Trichuris and hookworm infections associated with anaemia during pregnancy
Corresponding Author Theresa W. Gyorkos, Division of Clinical Epidemiology, McGill University Health Centre, Royal Victoria Hospital Campus, V Building, 687 Pine avenue West, Montreal, QC, Canada H3A 1A1. Tel.: +514 934 1934 ext. 44721; Fax: +514 934 8293; E-mail: firstname.lastname@example.org
Objective To assess the following associations between the second and third trimesters of pregnancy: (i) the intensity of soil-transmitted helminth (STH) infection and haemoglobin/anaemia, (ii) the effect of mebendazole treatment on the occurrence of STH infection, and (iii) the effect of mebendazole treatment on haemoglobin/anaemia.
Methods Data originated from a trial of 1042 pregnant women recruited in their second trimester and followed to delivery. Baseline assessments included socio-demographic/health information from questionnaires, haemoglobin/anaemia from HemoCue ascertainment of fingerprick blood, and the presence and intensity of STH (Ascaris lumbricoides, hookworms and Trichuris trichiura) infections from Kato–Katz examination. All women were given iron supplements; half were randomly allocated to receive single dose 500 mg mebendazole, and half, placebo. Haemoglobin/anaemia and STH infection status were determined again in the third trimester of pregnancy.
Results Complete information was available from 935 (89.7%) women. Mebendazole significantly reduced the prevalence and intensity of all three STH infections. Higher intensities of hookworm and Trichuris infections in the second trimester were associated with a higher risk of anaemia in the third trimester. Overall, women with moderate/heavy Trichuris infection were found to be at a higher risk of anaemia; the highest risk was observed among those with moderate/heavy hookworm co-infection (adjusted OR = 2.77; 95% CI: 1.26, 6.11). Mebendazole treatment did not reduce the risk of anaemia.
Conclusion Higher intensities of both Trichuris and hookworm infections are associated with anaemia in pregnancy. The importance of Trichuris infections during pregnancy requires renewed attention.
Infections au second trimestre par Trichuris et l’ankylostome, associées à l’anémie au troisième trimestre dans une population de femmes enceintes péruviennes
Objectifs: Evaluer les associations suivantes au cours du deuxième et du troisième trimestre de la grossesse: 1) l’intensité de l’infection par les helminthes transmis par le sol (HTS) et l’hémoglobine/l’anémie, 2) l’effet du traitement au mébendazole sur la présence d’infection par les HTS et 3) l’effet du traitement au mébendazole sur l’hémoglobine/l’anémie.
Méthodes: Les données proviennent d’un essai portant sur 1042 femmes enceintes recrutées au cours de leur deuxième trimestre et suivies jusqu’à l’accouchement. Les évaluations de base comprenaient des informations sociodémographiques/santéà partir de questionnaires, l’hémoglobine/l’anémie à partir du test HemoCue sur goute de sang digital et la présence et l’intensité d’infections par des HTS (Ascaris lumbricoides, Trichuris trichiura et ankylostomes) à partir de l’examen de Kato-Katz. Toutes les femmes ont reçu des suppléments de fer; la moitié d’entre elles ont été réparties aléatoirement pour recevoir d’une part, une dose unique de 500 mg de mébendazole et d’autre part, un placebo. Les statuts d’hémoglobine/anémie et d’infection par des HTS ont été déterminés à nouveau au cours du troisième trimestre de la grossesse.
Résultats: Une information complète était disponible pour 935 (89,7%) femmes. Le mébendazole a considérablement réduit la prévalence et l’intensité des infections par les trois HTS. Des intensités plus élevées d’infections par l’ankylostome et Trichuris au deuxième trimestre étaient associées à un risque plus élevé d’anémie au troisième trimestre. Globalement, les femmes avec une infection modérée/forte àTrichuris ont été trouvées avec un risque plus élevé d’anémie, le risque le plus élevéétant observé chez celles avec une coinfection modérée/forte avec l’ankylostome (OR ajusté = 2,77; IC95%: 1,26-6,11). Le traitement au mébendazole n’a pas réduit le risque d’anémie.
