This supports the conclusion that infection with Giardia may lead to clinically significant complications.
There is evidence that different infections may cause separate post-infectious syndromes,35 but we also found a high risk of combined IBS and chronic fatigue after Giardia infection. It is not clear whether this combination is unique to Giardia, or whether it is common also after other infections. Epidemiological studies in unselected groups have shown correlation between different unexplained somatic syndromes, including IBS and chronic fatigue,36 and a case–control study found that patients with IBS and other functional gastrointestinal disorders had more severe fatigue than controls.37
Female gender is one of several patient characteristics that are associated with IBS.34 There is evidence that post-infectious IBS is a clinically distinct subgroup of IBS,38 and the role of gender in this group varies between studies. Some studies show no significant difference and others find a higher prevalence among females, with an RR>2.3 In our study gender was not an effect modifier for IBS, but there was a slightly higher prevalence of IBS among females in both groups. Several factors may have influenced these findings. More female subjects were diagnosed with giardiasis. If this to some degree was caused by differences between genders in healthcare seeking behaviour, this might mean that males presented with more severe symptoms during the acute infection, which in turn could influence the risk of post-infectious complications. There were also more females among responders, and if IBS was more common among responders than non-responders, this might bias the results towards a relatively lower prevalence in females.
Fatigue is more common in females in the general population.12 39 In this study gender was not an effect modifier for chronic fatigue, and there was no significant gender difference in the prevalence of chronic fatigue in the groups.
In this study exposure to acute infection with Giardia was the dominating factor linked to both IBS and chronic fatigue. Most of our knowledge of the mechanisms of Giardia infection is based on studies in animals and has not been confirmed in humans.40 The results of infection are hypersecretion, malabsorption and reduced disaccharidase activity causing diarrhoea and other gastrointestinal symptoms,41 and a previous study from this outbreak has demonstrated high prevalence of duodenal inflammation in patients with persistent symptoms.17 The mechanisms explaining how an acute duodenal infection may cause chronic problems, including large bowel symptoms and systemic manifestations, are largely unclear. A broad approach to the pathogenesis is needed, and focus should be on both pathogen and host factors. The immune response may be important—in particular, T lymphocytes, as their role has been linked to both acute giardiasis,40 IBS3 and chronic fatigue.42
We report a high prevalence of IBS and chronic fatigue after giardiasis in a non-endemic area, where the affected population was relatively homogeneous and can be considered to be previously unexposed to Giardia. The majority of patients were young adults, whereas in endemic areas mostly children are affected. This study shows that the long-term clinical consequences of Giardia may be more severe than previously expected, but one should be cautious about interpreting the results of this study beyond the setting of outbreaks in developed countries. Our data offer new insight into both the potential pathogenicity of Giardia, and the epidemiology of unexplained somatic syndromes like IBS and chronic fatigue. This shows the need for more research to improve our knowledge of giardiasis, both in endemic and non-endemic areas.
We thank Trygve Hausken, Per Olav Vandvik and Jon Håvard Loge for valuable input regarding the selection of study variables and design of the questionnaire.
Funding An open grant from the Municipality of Bergen covered the cost for production and mailing of the questionnaire and for the services delivered by Statistics Norway. KAW was partly funded by the Norwegian Medical Association's Funds for Research in General Practice. All researchers are independent from the sponsors. The sponsors had no role in study design, in collection, analysis and interpretation of data, or in writing or deciding to submit the manuscript.
Competing interests All authors have completed the Unified Competing Interest form at http://www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare that except for the information stated about funding no authors (1) have received support from any organisation for the submitted work, (2) have any financial relationships with any organisations that might have an interest in the submitted work in the previous 3 years, or (3) have any other relationships or activities that might appear to have influenced the submitted work.
Ethics approval This study was approved by the Regional Committee for Medical and Health Research Ethics (project 150.07), and by the Ombudsman for Privacy in Research, Norwegian Social Science Data Services (project 17014).
Provenance and peer review Not commissioned; externally peer reviewed.
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.