The prevalence of soil transmitted helminth infections in minority indigenous populations of South-East Asia and the Western Pacific Region: A systematic review and meta-analysis

Introduction Soil transmitted helminth (STH) infections cause one of the most prevalent diseases in man. STHs disproportionately impact socio-economically disadvantaged communities including minority indigenous populations. This systematic review aimed to quantify the prevalence of STH infection within minority indigenous populations of the South-East Asia and Western Pacific Regions. Methods The systematic review was conducted in accordance with The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines following a published protocol. A random effects meta-analysis was used to estimate the pooled prevalence of STH infection, and meta-regression analysis was used to quantify associations with study characteristics. Where comparative data were available, sub-group analysis was conducted to evaluate the risk of STH infection in minority indigenous people relative to other population groups. The heterogeneity between studies was evaluated visually using Forest plots and was assessed quantitatively by the index of heterogeneity (I2) and Cochran Q-statistics. Results From 1,366 unique studies that were identified, 81 were included in the final analysis. The pooled prevalence of infection within minority indigenous populations was 61.4% (95% CI 50.8, 71.4) for overall STH infection; 32.3% (95% CI 25.7, 39.3) for Ascaris.lumbricoides; 43.6% (95% CI 32.6, 54.8) for Trichuris.trichiura; 19.9% (95% CI 15.7, 24.5) for hookworm and 6.3% (95% CI 3.2, 10.2) for Strongyloides.stercoralis. A significant increase in T. trichiura prevalence was observed over time. The stratified analysis showed that the prevalence of infection for STH overall and for each STH species were not significantly different in minority indigenous participants compared to other populations groups. Conclusion The prevalence of STH infection is high within minority indigenous populations across countries at very different levels of socio-economic development. The increasing prevalence of T. trichiura calls for the implementation of more effective therapies and control strategies.


Introduction
Soil transmitted helminthiasis is a Neglected Tropical Disease (NTD) [1] estimated to impact 1.5 billion people, [2] a figure which equates to 19% of the world's population. The four species of gastro-intestinal nematode commonly included in soil transmitted helminths (STH) are Ascaris lumbricoides (roundworms), Trichuris trichiura (whipworms), Necator americanus and Ancylostoma duodenale (hookworms). Ancylostoma ceylanicum is also an increasingly recognized hookworm species of public health importance. These parasites prevail in the tropics and subtropics and have their greatest impact on populations affected by poverty and disadvantage. [2][3][4][5] The impact STH infection creates a significant global health burden. In 2016, the WHO estimated a loss of 3.4 million disability adjusted life-years (DALY) worldwide, of which 42% was attributed to A. lumbricoides, 10% to T. trichiura and 48% to hookworm infection. [6] A significant proportion of the total disease burden is attributed to Years Lost due to Disability (YLD) which is estimated at 2.9 million. [6] The quantum of the YLD estimate is reflective of the chronic and debilitating morbidity associated with STH infections. The symptoms of morbidity are often difficult to quantify due to the effects of poverty, malnutrition and co-infection, which are common amongst those worst affected. [7] However, a number of morbidities have been well documented and include impaired growth and physical development, intestinal obstruction, anaemia, vitamin A deficiency, and poor intellectual and cognitive development. [4,8] Although not as well represented in the literature, Strongyloides stercoralis is another pathogenic STH of significance to human health. While the prevalence of Strongyloidiasis is difficult to quantify due to many cases being asymptomatic and traditional diagnostic methods lacking sensitivity, [9] global estimates project between 100 and 370 million infections.[9,10] S. stercoralis is differentiated from other STH species by an auto-infective capability within the lifecycle [11] and by its prevalence in both tropical and temperate climates. [12] Statistics for S. stercoralis are not included within DALY figures and although hyper-infection syndrome for this parasite is rare, it is often fatal in immunocompromised patients among whom mortality rates of 86% are reported. [13] The successful control of STH infections will be dependent upon a multi-faceted approach. Economic development is proven to be a significant factor in eradicating STH infections [14] and it is acknowledged that WASH (water, sanitation and hygiene) and education initiatives are fundamental to reducing disease transmission. [15] These approaches are combined with the primary focus of the WHO endorsed control strategy for A. lumbricoides, T. trichiura and hookworm infection which is the periodic administration of anthelmintic drugs to at-risk populations living in endemic areas. [16] Although well-developed treatment strategies have been developed for A. lumbricoides, T. trichiura and hookworm, [15] systematic action plans to address Strongyloidiasis are lacking. [10] There is a fundamental lack of epidemiological data for Strongyloides infection, a knowledge gap not limited to developing regions as evidenced by the call for its inclusion on the Australian Notifiable Disease List. [17] Although significant reductions in STH prevalence have been achieved over recent times, [18] infections continue to impose a significant global health burden and impact those most vulnerable within society. One population group that has been shown to be disproportionately affected by poverty and social disadvantage is indigenous people. [19] Although there are published studies on the impact of STH infections within discrete ethnic groups, there is nothing in the literature that quantifies STH infection risk in minority indigenous people as a collective. If the goals of the 2030 Agenda for Sustainable Development are to be achieved, the burden of disease amongst vulnerable populations needs to be evaluated to inform effective interventions.
This systematic review aimed to quantify the prevalence of STH infection amongst minority indigenous populations of the SEAR and WPR. These regions were chosen as WHO data attributes a high proportion of DALYs to be lost as a result of STH infection within these areas. [6] The SEAR and WPR also include a significant representation of indigenous populations [20] whilst providing an opportunity to compare the prevalence of STH infection across countries of differing socio-economic strata.

