ABSTRACT
Background
Eosinophilia is a common clinical finding with diverse etiologies. Helminthiasis is a major aetiology in tropical regions, but data from haematology consultation services at tertiary centres in endemic settings are limited.
Objectives
To determine the prevalence of helminthiasis among patients with eosinophilia at a tropical tertiary care centre and identify clinical and laboratory characteristics differentiating helminth from non‐helminth etiologies.
Methods
A prospective cohort study with predefined retrospective outcome classification was conducted at Siriraj Hospital, Thailand (January 2022 to November 2023) in adult patients with blood eosinophilia above 500 cells/μL. All patients underwent helminth screening through stool examination and serological testing for strongyloidiasis, gnathostomiasis, angiostrongyliasis, filariasis and cysticercosis, and received empirical ivermectin or albendazole with monthly follow‐up for 6 months.
Results
Among 131 patients, 76 (58.0%) were diagnosed with helminthiasis, comprising 49 definite (laboratory‐confirmed) and 27 probable (treatment‐response‐defined) infections. Strongyloidiasis (53.1%) was most common among definite infections, followed by gnathostomiasis (40.8%) and cysticercosis (26.5%), with multiple infections in 24.5% of definite cases. The median time to absolute eosinophil count (AEC) normalization was 1.0 month (IQR: 1.0–2.0), with 93.4% normalizing within 3 months. Probable infections were significantly associated with soil contact and forest travel (all p < 0.05), while definite infections were associated with organism‐specific exposures (all p < 0.05). Patients aged 75 years or above showed significantly lower AEC at presentation (adjusted geometric mean ratio 0.52, 95% CI: 0.31–0.89, p = 0.017). Leukopenia and hypoalbuminemia were more prevalent in non‐infectious etiologies.
Conclusions
Helminthiasis was identified in more than half of patients with eosinophilia at a tropical tertiary haematology service. Empirical anthelmintic therapy effectively normalized eosinophilia in the majority within 3 months, supporting its use as a standard initial approach in endemic settings. Detailed exposure anamnesis and concurrent evaluation for non‐helminth etiologies, particularly in patients with leukopenia or hypoalbuminemia, should be routinely performed.
Keywords: drug reactions, eosinophilia, helminths, prevalence, tropical medicine
1. Introduction
Eosinophilia, defined as a blood eosinophil count above 0.5 × 106 cells/mL in adults, is a common clinical finding with diverse etiologies. It is classified into three main categories: primary, secondary (reactive) and idiopathic. Primary eosinophilia or clonal eosinophilia is rare and arises from clonal haematopoiesis in hematologic malignancies. Secondary eosinophilia is a predominant cause, encompassing helminthiasis, allergic diseases, autoimmune disorders, neoplastic conditions and adverse drug reactions. Idiopathic eosinophilia is diagnosed when no aetiology is identified after thorough evaluation [1, 2]. Systematic assessment of secondary etiologies should take precedence in clinical practice, as secondary eosinophilia frequently resolves with targeted treatment of the underlying condition [1, 2, 3].
Helminthiasis‐associated eosinophilia is primarily driven by tissue‐migratory or tissue‐dwelling species. Common geohelminths include Ascaris lumbricoides, Strongyloides stercoralis, Trichuris trichiura and hookworm species [4, 5]. In Southeast Asia including Thailand, Gnathostoma spp., Opisthorchis viverrini and filarial species are regionally important contributors [4, 6]. Confirming a definitive helminth diagnosis is challenging due to inherent limitations of available detection platforms, variable serological test performance, widespread cross‐reactivity and the clinically silent nature of many infections, particularly at low worm burdens [4]. These system‐level diagnostic constraints render routine laboratory confirmation impractical in many clinical settings, even when helminthiasis remains the most probable aetiology.
Consequently, clinicians in endemic areas frequently administer empirical anthelmintic therapy using albendazole or ivermectin without awaiting laboratory confirmation [4]. Existing literature on helminthiasis‐associated eosinophilia has largely focused on travellers at travel medicine clinics [7, 8, 9] or refugee populations, [10, 11] where reported helminth prevalence ranges from 14% to 40%. Data from haematology services at tropical tertiary centres are notably absent, despite their substantially higher pre‐test probability of non‐helminth etiologies compared to travel clinic settings. This study therefore aimed to determine the prevalence of helminthiasis among patients presenting with eosinophilia to a haematology consultation service at a tropical tertiary centre in Thailand, and to identify exposure histories, clinical characteristics and laboratory parameters associated with helminthiasis.
2. Materials and Methods
This prospective cohort study with predefined retrospective outcome classification was conducted at the Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand, from January 2022 to November 2023. The study protocol was approved by the Siriraj Institutional Review Board (approval number Si 932/2020) and registered with the Thai Clinical Trial Registry (TCTR20211129002). All procedures adhered to the Declaration of Helsinki and its amendments. This study was conducted and reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.