Conclusions: Des intensités d’infections plus élevées par Trichuris et ankylostomes sont associées à l’anémie pendant la grossesse. L’importance des infections àTrichuris pendant la grossesse nécessite une attention renouvelée.
Asociación entre las infecciones por Trichuris y anquilostoma durante el segundo semestre y la anemia en el tercer trimestre, en mujeres peruanas embarazadas
Objetivos: Evaluar las siguientes asociaciones entre el segundo y tercer trimestre de embarazo: 1) la intensidad de la infección por helmintos transmitidos por el suelo (HTS) y hemoglobina/anemia; 2) el efecto del mebendazol en la incidencia de infecciones por HTS; y 3) el efecto del tratamiento con mebendazol sobre la hemoglobina/anemia.
Métodos: Los datos se originaron en un ensayo de 1042 mujeres embarazadas, reclutadas en su segundo trimestre y seguidas hasta dar a luz. Las evaluaciones de base incluían información socio-demográfica/salud de cuestionarios, hemoglobina/anemia constatado por HemoCue en sangre obtenida por punción en el dedo, y la presencia e intensidad de infecciones por HTS (Ascaris lumbricoides, anquilostoma y Trichuris trichiura) mediante la técnica de Kato-Katz. Todas las mujeres recibieron suplementos de hierro; la mitad fueron asignadas de forma aleatoria para recibir una dosis única de 500 mg de mebendazol, y la mitad, placebo. Hemoglobina/anemia, y el estatus de infección por HTS se determinaron de nuevo durante el tercer trimestre de embarazo.
Resultados: La información completa estaba disponible para 935 (89.7%) mujeres. El mebendazol redujo significativamente la prevalencia e intensidad de todas las tres infecciones por HTS. Una mayor intensidad de infección por anquilostoma y Trichuris durante el segundo trimestre estaba asociada a un mayor riesgo de anemia en el tercer trimestre. En general, las mujeres con una infección moderada/alta por Trichuris estaban en mayor riesgo de anemia; el mayor riesgo se observó entre aquellas con una co-infección moderada/alta por anquilostoma (OR ajustado = 2.77; 95% IC: 1.26, 6.11). El tratamiento con mebendazol no redujo el riesgo de anemia.
Conclusiones: Una alta intensidad de infecciones por Trichuris y anquilostoma estaba asociada con anemia durante el embarazo. La importancia de infecciones por Trichuris durante el embarazo requiere una atención renovada.
Anaemia is a major global public health problem. More than 50% of pregnant women, especially in developing countries, are estimated to be anaemic, frequently due to iron deficiency (WHO 2001; Kalaivani 2009). Where the prevalence of anaemia exceeds 20%, as in pregnant women and preschool-age children in the Amazon region of Peru, WHO categorizes its public health significance as moderate to severe (WHO 2001; Larocque et al. 2006; Casapia et al. 2007). The principal risk factors for iron-deficiency anaemia are insufficient dietary intake of iron, heavy blood loss at menstruation and parasite infections (de Benoist et al. 2008). Soil-transmitted helminths (STHs), such as hookworms (Necator americanus and Ancylostoma duodenale) and whipworms (Trichuris trichiura), contribute to iron-deficiency anaemia by ingesting blood and by damaging the intestinal mucosa during feeding. Two recent systematic reviews, one in pregnant women (Brooker et al. 2008) and the other in non-pregnant populations (Smith & Brooker 2010) documented consistent and global evidence that hookworm infections are associated with lower haemoglobin values.
In contrast, studies of the association between Trichuris infection and anaemia or haemoglobin levels, mostly conducted in child populations yielded mixed results (Robertson et al. 1992; Ramdath et al. 1995; Kung’u et al. 2009; Knopp et al. 2010; Midzi et al. 2010; Quihui-Cota et al. 2010). Studies specifically focussing on pregnant women found no independent association between Trichuris and anaemia/haemoglobin (Nurdiati et al. 2001; Muhangi et al. 2007; Ndyomugyenyi et al. 2009).
