Infection-induced anaemia: a cross-sectional study of 14,636 German travellers aged 20–49 years
- First Online:
- Cite this article as:
- Herbinger, K., Metzner, M., Schmidt, V. et al. Infection (2013) 41: 1079. doi:10.1007/s15010-013-0528-6
- 131 Views
Anaemia is a frequently diagnosed condition which can develop as a consequence of numerous factors, including infectious diseases (IDs). Travelling, especially in sub-/tropical regions, leads to an elevated risk of contracting IDs. The aim of our study was to assess the epidemiological significance of IDs in inducing anaemia among a large cohort of returned travellers.
This was a cross-sectional study in which data on 17,009 returned travellers aged 20–49 years who consulted the travel medicine clinic of the University of Munich between 1999 and 2011 were retrieved and analysed.
Of the returned travellers, 8.3 % (6.0 % of males/10.4 % of females) were diagnosed with anaemia. The prevalence of anaemia was significantly elevated among patients of African (21.4/28.3 %) and Asian (11.6/15.7 %) origin. When the study population was restricted to the 14,636 travellers of German origin, 7.1 % of the returned travellers (4.6/9.6 %) were diagnosed with anaemia. The prevalence was significantly elevated among patients who travelled for >30 days (5.7 of males/10.6 % of females) and for male travellers visiting friends and relatives (7.7 %). However, these correlations were confounded by malaria. The prevalence of anaemia was significantly elevated only among returned travellers diagnosed with malaria (36.1 of males/26.9 % of females) and with symptomatic intestinal Entamoeba histolytica infections (30.0/33.3 %).
Following the exclusion of confounding by malaria from the statistical analysis, the prevalence of anaemia was found to be significantly elevated among patients of African and Asian origin, and among patients of German origin who had travelled for >30 days, it could be mainly attributable to chronic, long-lasting causes. Although more than 550 travel-associated IDs were assessed in our study, only symptomatic intestinal Entamoeba histolytica infections and, to an even larger extent, malaria were determined to be of epidemiological significance for inducing anaemia among travellers.
Anaemia is a global public health condition affecting approximately 25 % of the world’s population, with the proportion varying immensely among different population groups and among regions depending on local conditions [1, 2]. Those with the highest risk of developing anaemia include preschool children (global prevalence of approx. 48 %) and women (approx. 30 %), especially pregnant women (approx. 42 %) . The prevalence of anaemia is estimated to range from about 5 to 25 % among females aged 15–49 years living in developed countries and from about 20 to 75 % among those living in tropical countries . The highest proportions of >50 % are observed in African countries, whereas the overall highest absolute number of affected individuals can be found in Southeast Asia (315 million) . The World Health Organization (WHO) considers that about one-half of cases of anaemia are due to iron deficiency, while the aetiopathogenesis of the remaining half is determined by a large number of different factors. Such factors include protozoal, helminthic, bacterial and viral infectious diseases (IDs) which are either proven or suspected causes of anaemia and are typical or specific to sub-/tropics .
Protozoal IDs (such as babesiosis [3–5], malaria  and connatal toxoplasmosis ) as well as bacterial IDs (such as Oroya Fever due to Bartonella bacilliformis infection ) and connatal viral IDs (such as cytomegalia ) may induce severe haemolysis. Additionally, splenomegaly and the “pooling” of erythrocytes caused by schistosomiasis , malaria , African trypanosomiasis  visceral leishmaniasis , brucellosis , histoplasmosis  and infections with lymphotrophic viruses (particularly mononucleosis and cytomegalia) may lead to normocytic normochromic anaemia . Helminthic IDs, such as ancylostomiasis [17, 18], ascariasis [18–20], diphyllobothriasis [21, 22], fascioliasis , schistosomiasis  and trichuriasis [18, 20], may lead to bleeding events and consequently to microcytic (due to iron deficiency) or macrocytic (due to vitamin B12 deficiency) anaemia. In very rare cases, this scenario has also been described for giardiasis . Other IDs of the intestine, such as campylobacter enteritis [24, 25], shigella enteritis  or intestinal entamoeba infections [27, 28], may lead to acute anaemia due to bleeding events. Chronic IDs, such as bacterial osteomyelitis , bartonellosis [8, 30], endocarditis lenta , leprosy , syphilis [33, 34], and tuberculosis , can lead to disorders of haematopoiesis as a result of abnormal iron distribution. In addition, a variety of viral IDs, such as dengue fever [36, 37], human immunodeficiency virus (HIV) infections [38, 39], mononucleosis [40, 41], parvovirus B19 infections  and others can induce anaemia by suppression of the bone marrow [1, 43].