The haematology consultation service identified all adult patients with newly detected eosinophilia during the study period. All enrolled patients were local residents in the catchment area of Siriraj Hospital or were referred from regional hospitals in Thailand; travellers, tourists and recent international arrivals were not included in this cohort. Inclusion criteria were age 18 years or older, absolute eosinophil count (AEC) above 500 cells/μL, and willingness to receive treatment and follow‐up for at least 6 months after diagnosis at the study institution. Written informed consent was obtained from all participants. Baseline demographic data, helminth exposure histories, allergy history and laboratory parameters including complete blood count, creatinine and liver function tests were collected using a standardized questionnaire. Patients were followed monthly for 6 months to enable systematic monitoring of treatment responses. Data were de‐identified and stored securely following institutional protocols. The primary objective was to determine the prevalence of helminthiasis among patients with eosinophilia. Secondary objectives were to identify exposure histories, clinical characteristics and laboratory parameters associated with helminthiasis.
2.1. Detection of Intestinal Helminths
Stool samples were examined for ova and parasites using the formalin‐ethyl acetate sedimentation technique, in which approximately 1 g of filtered stool was mixed with 10 mL of formalin‐ethyl acetate solution at a 3:1 ratio, centrifuged at 3000 rpm for 1 min and the resulting sediment examined microscopically.
2.2. Detection of Tissue Helminths by Immunodiagnosis
Serum samples were tested for strongyloidiasis, cysticercosis, gnathostomiasis, angiostrongyliasis and filariasis using validated immunoassays. Strongyloides IgG/IgM antibody was detected by ELISA (Abcam, Cambridge, United Kingdom). Cysticercus antibody was detected by Taenia solium IgG ELISA (Tecan Trading AG, Switzerland). Gnathostoma antibody was detected by in‐house immunoblotting as previously described [12]. Angiostrongylus cantonensis antibody was detected by ELISA and confirmed by immunoblotting [13, 14]. Filarial IgG4 antibody was detected by immunochromatography [15, 16].
2.3. Criteria for Diagnosis and Severity Classification
Eosinophil counts were recorded at diagnosis and at monthly intervals for 6 months following anthelmintic initiation. Eosinophilia severity was classified per WHO and International Consensus Classification criteria as mild (500–1500 cells/μL), moderate (1500–5000 cells/μL), or severe (above 5000 cells/μL) [17]. All enrolled patients received empirical anthelmintic treatment before definitive diagnosis, either ivermectin (200 μg/kg/day for three consecutive days) or albendazole (400 mg twice daily for 7 days), with the agent selected based on physician assessment and drug availability. Treatment was modified when specific helminth infections were subsequently confirmed on laboratory testing.
2.4. Classification of Study Groups
Patients were classified into three groups based on predefined criteria applied retrospectively after 6 months of follow‐up. The definite helminth infection group included patients with at least one positive laboratory result for helminthiasis whose eosinophil count normalized within 6 months. The probable helminth infection group comprised patients with negative laboratory results whose eosinophil count nonetheless normalized within 6 months and in whom no alternative aetiology was identified. The non‐infectious group comprised two distinct subsets. The first subset included patients with a clearly identified non‐helminth aetiology established at the time of eosinophilia diagnosis, such as drug reaction or hematologic malignancy, in whom a non‐parasitic cause was evident from the outset. The second subset included patients with negative helminth serology and stool examination who did not achieve eosinophil count normalization after completing empirical anthelmintic treatment; in these patients, the most probable underlying aetiology of eosinophilia was identified concurrently during the follow‐up period through ongoing clinical assessment, with conditions such as allergic disorders representing the most commonly assigned diagnosis in this subset. It is acknowledged that this second subset represents a diagnostically challenging population in whom concurrent subclinical helminth infection cannot be definitively excluded, as eosinophilia driven by a non‐helminth condition may persist despite successful helminth clearance.
2.5. Statistical Analysis
Data are presented as mean ± standard deviation or median with interquartile range for continuous variables, and as frequency with percentage for categorical variables. Group comparisons used the independent t‐test or Mann–Whitney U test for continuous variables, the Kruskal–Wallis test for more than two groups and the Chi‐squared or Fisher exact test for categorical variables. Log‐transformed AECs were analysed using multivariable linear regression adjusted for age, sex, hypertension, dyslipidemia, chronic kidney disease and diabetes mellitus, with results expressed as geometric mean ratios [18]. Missing outcome data were imputed under the missing at random assumption using multiple imputation with auxiliary variables showing a correlation coefficient above 0.1 with missingness, with the number of imputations set equal to the percentage of missing data [19]. A complete‐case sensitivity analysis was performed. Given the exploratory nature of multiple comparisons across exposure and laboratory variables, all analyses should be interpreted with appropriate caution regarding type I error. Prevalence ratios for risk factors and hematologic parameters were estimated using Poisson regression with robust variance estimation. All analyses were performed using R version 4.2.0 (R Foundation for Statistical Computing, Vienna, Austria). A p value below 0.05 was considered statistically significant.