Search strategy
The systematic review was undertaken in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [21] (S1 PRISMA checklist). Specifics on the search criteria and details of the study selection criteria are available in a published protocol. [22] In summary, four biomedical databases: Scopus, Web of Science, Medline (Ovid) and EMBASE (Ovid), were systematically searched using the criteria detailed in S1 Table, without restriction on the year of publication. In addition to the biomedical database search, reference lists from included publications were hand searched.

Study screening and selection criteria
All studies identified from the systematic search were imported into Endnote X9 (Clarivate Analytics) where duplicates were deleted. Following removal of the duplicates, studies were uploaded to Rayyan Qatar Computing Research Institute (QCRI) [23] and titles and abstracts were independently assessed by two authors (BG and KAA). The full text articles of shortlisted abstracts were independently screened by the same two authors against the inclusion and exclusion criteria.
Any discrepancies relating to the shortlisting of publications were discussed and where consensus could not be achieved, advice was sought from the third author (ACAC). Where further clarification was required, this was requested from the corresponding author of the relevant publication.

Inclusion criteria
Studies were included if they were representative cross-sectional surveys relating to human infection and provided sufficient data to facilitate the calculation of STH prevalence. Studies were required to include minority indigenous population participants within the SEAR or WPR.
In accordance with the protocol, [22] minority indigenous populations were defined when each of the following criteria were met: ■ Descendants of the original or earliest known inhabitants of an area; people who have historical continuity with pre-invasion and pre-colonial societies.
■ Distinct societies with languages, culture, customs, and social and political frameworks that vary significantly from those of the dominant population.
■ Groups of people with strong cultural ties and dependence upon the environment and its resources for their survival.
■ People self-identifying as indigenous.
■ Groups who face relative disadvantage or discrimination in multiple areas of social existence-success, education, healthcare, employment.
■ Numerically non-dominant groups in a country or area.
The WHO Global Burden of Disease (GBD) regional classification system [24] was used to define the countries located within the SEAR and WPR.

Exclusion criteria
Studies were excluded if they were not full text articles and did not publish in English. Publications were excluded if less than 90% of the participants (or, for the comparative studies, the minority indigenous category) met the minority indigenous population criteria. Data from case series with less than 10 participants and case studies; systematic and literature reviews; conference poster or abstracts and scientific correspondence e.g., letters to the editor, were excluded. Singapore was excluded from the search as it does not have any minority indigenous people according to the definitions used by this review.

Outcomes
The primary outcome of the study was prevalence of STH infection amongst minority indigenous populations of the SEAR and WPRs. Prevalence included STH infection overall and according to species: A.lumbricoides, T.trichiura, S.stercoralis and Hookworm species collectively.

Data extraction and quality assessment
Data were extracted from included studies using Microsoft Excel version 2016 (Microsoft, Redmond, Washington, USA) by BG and independently validated by KAA. Following pilot testing and refinement, a data extraction spreadsheet was used to record the following information: first author and year of publication; year and country in which the study was undertaken; study population classification (minority indigenous or other); species of infectious agent; diagnostic method; sex of study participants; size (n) of the study population and number of disease positive participants. Although the protocol [22] also intended to extract and analyze data by age, this was not undertaken due to the large variation in age classifications across publications.
Where studies evaluated the impact of intervention regimes, only pre-intervention baseline data were extracted. When surveys undertook a comparison of disease prevalence across minority indigenous and other population groups, data were extracted for both to facilitate a comparison.
A modified version of the Newcastle-Ottawa Quality Assessment (QA) Scale [25] was utilized to assess the quality of the studies analysed, the scores for which are detailed in S2 Table. The QA tool has scores ranging from 0 to 9, in accordance with the protocol, [22] scores between 1 and 4 were defined as low quality, scores between 5 and 7 were defined as medium quality, and scores between 8 and 9 defined as high quality.