Estimates of STH global prevalence vary but most reports concur that between one and two billion people, globally, are infected with an STH at any one moment in time, of whom pregnant women infected with hookworms constitute an important component (Bundy et al. 1995; de Silva et al. 2003; WHO 2005; Brooker 2010).
An analysis of the baseline (second trimester) data for an RCT investigating the effectiveness of deworming and iron supplementation in a population of more than 1000 pregnant women living in and around Iquitos in the Peruvian Amazon showed that women with moderate-to-heavy intensities of both hookworm and Trichuris infection were more than twice as likely to concurrently have anaemia as those with no or light infection (Larocque et al. 2005). An analysis of the trial proper documented the effect of mebendazole and iron supplementation on anaemia and on the presence and intensity of STH infections (Larocque et al. 2006). However, the interrelationships among STH infections (and in particular Trichuris infections), mebendazole treatment and anaemia during pregnancy (i.e. between the second and third trimesters) have not been previously studied.
Therefore, the objectives of our study were to assess: (i) the association between the intensity of hookworm and/or Trichuris infection in the second trimester and haemoglobin/anaemia in the third trimester, (ii) the effect of mebendazole treatment in the second trimester on the occurrence of STH infection in the third trimester, and (iii) the effect of mebendazole treatment in the second trimester on haemoglobin (and anaemia) status in the third trimester.
The data originated from a randomized clinical trial investigating the effect of mebendazole on birth weight (Larocque et al. 2006). Briefly, 1042 pregnant women were recruited in their second trimester. A questionnaire on socio-demographic and health information was administered, and fingerpick blood and stool specimens were collected. All women were given iron supplements (60 mg elemental iron/day); half were randomly allocated to receive single dose 500 mg mebendazole, and half, placebo. Blood and stool specimens were again collected during the third trimester of pregnancy. Haemoglobin concentration was measured by HemoCue, and malaria was determined by thick and thin smear. Intestinal helminths were quantified using the Kato–Katz method. Women with a haemoglobin level below 7 g/dl in the second trimester were excluded from further study and referred for clinical follow-up.
Infection intensity categories proposed by WHO (1998) were used. Anaemia was defined as a haemoglobin level below 11 g/dl, the threshold recommended by WHO for pregnant women (de Benoist et al. 2008).
The effectiveness of mebendazole against helminth infection was determined by comparing infection prevalences in the third trimester with prevalences at baseline (i.e. second trimester) between the mebendazole and placebo groups, using chi-squared tests. This analysis was carried out separately in women with and without infection in the second trimester.
Associations between independent variables and haemoglobin levels were determined by linear regression, and associations between independent variables and the risk of anaemia in the third trimester were determined by logistic regression analysis. First, simple regression analyses were performed and variables showing significant or near-significant (P < 0.10) associations with haemoglobin or anaemia were included in multiple regression models. For anaemia, crude and adjusted odds ratios and 95% confidence intervals were computed from the logistic regressions. For all the above analyses, independent variables tested included the intervention group (iron + mebendazole or iron + placebo) and presence of infection by hookworms, Ascaris and Trichuris in the second trimester. The reason for including helminth infections in the second rather than the third trimester is that those in the second trimester precede administration of the intervention. Data were analysed using SAS 9.1 software. The RCT had obtained Canadian and Peruvian ethical approvals and is registered (ISRCTN08446014).
Among the 1042 women included in the RCT (conducted in 2003), 935 (89.7%) had complete information and haemoglobin measures in both second and third trimesters. This included 474 women in the iron + mebendazole group and 461 in the iron + placebo group. Baseline characteristics of these pregnant women and associations between the prevalence of STH infections and anaemia in the second trimester are described in detail elsewhere (Larocque et al. 2005, 2006). Prevalences of hookworm, Ascaris and Trichuris infections in the second trimester were 46.3%, 63.9% and 82.0% respectively. However, while the prevalences of moderate-to-heavy Ascaris and Trichuris infections were, respectively, 32.6% and 29.1%, the prevalence of moderate-to-heavy hookworm infection was much lower (i.e. 5.2%). Lower education was associated with a significantly higher risk of infection by all three helminths under study while living in rural areas was associated with a higher risk for hookworm infection only (Table 1).