3. Results
3.1. Patient Characteristics
A total of 145 patients were initially enrolled; however, 14 patients were lost to follow‐up before completing the 6‐month protocol and were therefore excluded from the final analysis, resulting in 131 patients available for analysis. There was a slight male predominance (59.5%) and a mean age of 62 ± 17 years. Most patients (72.5%) had underlying comorbidities, predominantly hypertension and dyslipidemia. The median white blood cell count was 7500/μL and the median AEC was 1224/μL. Eosinophilia severity was mild in 61.1%, moderate in 31.3% and severe in 7.6% of patients. Demographic data of the patients included in this study are presented in Table 1.
TABLE 1.
Demographic data of eosinophilia patients participated in this study based on the response to empirical anthelminthic treatment.
| Characteristic | Total (n = 131) | Non‐infectious (n = 55) | Helminth infection a (n = 76) | p |
|---|---|---|---|---|
| Age (years) | 62 ± 17 | 63 ± 17 | 61 ± 16 | 0.25 |
| Male sex | 78 (59.5) | 32 (58.2) | 46 (60.5) | 0.79 |
| Any comorbidity | 95 (72.5) | 43 (78.2) | 52 (68.4) | 0.22 |
| DM | 46 (35.1) | 16 (29.1) | 30 (39.5) | 0.22 |
| HT | 77 (58.8) | 35 (63.6) | 42 (55.3) | 0.34 |
| DLP | 67 (51.1) | 27 (49.1) | 40 (52.6) | 0.69 |
| IHD | 5 (3.8) | 1 (1.8) | 4 (5.3) | 0.40 |
| CKD | 19 (14.5) | 7 (12.7) | 12 (15.8) | 0.62 |
| Cirrhosis | 4 (3.1) | 2 (3.6) | 2 (2.6) | 1.00 |
| Lymphoma | 2 (1.5) | 2 (3.6) | 0 (0.0) | 0.17 |
| Referred from other hospital | 16 (12.2) | 7 (12.7) | 9 (11.8) | 1.00 |
| Initial WBC (/μL) | 7500 (5860, 10,000) | 7280 (5500, 10,040) | 7677 (6100, 9655) | 0.44 |
| Leukocytosis (WBC > 11,000/μL) | 22 (16.8) | 11 (20.0) | 11 (14.5) | 0.40 |
| Leukopenia (WBC < 4500/μL) | 10 (7.6) | 8 (14.5) | 2 (2.6) | 0.017 |
| The initial percentage of eosinophils | 17.2 (12.2, 26.0) | 19.6 (12.5, 28.9) | 15.9 (12.1, 25.0) | 0.28 |
| Initial absolute eosinophil count (/μL) | 1224 (785, 2277) | 1280 (738, 2457) | 1220 (839, 2096) | 0.96 |
| Initial absolute eosinophil count | 0.40 | |||
| ≥ 500 to < 1500 | 80 (61.1) | 34 (61.8) | 46 (60.5) | |
| ≥ 1500 to < 5000 | 41 (31.3) | 15 (27.3) | 26 (34.2) | |
| ≥ 5000 | 10 (7.6) | 6 (10.9) | 4 (5.3) | |
| Initial Hb (g/dL) | 11.6 ± 2.2 | 11.2 ± 2.2 | 11.9 ± 2.1 | 0.055 |
| Missing | 1 | 0 | 1 | |
| Initial platelet count (/μL) | 233,489 ± 94,851 | 239,764 ± 81,804 | 228,947 ± 103,569 | 0.31 |
| Creatinine | 0.95 (0.77, 1.35) | 1.00 (0.78, 1.50) | 0.91 (0.77, 1.26) | 0.27 |
| Missing | 2 | 1 | 1 | |
| AST | 22 (18, 28) | 23 (19, 28) | 21 (18, 28) | 0.44 |
| Missing | 5 | 2 | 3 | |
| ALT | 17 (13, 23) | 17 (15, 23) | 18 (13, 24) | 0.73 |
| Missing | 4 | 2 | 2 | |
| Total bilirubin | 0.5 (0.3, 0.7) | 0.5 (0.3, 0.8) | 0.5 (0.3, 0.7) | 0.79 |
| Missing | 8 | 3 | 5 | |
| Serum globulin | 3.2 (2.9, 3.6) | 3.2 (3.0, 3.5) | 3.2 (2.8, 3.6) | 0.93 |
| Missing | 8 | 3 | 5 | |
| Serum albumin | 4.0 ± 1.2 | 3.8 ± 1.2 | 4.1 ± 1.1 | 0.006 |
| Missing | 4 | 1 | 3 | |
| Serum albumin < 3.5 mg/dL | 23 (18.1) | 16 (29.6) | 7 (9.6) | 0.004 |
| Serum albumin > 5.5 mg/dL | 2 (1.6) | 1 (1.9) | 1 (1.4) | 1.00 |
Note: Continuous data are presented as mean ± SD or median (Q1, Q3). Categorical data are presented as n (%). Missing data were identified across several laboratory parameters, with different patients contributing to missing values for different variables; total bilirubin and globulin had the highest proportion of missing values, affecting eight patients (6.1% of the cohort). Given that missing data were confined to a small number of patients per variable, the overall impact on statistical analyses is considered negligible. Bold values indicate statistical significance (p < 0.05).