International travel has increased by more than 50 % from 1999 (626 million trips) to 2011 (983 million trips) . More than 50 million people travel each year from industrialized countries to sub-/tropical destinations, and this number is continuing to rise . Depending on the travel destination, this growing mobility possibly exposes the traveller to an elevated risk of acquiring IDs and of potentially importing these “exotic” diseases to their home countries, especially during travels in sub-/tropical regions . The differential diagnosis of conditions/symptoms presented by returned travellers requires the treating physician to think broadly , as IDs are becoming a more prominent global issue . An example of this is the registration of the first case of fatal dengue haemorrhagic fever infection in Munich, Germany in 2009 . This infection was imported by a returned traveller into Germany, where dengue fever is not endemic and where most clinicians have no practical experience with this ID.
In spite of the high prevalence of anaemia in sub-/tropical countries, the large number of ID’s causing anaemia and the continuously increasing number of international travellers going into sub-/tropical countries, data on the risk of acquiring infection-induced anaemia among travellers are still lacking. The objective of our cross-sectional study was to assess the epidemiological significance of infection-induced anaemia among a large cohort of returned travellers consulting the outpatient travel medicine clinic of the Department of Infectious Diseases and Tropical Medicine (DITM) of the Ludwig-Maximilians-University in Munich, Germany, by analysing data on demographics, travel and laboratory diagnosis of IDs. The results of a comparable study on infection-induced thrombocytopenia were published in 2012 .
Patients and methods
Study population and inclusion criteria
Data were collected on 48,678 individuals consulting the DITM for treatment or general medical advice for the time period of January 1999 through to December 2011. Individuals were considered eligible for entry into the study if they satisfied three inclusion criteria: (1) laboratory values of haemoglobin (Hb), haematocrit (Ht) and erythrocytes (red blood cells, RBCs) were available—this criterion was met by 32,474 individuals (66.7 %); (2) travel destination was known—met by 26,580 (54.6 %) individuals; (3) duration of travel was known—met by 26,482 (54.4 %) individuals.
All three criteria were fulfilled by 22,756 (46.7 %) individuals, and these individuals comprised the study population.
Laboratory analysis and anaemia
Blood samples were obtained from all travellers at first visit to the DITM prior to any therapeutic drug administration and collected in ethylenediaminetetraacetic acid-coated tubes. They were analysed in a fully automated haematology analyser (model KX-21 N; Sysmex Digitana AG, Horgen, Switzerland) without previous storage or freezing. We defined a case of anaemia as a returned traveller aged 20–49 years fulfilling at least one of the following diagnostic criteria on the first visit to the DITM: (1) RBC count of <4.3 × 106/mm3 for males/<4.0 × 106/mm3 for females; (2) Hb of <13.3 g/dL for males/<12.0 g/dL for females; (3) Ht of <38.8 % for males/<35.4 % for females.
Anaemia was defined as normochromic if the mean corpuscular haemoglobin (MCH) was 26.7–31.9 pg/cell, hypochromic if the MCH was <26.7 pg/cell and hyperchromic if the MCH was >31.9 pg/cell. Anaemia was defined as normocytic if the mean corpuscular volume (MCV) was 79.0–93.3 μm3, microcytic if the MCV was <79.0 μm3 and macrocytic if the MCV was >93.3 μm3 .