Table 1. Association between socio-economic variables and the prevalence of STH infections in 935 second-trimester pregnant women in Iquitos, Peru in 2003
| Less than primary||138||5.84||3.69–9.23||1.46||0.95–2.27||2.93||1.66–5.17|
| Primary completed||558||2.94||2.10–4.11||1.41||1.03–1.92||2.75||1.91–3.96|
| Secondary completed||239||1.00||–||1.00||–||1.00||–|
|Area of residence|
The prevalences and intensities of all three STH infections in the third trimester were significantly lower in the mebendazole group than in the placebo group, both in women who tested positive for each respective STH infection earlier in their second trimester and in women who had earlier tested negative for these respective STHs (Table 2; intensity data not shown).
Table 2. Effect of mebendazole treatment in the second trimester on the prevalence of STH infections in the third trimester
|Ascaris||No||Placebo + iron||162||109||67.3||53||32.7||<0.001|
|Mebendazole + iron||176||161||91.5||15||8.5|
|Yes||Placebo + iron||299||25||8.4||274||91.6||<0.001|
|Mebendazole + iron||298||232||77.9||66||22.1|
|Hookworms||No||Placebo + iron||253||222||87.7||31||12.3||0.002|
|Mebendazole + iron||249||238||95.6||11||4.4|
|Yes||Placebo + iron||208||24||11.5||184||88.5||<0.001|
|Mebendazole + iron||225||69||30.7||156||69.3|
|Trichuris||No||Placebo + iron||80||58||72.5||22||27.5||0.015|
|Mebendazole + iron||88||77||87.5||11||12.5|
|Yes||Placebo + iron||381||17||4.5||364||95.5||<0.001|
|Mebendazole + iron||386||161||41.7||225||58.3|
Overall, the mean haemoglobin level increased from 11.03 to 11.39 g/dl between the second and the third trimester (i.e. a mean increase of 0.36 g/dl (95% CI: 0.28 to 0.44 g/dl). After adjusting for confounders, both hookworm and Trichuris infections were associated with haemoglobin status in the third trimester (Table 3). For every additional 1000 epg increment in either hookworm or Trichuris infection, the haemoglobin level fell (by 0.081 g/dl among hookworm-infected women and by 0.046 g/dl among Trichuris-infected women). Trichuris was the only STH found to be associated with the variation in haemoglobin level between the second and third trimester, with each 1000 epg increment in intensity being associated with a 0.045 g/dl reduction in haemoglobin variation after adjustment for covariates.
Table 3. Determinants of haemoglobin in the third trimester and of the variation in haemoglobin levels between the second and the third trimesters
|Age||Per 10 years||0.106||0.102||–||–||−0.044||0.532||–||–|
|Area of residence||Periurban||reference||–||–||–||reference||–||–||–|
|Education||Less than primary||−0.078||0.514||–||–||−0.226||0.081||−0.080||0.469|
|Gestational age||Per 4 weeks||0.066||0.248||–||–||0.183||0.003||0.111||0.208|
|Time interval between first and second assessments||Per 4 weeks||−0.049||0.356||–||–||−0.138||0.014||0.003||0.970|
|Group||Placebo + iron||reference||–||–||–||reference||–||–||–|
|Mebendazole + iron||−0.006||0.937||–||–||−0.031||0.693||–||–|
|Haemoglobin in second trimester||Per g/dl||–||–||–||–||−0.583||<0.001||−0.581||<0.001|
|Ascaris in second trimester||Per 10 000 epg||−0.035||0.124||–||–||−0.020||0.413||–||–|
|Hookworm in second trimester||Per 1 000 epg||−0.102||<0.001||−0.081||0.009||0.006||0.866||–||–|
|Trichuris in second trimester||Per 1 000 epg||−0.054||<0.001||−0.046||0.002||−0.039||0.011||−0.045||<0.001|
Heavy hookworm infection and both moderate and heavy Trichuris infections in the second trimester were significantly associated with an increased risk of anaemia in the third trimester, but after adjustment for confounders, only Trichuris infections remained statistically significantly associated with anaemia (Table 4). The association between Trichuris infection in the second trimester and the risk of anaemia in the third trimester remained statistically significant for all categories of hookworm infection intensity, even the hookworm-uninfected group (Table 5). The magnitude of the association between Trichuris and anaemia was not changed by adjustment for intervention group (data not shown).