Abbreviations: ALT, alanine transaminase; AST, aspartate transaminase; CKD, chronic kidney disease; DLP, dyslipidemia; DM, diabetes mellitus; Hb, haemoglobin; HT, hypertension; IHD, ischemic heart disease; WBC, white blood cell count.
Definite helminth infection and probable helminth infection groups.
Compared with the helminth infection group, the non‐infectious group had significantly higher rates of leukopenia (WBC below 4500/μL; 14.5% vs. 2.6%, p = 0.017) and hypoalbuminemia (serum albumin below 3.5 g/dL; 29.6% vs. 9.6%, p = 0.004). Mean albumin levels were significantly lower in the non‐infectious group (3.8 ± 1.2 vs. 4.1 ± 1.1 g/dL, p = 0.006). Other hematologic parameters showed no significant differences between groups (Table 1).
3.2. Prevalence of Helminth Infection
Forty‐nine patients (37.4%) met criteria for definite helminth infection and 27 (20.6%) for probable helminth infection, yielding a combined helminth prevalence of 58.0%. The remaining 55 patients (42.0%) were classified as non‐infectious. Among definite infections, strongyloidiasis was the most common (26 of 49, 53.1%), followed by gnathostomiasis (20 of 49, 40.8%) and cysticercosis (13 of 49, 26.5%). No intestinal parasites other than Strongyloides were identified on stool examination. Serological testing was positive for tissue parasites in 30 patients. Single‐helminth infections accounted for 75.5% (37 of 49) of definite cases, while multiple helminth infections were present in 24.5% (12 of 49). The most common co‐infections were strongyloidiasis with gnathostomiasis and the combination of strongyloidiasis, cysticercosis and filariasis, each occurring in three patients (Figure 1).
FIGURE 1.
Upset plot of single and multiple definite infections.
3.3. Treatment Response After Receiving Empirical Anthelmintic Drugs
Most patients received ivermectin (119 of 131, 90.8%), with the remainder receiving albendazole (12 of 131, 9.2%). The median response time was 1 month (IQR: 1.0–2.0), with 93.4% of patients achieving normalization by 3 months of follow‐up (Figure 2). Response time varied by infection status: patients with single definite helminth infections had a median response time of 1 month (IQR: 1–2), while those with multiple definite infections required longer (median 1.5 months, IQR: 1–2.5). Patients with probable infections had a median response time of 1 month (IQR: 1–2).
FIGURE 2.
Response time to achieve normalization of absolute eosinophil count in definite and probable helminth infection groups after the initiation of empirical anthelminthic medication. This figure shows the percentage of patients whose eosinophils returned to normal levels. AEC, absolute eosinophil count.
3.4. Comparison of Helminth Exposure Risk Factors
Exposure histories and clinical findings are detailed in Table 2. Walking barefoot on soil was the only significant univariate risk factor distinguishing the helminth group from the non‐infectious group (13.2% vs. 1.8%, p = 0.025). Organism‐specific analyses revealed distinct exposure patterns. Gnathostoma spp. infection was significantly associated with contact with dirty water (RR: 2.74, 95% CI: 2.03, 3.71; p < 0.001). Cysticercosis was associated with contact with dirty water (RR: 5.08, 95% CI: 3.11, 8.29, p < 0.001) and eating raw pork (RR: 4.79, 95% CI: 2.99, 7.65, p < 0.001). Filarial infection was associated with forest travel or residence in an endemic area (RR: 8.57, 95% CI: 4.25, 17.3, p < 0.001). In the probable helminth infection group, walking barefoot (RR: 2.68, 95% CI: 1.53, 4.69, p = 0.001), soil contact occupation (RR: 2.20, 95% CI: 1.13, 4.28; p = 0.020) and forest travel (RR: 3.08, 95% CI: 2.24, 4.23; p < 0.001) were each significantly associated with infection. No patient with exposure to eating raw pork had a probable infection. Signs of rash and the presence of anaemia were not significant (Table S1).
TABLE 2.
Comparison of exposure risk factors between eosinophilia responders and non‐responders to empirical anthelminthic treatment.
| Exposure risk factors for helminth infection | Total (N = 131) | Non‐helminth infection (n = 54) | Helminth infection (n = 77) | p |
|---|---|---|---|---|
| Walking barefoot on soil | 11 (8.4) | 1 (1.8) | 10 (13.2) | 0.025 |
| Contact with dirty water | 3 (2.3) | 0 (0.0) | 3 (3.9) | 0.26 |
| Occupation involving contact with soil | 11 (8.4) | 2 (3.6) | 9 (11.8) | 0.12 |
| Recent forest travel | 2 (1.5) | 0 (0.0) | 2 (2.6) | 0.51 |
| Pet (cat or dog) owner | 4 (3.1) | 2 (3.6) | 2 (2.6) | 1.00 |
| Consumption of raw fish | 3 (2.3) | 1 (1.8) | 2 (2.6) | 1.00 |
| Consumption of raw beef | 24 (18.3) | 9 (16.4) | 15 (19.7) | 0.62 |
| Consumption of raw pork | 3 (2.3) | 1 (1.8) | 2 (2.6) | 1.00 |
| Consumption of other raw meat | 3 (2.3) | 1 (1.8) | 2 (2.6) | 1.00 |
| Consumption of raw vegetables | 115 (87.8) | 48 (87.3) | 67 (88.2) | 0.88 |
| Drinking unfiltered water | 5 (3.8) | 2 (3.6) | 3 (3.9) | 1.00 |
Note: Data are presented as n (%). Bold values indicate statistical significance (p < 0.05).