Variables and statistical analysis
Among the study population of 22,756 individuals, more than 550 different IDs were confirmed by laboratory findings. To assess the epidemiological significance of infection-induced anaemia among these returned travellers, we analysed for the prevalence of anaemia in terms of variables on demographics (gender, age, origin), travel (travel destination, duration of travel, type of travel) and diagnosis of IDs. In this study, origin was defined as country of birth, and travellers visiting friends and relatives (VFRs) were defined as those individuals who travelled to visit friends and relatives abroad, regardless of whether they are born outside of Germany (non-German origin) or in Germany (German origin). To avoid confounding by gender, data were presented separately in strata for male and female patients, as the proportion of patients with anaemia is known to be higher in females.
Approximative tests (χ2 tests and χ2 tests for linear trends) and t tests (parametric test) were conducted using Stata software ver. 9.0. (Stata Corp, College Station, TX) and EpiInfo ver. 3.3.2. (Centers for Disease Control and Prevention, Atlanta, GA). Significant differences were defined as P values of <0.05 or as non-overlapping of 95 % confidence intervals (95 % CI) of proportions.
Ethical clearance and informed patient consent
Ethical clearance of the study protocol was provided by the Ethical Committee of the Medical Faculty at the University of Munich, Germany. Clinical and laboratory data were used only from patients who had signed a general informed consent form to participate in any study or, in the case of minors, who could provide a general informed consent form that was signed by their legal guardians.
Data from 22,756 patients [11,321 (49.7 %) males; 11,435 (50.3 %) females] were included in the study. The age of the male patients ranged from 0.8 to 86 years and that of the female patients from 0.5 to 97 years. The median age of the male and female patients was 36.4/33.5 years, respectively, and the mean age was 38.7 (95 % CI 38.4–39.0)/36.9 (95 % CI 36.6–37.2) years, respectively. The prevalence of anaemia was significantly (P < 0.01) lower among males (7.2 %) than among females (10.3 %). The majority of patients were in the age group 20–49 years (74.1 % of males/75.4 % of females) followed by the age groups 50–97 years (21.1/19.6 %) and 0–19 years (4.8/5.0 %). The prevalence of anaemia was significantly (P < 0.01 each) highest in the age group 50–97 years (11.9 of males/11.0 % of females), followed by the age groups 20–49 years (6.0/10.4 %), and 0–19 years (4.9/6.1 %). To avoid confounding by age, the study population was restricted to patients aged 20–49 years (17,009 of the 22,756 patients, 74.7 %).
Origin (place of birth) of 17,009 returned travellers aged 20–49 years (cases: 1,404; prevalence: 8.3 %) consulting the outpatient travel medicine clinic of the Department of Infectious Diseases and Tropical Medicine of the University of Munich, Germany, for the first time between 1999 and 2011
Cases (n)a (%)b
Total (n)c (%)d
Cases (n)a (%)b
Total (n)c (%)d
The proportion of patients diagnosed with malaria was 1.4 % males/0.6 % females. Among those of African origin, these proportions (9.1/8.2 %) were significantly (P < 0.01 each) higher, while among those of Asian origin, this proportion (1.2/1.5 %) was not significantly (P = 0.73/P = 0.20) different. To avoid confounding by origin, patients were restricted to those of German origin (14,636 of 17,009; 86.0 %).
Among the 14,636 (7,148 males/7,488 females) patients aged 20–49 years of German origin, the prevalence of anaemia was significantly (P < 0.01) lower among males (4.6 %) than among females (9.6 %). The destination of most patients had been Asia (44.8 % males/43.1 % females), followed by Africa (30.1/31.6 %) and Latin America (19.3/19.5 %). The highest prevalence of anaemia among male patients was found in those with a travel destination of Africa (5.2 %); for females, this was Eastern Europe (12.4 %). However, there was no significant correlation found between any travel destination and the prevalence of anaemia. About one-third of the patients travelled for each of the following travel durations: of 1–14 days (31.1 % males/30.9 % females), 15–30 days (34.8/35.7 %) and >30 days (34.1/33.4 %). The prevalence of anaemia among travellers with a travel duration of 1–14 days was significantly (P < 0.01 each) lower (3.5/8.7 %) and among those with a travel duration of >30 days was significantly (P < 0.01/P = 0.03) higher (5.7/10.6 %). The χ2 tests for trends assessed a significant (P < 0.01 males/P = 0.01 females) linear trend between the increased duration of travel and the elevated prevalence of anaemia.