Table 4. Second-trimester determinants for the risk of anaemia in the third trimester
| Less than primary||138||48||34.8||0.88||0.57–1.37||–||–|
| Primary completed||558||186||33.3||0.83||0.60–1.14||–||–|
| Secondary completed||239||90||37.7||1.00||–||–||–|
|Zone of residence|
|Gestational age at first assessment|
| Fourth month||483||165||34.2||1.00||–||–||–|
| Fifth month||293||108||36.9||1.13||0.83–1.52||–||–|
| Sixth month||159||51||32.1||0.91||0.62–1.33||–||–|
|Time interval between first and second assessment|
| 28–61 days||255||88||34.5||1.00||–||1.00||–|
| 62–83 days||217||68||31.3||0.87||0.59–1.27||0.82||0.55–1.21|
| 84–91 days||269||85||31.6||0.88||0.61–1.26||0.86||0.60–1.25|
| 92–147 days||194||83||42.8||1.42||0.97–2.08||1.35||0.92–2.00|
|Ascaris in second trimester|
| None (0 epg)||338||114||33.7||1.00||–||–||–|
| Light (1–4 999 epg)||292||99||33.9||1.01||0.72–1.40||–||–|
| Moderate (5 000–49 999 epg)||284||102||35.9||1.10||0.79–1.53||–||–|
| Heavy (50 000–166 800 epg)||21||9||42.9||1.47||0.60–3.60||–||–|
|Hookworms in second trimester|
| None (0 epg)||502||160||31.9||1.00||–||1.00||–|
| Light (1–1 999 epg)||384||143||37.2||1.27||0.96–1.68||1.11||0.82–1.49|
| Moderate (2 000–3 999 epg)||33||12||36.4||1.22||0.59–2.54||0.91||0.43–1.95|
| Heavy (4 000–13 848 epg)||16||9||56.3||2.75||1.01–7.51||2.05||0.72–5.83|
|Trichuris in second trimester|
| None (0 epg)||168||45||26.8||1.00||–||1.00||–|
| Light (1–999 epg)||495||159||32.1||1.29||0.88–1.91||1.25||0.84–1.87|
| Moderate (1 000–9 999 epg)||251||109||43.4||2.10||1.37–3.20||1.95||1.25–3.05|
| Heavy (10 000–25 200 epg)||21||11||52.4||3.01||1.20–7.56||2.63||1.02–6.79|
Table 5. Association between hookworm and Trichuris co-infection in the second trimester of pregnancy and risk of anaemia in the third trimester of pregnancy
|1 000–25 200||94||40||42.6||2.22||1.27–3.89||2.13||1.21–3.74|
|1 000–25 200||145||64||44.1||2.37||1.43–3.92||2.40||1.44–3.97|
|2 000–13 848||0–999||16||5||31.3||1.36||0.44–4.20||1.31||0.42–4.07|
|1 000–25 200||33||16||48.5||2.82||1.29–6.17||2.77||1.26–6.11|
Deworming pregnant women with mebendazole in their second trimester was found to significantly reduce all three STH infections (i.e. Ascaris, Trichuris and hookworm) in the third trimester. This reduction was observed in pregnant women who had been both infected and uninfected in their second trimester, demonstrating the beneficial effect of preventive chemotherapy on infections not detected on a screening stool examination. While this result reflects the less than optimal positive predictive value of the Kato–Katz test, it does provide empirical evidence in favour of including deworming in antenatal care programmes in areas of high STH endemicity, as advocated by WHO (Larocque & Gyorkos 2006; WHO 2006).