3.5. Non‐Helminth Etiologies
Among the 55 patients with non‐helminth etiologies, allergic disorders were most common (20 of 55, 36.3%), followed by hypereosinophilic syndrome (13 of 55, 23.6%), drug reaction (9 of 55, 16.4%), drug allergy with systemic symptoms (3 of 55, 5.5%), autoimmune disease (4 of 55, 7.3%), clonal (primary) eosinophilic disorders (3 of 55, 5.5%), idiopathic eosinophilia (2 of 55, 3.6%) and exogenous steroid‐related adrenal insufficiency (1 of 55, 1.8%). Eosinophilia secondary to solid neoplastic conditions was excluded in all patients through individualized diagnostic workup as clinically indicated [1, 2].
3.6. Variation in Eosinophil Counts and Percentages
Analysis of AECs across different diagnostic categories revealed distinct patterns (Figure 3). Multiple helminth infections demonstrated the highest median initial AEC (2630/μL), followed by autoimmune disease (1968/μL) and hypereosinophilic syndrome (1598/μL). A distinct pattern was observed within the helminth infection group: multiple infections showed a significantly higher median AEC level than a single infection (median 2630/μL vs. 1016/μL, respectively; p = 0.001) in univariate analysis. Also, the median initial AEC in the patients defined as probable infection was significantly lower than in the multiple helminth infection group (1188/μL vs. 2630/μL, respectively; p = 0.008) in univariate analysis. Allergic disorders demonstrated relatively lower AEC values and narrower distributions than other groups with more than three patients (Figure 3). Idiopathic eosinophilia and exogenous steroid adrenal insufficiency had insufficient cases for meaningful distribution analysis. Despite these differences in median AEC values across diagnostic categories, considerable overlap exists between groups, and AEC alone should not be used as a specific biomarker for clinical decision making in differentiating helminth from non‐helminth etiologies.
FIGURE 3.
Initial absolute eosinophil counts in patients based on the presumptive etiologies of eosinophilia. Box‐plot diagram represents the distribution of initial eosinophil counts in eosinophilia patients (median, IQR). The vertical dashed black lines are the first and third quartiles (784 and 2277 cells//μL, respectively) of the total study population.
The median percentage of eosinophils in the definite helminth infection, probable helminth infection and non‐helminth infection groups was 16.9% (IQR: 11.8%–25.2%, range: 7.4%–58%), 15% (IQR: 12.1%–21.6%, range: 8.9%–55.5%) and 19.6% (IQR: 12.5%–28.9%, range: 5.9%–57.7%), respectively, but was not significantly different (p = 0.51). Analysis of eosinophil counts by quartiles showed a decrease in the percentage of patients with parasites as eosinophil counts increased. The highest proportions of helminth cases were observed in patients with modest eosinophil percentage elevation ≤ 17.2% (Figure 4).
FIGURE 4.
Percentage of patients classified as probable or definite helminth infection by quartiles of eosinophil percentage in the study population.
3.7. Age, Sex and Laboratory Parameter Associations
We conducted a linear regression analysis to examine further the relationship between AEC and demographic factors in the helminth infection group (Table S2). Linear regression within the helminth infection group revealed a non‐significant trend towards lower AEC in females after adjustment. In contrast, a significant age‐related attenuation was observed: patients aged below 45 years had higher initial AEC than those aged 75 years or above on both unadjusted (geometric mean ratio 0.58, 95% CI: 0.35–0.95; p = 0.034) and adjusted analyses (adjusted geometric mean ratio 0.52, 95% CI: 0.31–0.89; p = 0.017), after controlling for multiple infections, diabetes mellitus, hypertension, chronic kidney disease and dyslipidemia. Sex differences in initial AEC remained non‐significant after full adjustment. Patients with drug reaction had significantly lower albumin (mean difference −0.88, 95% CI: −1.65 to −0.11; p = 0.026) and albumin‐to‐globulin ratio (mean difference −0.30, 95% CI: −0.60 to −0.39; p = 0.045) compared with the helminth infection group. No significant laboratory parameter differences were found for allergic disorders, autoimmune disease, hypereosinophilic syndrome, or primary eosinophilic disorders relative to helminth infections. Complete‐case sensitivity analysis yielded consistent results (Tables S3 and S4).
4. Discussion
This prospective cohort study provides one of the first systematic assessments of helminth prevalence and clinical predictors of helminthiasis in patients presenting with eosinophilia to a haematology consultation service at a tropical tertiary centre. Unlike prior studies focused on travellers at travel medicine clinics or refugee populations, this study captures a clinically distinct population in an endemic setting, offering data more applicable to physicians managing eosinophilia in tropical regions. A combined helminth prevalence of 58% was observed, substantially exceeding the 14%–40% reported in travel clinic settings [7, 8, 9, 10, 11] and the identification of leukopenia and hypoalbuminemia as early markers favouring non‐helminth etiologies provides actionable diagnostic guidance not previously described in this context.