Data on travel and diagnosis of 14,636 returned travellers (cases: 1,043; prevalence: 7.1 %) aged 20–49 years of German origin consulting the outpatient travel medicine clinic of the Department of Infectious Diseases and Tropical Medicine (DITM) of the University of Munich, Germany, for the first time between 1999 and 2011
Cases (n)a (%)b
Total (n)c (%)d
Cases (n)a (%)b
Duration of travel
Type of travel
Other or unknown
Intestinal entamoeba inf.
1 (< 0.1)
Bacterial IDsh (n)
The most frequently detected protozoal diseases with the potential to induce anaemia were giardiasis (4.4 % among males/4.5 % among females), intestinal entamoeba infections (1.4/0.8 %) and malaria (0.9/0.3 %). The prevalence of anaemia was significantly (P = 0.02/P < 0.01) elevated among patients with symptomatic intestinal Entamoebahistolytica infections (30.0/33.3 %). The highest prevalence of anaemia (36.1 % of males/26.9 % of females) was found among patients with malaria. Of the 61 cases of malaria in male patients, 31 (50.8 %) were caused by Plasmodiumfalciparum, 22 (36.1 %) by P. vivax and three (4.9 %) by P. ovale; the species was unknown for the remaining five cases. Among the 31 and 22 patients with malaria caused by P. falciparum and P. vivax, respectively, we noted a significant elevation (P < 0.01 each) in the prevalence of anaemia (41.9 and 36.4 %, respectively). Of the 26 cases of malaria in female patients, 15 (57.7 %) were caused by P.falciparum, ten (38.5 %) by P. vivax and one (3.8 %) by P. ovale. Similar to the findings for male patients, we noted a significant elevation (P < 0.01) in the prevalence of anaemia (33.3 %) in the 15 female patients with malaria caused by P.falciparum (Table 2).
The proportion of patients diagnosed with malaria was significantly (P < 0.01 for males/P = 0.03 for females) higher among those who had travelled for >30 days (1.4/0.6 %). This proportion was significantly (P = 0.01 each) higher among male VFRs (2.0 %), but not among female VFRs (P = 0.73/0.4 %).
Confounding by malaria
As malaria correlates significantly with the variable “African origin”, data were restricted to patients aged 20–49 years without a diagnosis of malaria to avoid confounding by malaria; this analysis included 8,271 males/8,572 females. Among this group, the prevalence of anaemia was 5.6 % for males and 10.3 % for females and was significantly (P < 0.01 each) higher among those of African origin (20.3/26.6 %).
As malaria correlates significantly with the variables “duration of travel of >30 days” and “male VFRs”, we then analysed data obtained only from patients aged 20–49 years, of German origin and without a diagnosis of malaria to avoid confounding by malaria; this analysis included 7,087 males/7,462 females. Among this group, the prevalence of anaemia was 4.3 % for males and 9.5 % for females, and it was significantly (P = 0.03) higher among males and not significantly (P = 0.053) higher among female patients who had travelled for >30 days (5.1/10.5 %); however, it was not significantly (P = 0.06) higher among male VFRs (6.3 %).