Trichuris infection in the second trimester of pregnancy was a significant and independent determinant of lower haemoglobin levels and anaemia in the third trimester. The negative association between the intensity of Trichuris infection and the increase in haemoglobin indicates that this parasite reduces the efficacy of iron supplementation. Thus, even though all the pregnant women in this trial had received iron supplementation as per protocol (Larocque et al. 2006), and half had additionally received mebendazole, higher intensities of Trichuris in the second trimester still contributed to anaemia in the third trimester. Moreover, co-infection with high intensities of hookworm infections heightened the impact of Trichuris infections on anaemia.
The association between Trichuris infection and the risk of anaemia is consistent with findings from previous studies in children (Robertson et al. 1992; Ramdath et al. 1995; Quihui-Cota et al. 2010) but contrasts with findings of no association in three studies of pregnant women (Nurdiati et al. 2001; Muhangi et al. 2007; Ndyomugyenyi et al. 2009). This apparent discrepancy may be explained in part by study settings, as two of the pregnant populations had a Trichuris infection prevalence of less than 10% (Muhangi et al. 2007; Ndyomugyenyi et al. 2009). In the third study, although the prevalence of Trichuris ranged between 40% and 50%, depending on trimester, most intensities were classified as low (Nurdiati et al. 2001). Stephenson et al. (2000) reviewed the mechanisms by which Trichuris infection causes anaemia. These include blood loss from mechanical damage by adult worms in burrowing into the intestinal epithelium and their subsequent feeding behaviours, and infection-attributable loss of appetite resulting in lower food (and iron) intake. The impact of both these direct and indirect mechanisms rises as the worm burden grows and results in increasingly malnourished states.
Heavy hookworm infection in the second trimester was found to be associated with anaemia in the third trimester but the association was no longer statistically significant after adjustment for potential confounders. This may be due to the small number of pregnant women detected with heavy hookworm infections in their second trimester. Hookworm-Trichuris co-infection was nonetheless important in raising the risk of anaemia in the third trimester.
Recent systematic reviews which summarize the effectiveness of mebendazole on haemoglobin levels or anaemia in either pregnant or non-pregnant populations document the insufficient nature of cumulative evidence to date on this topic (Brooker et al. 2008; Smith & Brooker 2010). However, the facts that even moderate intensity Trichuris infections alone, or co-occurring with hookworm, significantly increases the risk of anaemia during pregnancy and that important reductions in STH infection and anaemia (e.g. Torlesse & Hodges 2001) can be achieved with different deworming drugs, lend support to integrating deworming into antenatal care programmes (after the first trimester), especially in areas where either hookworm or Trichuris infections are of public health importance. The deworming drugs recommended by WHO during pregnancy are the same as for other population subgroups, namely albendazole, mebendazole, levamisole and pyrantel (Savioli et al. 2003; WHO 2006). Although higher efficacy rates may be obtained with doses administered over several days, because of their single dose format albendazole and mebendazole are the most frequently used in large-scale programmes (Albonico et al. 2006; Bethony et al. 2006). Drug-specific efficacy rates differ for hookworm infections and for Trichuris infections and may be low, but egg reduction rates, in contrast, are normally much higher (Keiser & Utzinger 2008). Consideration of the relative proportions of hookworm and Trichuris worm burdens present in a programme’s target populations, in addition to issues of availability and cost, will inform the choice of deworming drug.
Ultimately, this type of deworming intervention should complement a more comprehensive and community-wide approach to the prevention and control of STH infections, especially through improvements in dietary nutrition, hygiene education and sanitation.
This data analysis was conducted within the framework of a Canadian Institutes of Health Research Interdisciplinary Enhancement Grant (#HOA-80064) awarded to Theresa W. Gyorkos. The Fonds de la recherche en santé du Québec provides partial institutional support to TWG’s research programme at the Research Institute of the McGill University Health Centre.