Laboratory confirmation was achievable in only 64% of helminth‐classified patients, reflecting the restricted range of available serological assays and the absence of tests for certain tissue parasites including zoonotic filariasis, which is increasingly reported in Thailand [20]. Additionally, both ivermectin and albendazole lack efficacy against trematode and cestode infections such as sparganosis, opisthorchiasis and paragonimiasis, meaning that infections with these organisms may remain undetected and unresponsive to empirical treatment [21]. Albendazole is the treatment of choice for intestinal nematodes, particularly soil‐transmitted helminths prevalent in Thailand, [21, 22] while ivermectin is effective against lymphatic filariasis, strongyloidiasis and gnathostomiasis [21, 23]. Both ivermectin and albendazole have well‐established safety profiles and their administration to patients in whom helminth infection is ultimately not confirmed carries low risk of serious adverse effects at standard therapeutic doses, supporting their use in empirical treatment protocols in endemic settings [21, 22]. The observed median time to AEC normalization of 1 month, with 93% of patients responding within 3 months, provides practical benchmarks for monitoring. Treatment effectiveness should be evaluated no earlier than 1 month after initiation, with 3 months serving as a key decision point, and empirical treatment should be balanced against prompt evaluation for non‐helminth etiologies in patients with organ damage or persistent symptoms.
The 58% combined helminth prevalence warrants careful interpretation, as the probable infection group was defined by AEC normalization following anthelmintic treatment without laboratory confirmation. The definite helminth prevalence of 37.4% alone remains considerably higher than rates from travel clinic settings, and together these findings support empirical anthelmintic therapy as a standard initial approach in tropical tertiary settings with high helminth burden, alongside concurrent evaluation for non‐helminth etiologies.
This study highlights the value of detailed exposure history in identifying helminth infections. Walking barefoot on soil was the only significant univariate risk factor for the combined helminth group, while organism‐specific analyses revealed more granular associations, including gnathostomiasis with dirty water contact, cysticercosis with raw pork consumption and filariasis with forest travel. A structured exposure history can therefore meaningfully guide empirical treatment decisions even when laboratory confirmation is unavailable. It is noted that household or family contact with infected persons was not systematically assessed in this study, which represents a potential gap in exposure characterization; future studies should consider incorporating shared household exposure data to further refine helminth risk profiling.
Multiple helminth infections exhibited distinct clinical features, including higher initial AEC and a trend towards longer treatment response times compared with single infections, consistent with prior observations of polyparasitism [10, 24] Cross‐reactivity among serological assays may lead to overestimation of co‐infection rates, and this possibility should be considered when interpreting multiple positive results [4] Extended follow‐up is warranted in patients with suspected multiple infections to ensure complete AEC normalization.
Age‐related attenuation of eosinophilic response represents an important diagnostic consideration. Elderly patients aged 75 years or above presented with significantly lower initial AEC compared with younger patients, even after adjustment for confounders. This is consistent with known age‐related decline in IL‐5‐mediated eosinophil degranulation and proliferation [25] Clinicians should be aware that reliance on AEC thresholds alone may underestimate helminth infection burden in elderly patients, and that a lower index of suspicion threshold may be appropriate in this group when exposure history is suggestive.
The higher prevalence of leukopenia in non‐helminth etiologies likely reflects immune‐mediated bone marrow suppression, particularly in drug reactions [26]. Hypoalbuminemia in non‐helminth cases, especially drug reactions, may result from decreased protein synthesis, increased catabolism, or protein‐losing states [27]. The concurrent presence of eosinophilia, leukopenia, hypoalbuminemia, and a reduced albumin‐to‐globulin ratio could therefore raise clinical suspicion for drug‐induced aetiology or other systemic disease, rather than helminth infection, potentially enabling earlier targeted intervention.
This study has several limitations. Its single‐centre design at a tertiary referral hospital may limit generalizability, as the study population may over‐represent urban and health‐literate patients compared with the broader at‐risk community in Thailand, suggesting that true population‐level helminth prevalence may be even higher than observed. The small sample sizes for specific non‐helminth etiologies limit definitive conclusions for some laboratory parameter comparisons. The stratified analyses presented in Tables [Link], [Link] should be interpreted in the context of relatively small subgroup sizes, and the observed overlap between groups is acknowledged. Future research priorities should include multicentre prospective cohort studies across diverse tropical settings to validate these findings, development and external validation of a clinical prediction score incorporating leukopenia and hypoalbuminemia for early identification of non‐helminth etiologies, and formal cost‐effectiveness analysis of empirical anthelmintic treatment versus a workup‐first approach in endemic tertiary centres.