Study variables
The mortality strata for each country of study was attributed according to the WHO definitions [26], and was evaluated as a study variable. The other study variables used for the subgroup analysis included: WHO region, country of study, year of data collection, study location (community/school), number of samples analysed (singular/multiple), diagnostic method, number of helminth infections, study participant sex, helminth species (for hookworm) and QA grade.

Data analysis
For the studies that identified overall STH infection, and for data extracted by species (A.lumbricoides, T.trichiura, hookworm and S.stercoralis), a random-effects meta-analysis was used to estimate the pooled prevalence of infection. The meta-analysis was undertaken using the Freeman-Tukey double arcsine transformation to address confidence limits outside the 0 to 1 range and variance instability. [27] This was implemented in Stata using the metaprop command. [28] The heterogeneity between studies in minority indigenous populations was assessed using Cochran's Q test and was quantitatively evaluated with the index of heterogeneity squared (I 2 ) statistic with 95% CI. [29] Heterogeneity between studies was classified low, moderate and high when I 2 values were below 25%, between 25% and 75% and above 75%, respectively. [29] In an attempt to account for the high heterogeneity that was identified, meta-regression was undertaken using the study characteristics as covariates. The meta-regression was conducted using the robust variance estimation (RVE) method to manage non-independent effect sizes without knowledge of the within-study covariance structure. [30] Where comparative data were available, sub-group analysis was conducted to evaluate the risk of helminth infection in minority indigenous communities relative to other population groups. Where differences in infection prevalence were identified across study variables, or between population groups, bivariate meta-regression was used to evaluate their significance (p-value <0.05) when three or more data sets were available for each comparison.
Funnel plots were utilized to evaluate potential publication bias and asymmetry was assessed using Egger's method with a p-value <0.05 denoting significant bias. [31] Analysis was conducted using Stata/MP version 16.1 (StataCorp, College Station, TX).

Results
The search identified 1,366 unique studies from which 157 were shortlisted following title and abstract screening. Following the full text review, 81 studies were included in the final analysis (Fig 1); the characteristics of the studies are provided in Table 1. Publication bias of the included studies was evidenced by the asymmetrical shape of the funnel plot (Fig 2) and a p value = 0.025 calculated with Egger's regression test.

Prevalence of overall STH Infection
Out of the 81 studies, 49 enabled the overall prevalence of STH infection to be calculated. Details on the pooled prevalence of infection and bivariate meta-regression across the study covariates are detailed in Tables 2 and 3, respectively.
The pooled prevalence of STH infection across the 49 studies, which represented 15,238 minority indigenous participants, was 61.4% (95% CI 50.8, 71.4), with high (I 2 = 99.4%) and significant (p = 0.000) heterogeneity shown between studies (Fig 3). Eighty-six percent of the studies were undertaken in the WPR and 61% of the studies that reported overall STH prevalence, had been undertaken within Malaysia. The prevalence of infection was found to be significantly higher in the WPR at 66.3% (95% CI 55.2, 76.6) compared to the SEAR at 30.3% (95% CI 15.6, 47.3; p = 0.010). The only other study covariate found to have a significant effect on overall STH prevalence, was the use of serology as a diagnostic method relative to microscopy (p = 0.000). Where studies detailed the number of infections, the prevalence of single and multiple species infections were found to be comparable.
Five studies provided data that could be used to compare STH infection prevalence between minority indigenous and other population groups. Although the prevalence of infection was higher in minority indigenous populations (41.9%, 95% CI 15.6, 70.9) relative to other groups (37.5%, 95% CI 10.6, 69.5) this was not found to be significant (p = 0.870).