Helminthic, bacterial and viral IDs
The most frequently detected helminthic diseases with the potential to induce anaemia were schistosomiasis (0.5 % for males/0.4 % for females), ancylostomiasis (0.2/0.1 %), ascariasis (0.1/0.1 %) and trichuriasis (0.2/0.1 %). The most frequently detected bacterial diseases with the potential to induce anaemia were campylobacter enteritis (3.3/2.8 %), shigella enteritis (1.2/1.3 %) and salmonella enteritis (1.1/1.0 %). The most frequently detected viral diseases with the potential to induce anaemia were dengue fever/dengue haemorrhagic fever (1.2/1.1 %), mononucleosis (0.4/0.2 %), cytomegalia (0.2/0.1 %) and first diagnosis of HIV infections (0.2/0.1 %). However, no significantly elevated prevalence of anaemia was not found among patients diagnosed with any of these diseases (Table 2).
Due to the small number of patients in various disease categories (n < 10 each), we did not consider the following IDs in our statistical analysis although they have been reported in the literature (either proven or suspected) to cause anaemia: Babesiosis (n = 1), brucellosis (n = 1), diphyllobothriasis (n = 1), fascioliasis (n = 2), leprosy (n = 2), parvovirus B19 (n = 2), syphilis (n = 7), tuberculosis (n = 1) and visceral leishmaniasis (n = 1).
Types of anaemia
Among the 14,636 (7,148 males/7,488 females) patients aged 20–49 years of German origin, 327 (4.6 %) males and 716 (9.6 %) females were diagnosed as anaemic. Most of the anaemic males were diagnosed with normocytic normochromic (57.5 %) anaemia, followed by macrocytic hyperchromic (21.4 %), normocytic hyperchromic (6.1 %) and microcytic hypochromic (4.6 %) anaemia. Most of the anaemic females were diagnosed with normocytic normochromic (49.0 %) anaemia, followed by macrocytic hyperchromic (21.5 %), microcytic hypochromic (8.8 %) and normocytic hypochromic (7.8 %) anaemia. Among the six anaemic patients with symptomatic intestinal E. histolytica infection and 29 anaemic patients with malaria, the great majority (83.3 and 79.3 % respectively) were diagnosed with normocytic normochromic anaemia.
Our study is the largest single-centre, single-population study performed to date on imported IDs. We analysed data on the demographics, travel itineraries and diagnosis of IDs of 22,756 returned travellers consulting the outpatient travel medicine clinic of the Ludwig-Maximilians-University of Munich between 1999 and 2011. More than 550 different IDs were confirmed based on laboratory findings, including protozoal, helminthic, bacterial and viral IDs which have been proven/suspected to cause anaemia. As travellers have an increased risk for acquiring IDs, the objective of this cross-sectional study was to assess the epidemiological significance of IDs in inducing anaemia in this large patient cohort as studies on this subject are still lacking.
To avoid confounding by age and origin, we restricted the study population to travellers aged 20–49 years of German origin. The prevalence of anaemia is known to be significantly higher among females, and therefore all data were stratified by gender. As predicted, following stratification the prevalence of anaemia among this group was found to be 4.6 % for males and 9.6 % for females. We identified three protozoal, four helminthic, three bacterial and four viral IDs with a sample size of more than ten each which were proven or suspected to cause anaemia. Among these, only malaria and symptomatic intestinal E. histolytica infection led to a significantly elevated prevalence of anaemia.
About 80 % of the anaemic patients in our study cohort with either malaria or symptomatic intestinal E. histolytica infection were diagnosed with normocytic normochromic anaemia; this result was expected as both of these IDs can induce acute anaemia. Malaria causes haemolytic anaemia 10 days or longer after the initial infection. Among the patients consulting the DITM, more than one-half of the malaria cases was caused by P.falciparum, which has been shown to induce the highest proportion of anaemia. Again this result was expected as the parasitaemia of P.falciparum is not strongly limited among patients without semi-immunity and therapy, while it is for other human pathogenic P. species [52, 53].
About 40 % of the cases of malaria were patients of African origin, while only 4.4 % of all patients aged 20–49 years were those of African origin. Consequently, the significant correlation between the variables “African origin” and “prevalence of anaemia” was confounded by malaria; however, the influence was minor. The results of a subanalysis on only patients without malaria revealed that patients of African origin still had by far the highest prevalence of anaemia, followed by patients with Asian origin. The high prevalence of anaemia among patients of Asian origin was not confounded by malaria. We found no significant association between other ethnic origins and prevalence of anaemia.