5. Conclusion
Helminthiasis was identified in more than half of patients presenting with eosinophilia to a tropical tertiary haematology service, with strongyloidiasis and gnathostomiasis as the most common laboratory‐confirmed infections. Even based on laboratory‐confirmed cases alone, helminth prevalence substantially exceeds rates from travel clinic settings, supporting empirical anthelmintic therapy as a standard initial approach in tropical endemic settings, with 3 months serving as a practical benchmark for treatment response monitoring.
Detailed anamnesis of helminth‐related exposures should be routinely performed, as soil contact, dirty water exposure, and forest travel can guide treatment decisions even without laboratory confirmation. Leukopenia and hypoalbuminemia were identified as accessible markers favouring non‐helminth etiologies and should prompt concurrent investigation.
Funding
This study was funded by the Faculty of Medicine Siriraj Hospital, Mahidol University, Thailand (R016532001).
Conflicts of Interest
The authors declare no conflicts of interest.
Supporting information
Table S1: The risk ratios for helminth exposure risk factors, as well as signs and symptoms associated with helminthiasis, were analysed in the definite and probable helminth infection groups. The risk ratios for specific helminth infections were calculated to assess their association with these factors.
Table S2: Initial AEC stratified by age and sex in the helminth infection group.
Table S3: Comparative analysis of laboratory parameters between the helminth infection and non‐helminth infection groups.
Table S4: Comparative analysis of laboratory parameters between the helminth infection and non‐helminth infection groups by complete case analysis.
Data S1: tmi70191‐sup‐0005‐Supinfo.doc.
Acknowledgements
The authors thank Dr. Anthony Tan for their collaboration in statistical analyses and editing the English language.
Data Availability Statement
The data that supports the findings of this study are available in the Supporting Information of this article.
References
- 1. Butt N. M., Lambert J., Ali S., et al., “Guideline for the Investigation and Management of Eosinophilia,” British Journal of Haematology 176, no. 4 (2017): 553–572. [DOI] [PubMed] [Google Scholar]
- 2. Thomsen G. N., Christoffersen M. N., Lindegaard H. M., et al., “The Multidisciplinary Approach to Eosinophilia,” Frontiers in Oncology 13 (2023): 1193730. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Larsen R. L. and Savage N. M., “How I Investigate Eosinophilia,” International Journal of Laboratory Hematology 41, no. 2 (2019): 153–161. [DOI] [PubMed] [Google Scholar]
- 4. Checkley A. M., Chiodini P. L., Dockrell D. H., et al., “Eosinophilia in Returning Travellers and Migrants From the Tropics: UK Recommendations for Investigation and Initial Management,” Journal of Infection 60, no. 1 (2010): 1–20. [DOI] [PubMed] [Google Scholar]
- 5. O'Connell E. M. and Nutman T. B., “Eosinophilia in Infectious Diseases,” Immunology and Allergy Clinics of North America 35, no. 3 (2015): 493–522. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Roose S., Vande Velde F., Vlaminck J., Geldhof P., and Levecke B., “Serological Diagnosis of Soil‐Transmitted Helminth Infections: A Scoping Review,” PLoS Neglected Tropical Diseases 18, no. 4 (2024): e0012049. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Meltzer E., Percik R., Shatzkes J., Sidi Y., and Schwartz E., “Eosinophilia Among Returning Travelers: A Practical Approach,” American Journal of Tropical Medicine and Hygiene 78, no. 5 (2008): 702–709. [PubMed] [Google Scholar]
- 8. Barrett J., Warrell C. E., Macpherson L., et al., “The Changing Aetiology of Eosinophilia in Migrants and Returning Travellers in the Hospital for Tropical Diseases, London 2002‐2015,” Journal of Infection 75, no. 4 (2017): 301–308. [DOI] [PubMed] [Google Scholar]
- 9. Salzer H. J. F., Rolling T., Vinnemeier C. D., et al., “Helminthic Infections in Returning Travelers and Migrants With Eosinophilia,” Travel Medicine and Infectious Disease 20 (2017): 49–55. [DOI] [PubMed] [Google Scholar]
- 10. Webster J., Stauffer W., Mitchell T., et al., “Cross‐Sectional Assessment of the Association of Eosinophilia With Intestinal Parasitic Infection in U.S.‐Bound Refugees in Thailand,” American Journal of Tropical Medicine and Hygiene 106, no. 5 (2022): 1552–1559. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Posey D. L., Blackburn B. G., Weinberg M., et al., “High Prevalence and Presumptive Treatment of Schistosomiasis and Strongyloidiasis Among African Refugees,” Clinical Infectious Diseases 45, no. 10 (2007): 1310–1315. [DOI] [PubMed] [Google Scholar]
- 12. Nopparatana C., Setasuban P., Chaicumpa W., and Tapchaisri P., “Purification of Gnathostoma Spinigerum Specific Antigen and Immunodiagnosis of Human Gnathostomiasis,” International Journal for Parasitology 21, no. 6 (1991): 677–687. [DOI] [PubMed] [Google Scholar]