Prevalence of Ascaris lumbricoides infection
Out of the 81 studies, 64 reported on the prevalence of A. lumbricoides infection. Details on the pooled prevalence of infection and bivariate meta-regression across the study covariates are detailed in Tables 4 and 5, respectively.
The pooled prevalence of A. lumbricoides infection across the 64 studies, representing 21,495 minority indigenous participants, was 32.3% (95% CI 25.7, 39.3- Fig 4). Although there was significant heterogeneity between publications, the only study covariates of significance were WHO region and country. The WPR, where 70% of the studies were undertaken, had a significantly higher prevalence of infection at 39.8% (95% CI 31.9, 47.9) than the SEAR at 16.5% (95% CI 8.22, 26.8; p = 0.002). Where sufficient data were available to allow the country of study to be analyzed as a covariate, prevalence was found to be significantly higher in China (67.8%, 95% CI 39.0, 90.7; p = 0.002) and Malaysia (38.3%, 95% CI 31.8, 44.9; p = 0.022) than elsewhere.
Eight studies provided data that facilitated a comparison of A.lumbricoides infection prevalence between minority indigenous and other population groups. Although not significant (p = 0.860), the prevalence of infection was found to be higher in minority indigenous participants (41.0%, 95% CI 25.7, 57.2) compared to those from other population groups (25.2%, 95% CI 8.4, 47.2).    Heterogeneity was found to be high (I 2 >75%) for A.lumbricoides prevalence within all covariates, with the exception of QA grade. Studies classified with a medium QA score (5-7) showed moderate heterogeneity (I 2 25-75%).

Prevalence of Trichuris trichiura infection
Sixty -five of the 81 studies reported on the prevalence of T. trichiura infection, representing a cumulative study population of 20,466 minority indigenous participants. The pooled prevalence of infection across the study covariates and the subsequent bivariate meta-regression are detailed in Tables 6 and 7, respectively. The pooled prevalence of T. trichiura infection within minority indigenous populations was 43.6% (95% CI 32.6, 54.8- Fig 5). There was significant heterogeneity between studies, with WHO region, country of study, period of data collection, and QA grade shown to be significant study co-variates. The prevalence of infection was shown to be significantly higher in the WPR at 55.8% (95% CI 44.2, 67.1) compared to the SEAR at 10.3% (95% CI 5.2, 16.9; p = 0.000). Where sufficient data were available to evaluate the country of study as a covariate, infection prevalence was significantly higher in China (51.6%, 95% CI 13.0, 89.1; p = 0.018) and Malaysia (67.8%, 95% CI 56.7, 78.0; p = 0.000). T. trichiura infection was found to be significantly higher in 2001-2020 (52.1%, 95% CI 37.0, 67.0) compared to 1949-1980 (17.6%, 95% CI 4.0, 37.9; p = 0.000). High QA grade studies were shown to have a significantly higher prevalence of T. trichiura infection (91.6%, 95% CI 71.62, 99.99) than low QA grade studies (25.1%, 95% CI 11.56, 41.74).
Eight studies reported data that facilitated a comparison of infection prevalence between minority indigenous and other population groups. Although the differential in T. trichiura prevalence was not significant (p = 0.115), it was higher in minority indigenous study participants (42.5%, 95% CI 26.9, 58.9) in comparison to those from other population groups (24.6%, 95% CI 15.5, 35.1).

Prevalence of hookworm infection
Sixty-eight studies presented data on hookworm infection, representing a cumulative minority indigenous study population of 21,967 participants. The pooled prevalence of infection across the study covariates and the subsequent bivariate meta-regression are detailed in Tables 8 and  9, respectively. The pooled prevalence of hookworm infection was 19.9% (95% CI 15.7, 24.5) within minority indigenous populations (Fig 6). The heterogeneity between studies was found to be high (I 2 = 98.5%) and significant (p = 0.000). The country of study was found to be the only significant study covariate and although there were insufficient data to evaluate all countries represented, four countries were found to have a significantly higher prevalence of infection than other countries. These countries were: China (49.8%, 95% CI 20.8, 78.9; p = 0.009), India (20.4%, 95% CI 12.7, 29.3; p = 0.010), Malaysia (17.2%, 95% CI 13.3, 21.5; p = 0.001)) and the Philippines (16.0%, 95% CI 11.2, 21.4; p = 0.014).
Eight studies detailed the species of hookworm they identified. Based on these publications, N. americanus was more prevalent (44.9%, 95% CI 23.8, 67.0) than A. duodenale (11.6%, 95% CI 1.3, 29.7) and A.ceylanicum was reported in one study only. [32] In addition to these eight studies, Eight studies presented data that enabled a comparison of hookworm infection prevalence to be evaluated between minority indigenous and other populations. Although the difference between population groups was not found to be significant (p = 0.597), it was higher in minority indigenous participants (16.7%, 95% CI 3.9, 35.7) than those from other population groups (10.7%, 95% CI 1.6, 26.3).