According to the WHO, anaemia is frequent in sub-/tropical countries, especially in Africa and Asia. The great majority of cases of anaemia are induced by chronic, long-lasting causes (mainly nutritional reasons, such as iron deficiency or chronic diseases) than acute causes (such as travel-associated IDs) which can lead to the development of acute anaemia within a short period of time . This might also be the case among travellers. Our results show a significant linear trend between the duration of travel and prevalence of anaemia, as chronic, long-lasting causes increase the risk for anaemia while travelling. However, this correlation was partially confounded by malaria. Among patients who had travelled for >30 days and consulted the DITM, the proportion of those diagnosed with malaria was elevated.
After exclusion of cases of malaria, the analysis still showed a significant correlation between duration of travel and anaemia. While malaria was a minor confounder of duration of travel, it was a major confounder for male VFRs. Based on our results, this population group would appear to have a significantly higher risk for anaemia than other types of travel. Among male VFRs, the proportions of cases of malaria were elevated, as were the number of travellers going to Africa, where malaria is highly endemic in most areas. However, the results of our study show that the variables “type of travel” and “destination of travel” do not significantly affect the prevalence of anaemia.
Symptomatic intestinal E. histolytica infection was the only ID other than malaria to have a significant correlation with anaemia, although more than 550 IDs were assessed in our study. Intestinal E. histolytica infections induce acute anaemia by causing intestinal bleeding events (haemorrhagic anaemia). The number of such cases detected in this study was, however, low, due to difficulties in detecting E. histolytica as stool microscopy cannot distinguish this entamoeba species from others such as E.dispar and E. moshkovskii . In contrast to malaria, intestinal E. histolytica infections did not confound any of the correlations mentioned above.
There are a number of limitations to our study. The study population was restricted to patients aged 20–49 years, as prevalence of anaemia was significantly lower in those aged <20 years and higher in those aged >49 years. Consequently, no conclusion on infection-induced anaemia can be drawn for young and elderly travellers. Our study has a cross-sectional design, and as such all data on exposure (e.g. travel-associated IDs) and outcome (anaemia) were assessed on the same day of consultation at the DITM. Consequently, no data on anaemia before travel were available. Additionally, as no follow-ups of patients were conducted, no data on the duration of IDs or anaemia which developed after consultation at the DITM could be collected. Therefore, any interpretation of causal associations between IDs and anaemia must be made with caution.
The results of this study demonstrate that among our study cohort, patients of African and Asian origin and patients of German origin who had travelled for >30 days had an elevated risk for infection-induced anaemia. The prevalence of anaemia is high among people from Africa and Asia, mainly due to chronic, long-term causal factors (especially nutritional reasons such as iron deficiency or chronic diseases) but not to acute infections. This might also be the case among the travellers assessed in our study. Travellers abroad for longer periods wee found to be more likely to develop anaemia, as chronic, long-lasting causes would appear to induce more cases of anaemia than acute travel-associated IDs. Despite the high number of IDs which have been described in the literature as proven or suspected causes of anaemia, only symptomatic intestinal E. histolytica-infections and, in particular, malaria were found to be of epidemiological significance among returned travellers included in our study. Malaria was more frequently detected at DITM and confounded the pretended elevated prevalence of anaemia among male VFRs, while cases of symptomatic intestinal E. histolytica infections were rare and without confounding effects. The results of this study show that the influence of acute travel-associated IDs on the risk of developing anaemia during or shortly after travel is very small compared the risk presented by chronic, long-lasting causes. These results may prove to be useful to clinicians who work with patients diagnosed with anaemia after their return from travelling, particularly those returning from tropical/subtropical regions.
The authors thank all patients in this study for their cooperation. The authors also wish to acknowledge their appreciation of the critical advice provided by Mrs. Catherine Mason.
Conflict of interest