- 13. Eamsobhana P., Yong H. S., Mak J. W., et al., “Detection of Circulating Antigens of Parastrongylus Cantonensis in Human Sera by Dot‐Blot ELISA and Sandwich ELISA Using Monoclonal Antibody,” Southeast Asian Journal of Tropical Medicine and Public Health 28, no. 3 (1997): 624–628. [PubMed] [Google Scholar]
- 14. Eamsobhana P., Yoolek A., Suvouttho S., et al., “Purification of a Specific Immunodiagnostic Parastrongylus Cantonensis Antigen by Electroelution From SDS‐Polyacrylamide Gels,” Southeast Asian Journal of Tropical Medicine and Public Health 32, no. 2 (2001): 308–313. [PubMed] [Google Scholar]
- 15. Wongkamchai S., Monkong N., Mahannol P., Taweethavonsawat P., Loymak S., and Foongladda S., “Rapid Detection and Identification of Brugia Malayi, B. Pahangi, and Dirofilaria Immitis by High‐Resolution Melting Assay,” Vector‐Borne and Zoonotic Diseases 13, no. 1 (2013): 31–36. [DOI] [PubMed] [Google Scholar]
- 16. Wongkamchai S., Satimai W., Loymek S., Nochot H., and Boitano J. J., “An ELISA Kit With Two Detection Modes for the Diagnosis of Lymphatic Filariasis,” Journal of Helminthology 89, no. 5 (2015): 552–558. [DOI] [PubMed] [Google Scholar]
- 17. Shomali W. and Gotlib J., “World Health Organization and International Consensus Classification of Eosinophilic Disorders: 2024 Update on Diagnosis, Risk Stratification, and Management,” American Journal of Hematology 99, no. 5 (2024): 946–968. [DOI] [PubMed] [Google Scholar]
- 18. Pongdee T., Manemann S. M., Decker P. A., et al., “Rethinking Blood Eosinophil Counts: Epidemiology, Associated Chronic Diseases, and Increased Risks of Cardiovascular Disease,” Journal of Allergy and Clinical Immunology: Global 1, no. 4 (2022): 233–240. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. van Buuren S., Boshuizen H. C., and Knook D. L., “Multiple Imputation of Missing Blood Pressure Covariates in Survival Analysis,” Statistics in Medicine 18, no. 6 (1999): 681–694. [DOI] [PubMed] [Google Scholar]
- 20. Sarasombath P. T., Sitthinamsuwan P., Wijit S., et al., “Integrated Histological and Molecular Analysis of Filarial Species and Associated Wolbachia Endosymbionts in Human Filariasis Cases Presenting Atypically in Thailand,” American Journal of Tropical Medicine and Hygiene 111, no. 4 (2024): 829–840. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Kappagoda S., Singh U., and Blackburn B. G., “Antiparasitic Therapy,” Mayo Clinic Proceedings 86, no. 6 (2011): 561–583. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Adegnika A. A., Zinsou J. F., Issifou S., et al., “Randomized, Controlled, Assessor‐Blind Clinical Trial to Assess the Efficacy of Single‐ Versus Repeated‐Dose Albendazole to Treat Ascaris Lumbricoides, Trichuris Trichiura, and Hookworm Infection,” Antimicrobial Agents and Chemotherapy 58, no. 5 (2014): 2535–2540. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Nogrado K., Adisakwattana P., and Reamtong O., “Human Gnathostomiasis: A Review on the Biology of the Parasite With Special Reference on the Current Therapeutic Management,” Food and Waterborne Parasitology 33 (2023): e00207. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Heukelbach J., Poggensee G., Winter B., Wilcke T., Kerr‐Pontes L. R. S., and Feldmeier H., “Leukocytosis and Blood Eosinophilia in a Polyparasitised Population in North‐Eastern Brazil,” Transactions of the Royal Society of Tropical Medicine and Hygiene 100, no. 1 (2006): 32–40. [DOI] [PubMed] [Google Scholar]
- 25. Mathur S. K., Schwantes E. A., Jarjour N. N., and Busse W. W., “Age‐Related Changes in Eosinophil Function in Human Subjects,” Chest 133, no. 2 (2008): 412–419. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Curtis B. R., “Non‐Chemotherapy Drug‐Induced Neutropenia: Key Points to Manage the Challenges,” Hematology. American Society of Hematology. Education Program 2017, no. 1 (2017): 187–193. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Soeters P. B., Wolfe R. R., and Shenkin A., “Hypoalbuminemia: Pathogenesis and Clinical Significance,” JPEN Journal of Parenteral and Enteral Nutrition 43, no. 2 (2019): 181–193. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1: The risk ratios for helminth exposure risk factors, as well as signs and symptoms associated with helminthiasis, were analysed in the definite and probable helminth infection groups. The risk ratios for specific helminth infections were calculated to assess their association with these factors.
Table S2: Initial AEC stratified by age and sex in the helminth infection group.
Table S3: Comparative analysis of laboratory parameters between the helminth infection and non‐helminth infection groups.
Table S4: Comparative analysis of laboratory parameters between the helminth infection and non‐helminth infection groups by complete case analysis.
Data S1: tmi70191‐sup‐0005‐Supinfo.doc.
Data Availability Statement
The data that supports the findings of this study are available in the Supporting Information of this article.