Prevalence of Strongyloides stercoralis infection
Twenty studies over a cumulative 7,020 minority indigenous participants reported on the prevalence of S.stercorlais infection. The prevalence of infection analyzed by study co-variates is detailed in Table 10 and the subsequent meta-analysis in Table 11. The pooled prevalence of infection within minority indigenous populations was 6.3% (95% CI 3.2, 10.2) with a high and significant degree of heterogeneity between studies (Fig 7). From the study co-variates analyzed, diagnostic method and sex where the only two covariates to There was only one study that provided data enabling a comparison of S. stercorlais prevalence between minority indigenous and other population participants. Although it was not possible to evaluate the significance of the results, it is noted that prevalence was higher in minority indigenous participants (13.6%, 95% CI 11.2, 16.4) compared to those in other population groups (5.1%, 95% CI 2.8, 9.1).

Discussion
The systematic review shows a high prevalence of STH infection amongst minority indigenous populations. It is likely that the true prevalence of infection is higher due to the low sensitivity of diagnostic methods used. [36] This potential under-estimation of infection is particularly likely in minority indigenous communities, for whom the provision of faecal samples presents a significant obstacle due to cultural beliefs, thereby creating a challenge to the recommended serial sampling over multiple days. [37,38] The results from our review show the prevalence of infection to be consistently higher in the WPR than the SEAR, although WHO figures show the DALYs to be higher overall within the SEAR. [6] Although there are many potential confounders, the higher prevalence of  infection within the WPR identified by this review may reflect a higher burden of disease within indigenous minority populations in this region. Although research shows the prevalence and intensity of STH infection to be related to socioeconomic status and hygiene conditions, [39][40][41][42][43] it is interesting to note that the review found no significant difference in disease prevalence for some STH between countries that have very different socio-economic profiles. For example, the review shows there to be no significant difference in overall STH infection in minority indigenous populations between Australia, which in 2020 ranked eighth on the Human Development Index (HDI), and India which ranked 131 st . [44] This re-enforces the fact that vulnerable population groups within otherwise highly developed countries continue to be at risk of NTDs such as STH infection. Although it is hoped that economic development and preventative chemotherapy programs have led to a reduction in the global burden of STH infection over time, [18] results from the systematic review show the prevalence of overall STH infections within minority indigenous populations to have remained static. When the review analyzes the prevalence of infection by species, some interesting trends are observed. In particular, the prevalence. of S. stercorlais and T. trichiura infections within minority indigenous populations have increased over time, with the increasing prevalence of T. trichiura being significant. The increasing trend in S. stercoralis prevalence may in part be due to developments in diagnostic capabilities as the parasite is very difficult to detect by microscopy; [45] but may also reflect the treatment challenges presented by its autoinfection capability. [46] The significant increase in T.trichiura infection within this  vulnerable population group however warrants further investigation. Although the WHO recommend the administration of albendazole or mebendazole as part of their STH control strategy, [2] these drugs are shown to have limited efficacy against T. trichiura. [47,48] Although the review provides an indication of STH prevalence within indigenous minority populations as a collective, research showing the significant heterogeneity in infection prevalence and intensity between individuals within a population is noted. [36] There is an argument that infection intensity would be a more useful metric than prevalence, as morbidity severity is relative to infection intensity and heavily infected individuals present a major source of infection for their community. [36] If the 2021-2030 NTD road map targets [49] are to be achieved, countries need to address the impact of STH infections within their vulnerable indigenous populations. By impacting productivity and human development, STH infections re-enforce poverty, [50] which already disproportionately affects these communities [19]. To be effective, interventions need to be culturally appropriate [51] and as a result of disruptions to public health programmes caused by the Coronavirus Disease 2019 (COVID 19) pandemic, they will need to be increasingly innovative if 2021-2030 targets are to be achieved.[52] This systematic review provided information on STH prevalence amongst minority indigenous populations of the SEAR and WPRs and showed where further data and research are required. However, the limitations of systematic reviews and the scope of data need to be taken into consideration when results of the systematic review are used to inform public health policy. The following limitations of the review are noted. Publication bias is an inherent potential limitation of the systematic review process. As a result of resource constraints data extraction was limited to articles published in English. The accuracy of estimating disease prevalence may be impacted by the inclusion of small study populations. The review did not take into consideration the effect of treatment and intervention regimes which may impact infection prevalence over time. The definition of a minority indigenous population is not based upon a universal classification.

Conclusion
STH infections continue to create a significant global health burden within vulnerable communities. Soil transmitted helminthiasis is prevalent within indigenous communities who reside in countries across the spectrum of WHO mortality strata. To stop the ongoing impacts of STH infection upon the poverty cycle, accurate relevant prevalence and infection intensity data are required to inform innovative and culturally appropriate interventions.