Spatiotemporal Distribution of Dengue and Chikungunya in the Hindu Kush Himalayan Region: A Systematic Review

The risk of increasing dengue (DEN) and chikungunya (CHIK) epidemics impacts 240 million people, health systems, and the economy in the Hindu Kush Himalayan (HKH) region. The aim of this systematic review is to monitor trends in the distribution and spread of DEN/CHIK over time and geographically for future reliable vector and disease control in the HKH region. We conducted a systematic review of the literature on the spatiotemporal distribution of DEN/CHIK in HKH published up to 23 January 2020, following Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines. In total, we found 61 articles that focused on the spatial and temporal distribution of 72,715 DEN and 2334 CHIK cases in the HKH region from 1951 to 2020. DEN incidence occurs in seven HKH countries, i.e., India, Nepal, Bhutan, Pakistan, Bangladesh, Afghanistan, and Myanmar, and CHIK occurs in four HKH countries, i.e., India, Nepal, Bhutan, and Myanmar, out of eight HKH countries. DEN is highly seasonal and starts with the onset of the monsoon (July in India and June in Nepal) and with the onset of spring (May in Bhutan) and peaks in the postmonsoon season (September to November). This current trend of increasing numbers of both diseases in many countries of the HKH region requires coordination of response efforts to prevent and control the future expansion of those vector-borne diseases to nonendemic areas, across national borders.


Introduction
Dengue (DEN) is one of the fastest spreading infectious human diseases of the twenty-first century, and chikungunya (CHIK) is an emerging public health threat worldwide [1]. DEN is caused by the dengue virus (DENV), which is distinguished in 4 serotypes, DENV-1 to -4 and CHIK, by the chikungunya virus (CHIKV) [2,3]. According to estimates of the World Health Organization (WHO), around 100 million DEN infections occur worldwide annually, and approximately 2.5 billion of the world's population live in DEN-endemic areas [4]. Thus, it has a major socioeconomic and public health impact on the epidemic

Methods
A systematic literature review was performed in order to summarize information on the spatiotemporal distribution of DEN/CHIK in the HKH region, following the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines. All Web of Sciences (WoS) databases (Web of Science Core Collection, Biological Abstracts, BIOSIS Citation Index, Current Contents Connect, Data Citation Index, Derwent Innovations Index, KCI-Korean Journal Database, Medline, Russian Science Citation Index, SciELO Citation Index, and Zoological Record) were searched for peer-reviewed articles in the English language published up to 23 January 2020. The following search terms were applied in title, abstract, and keywords (topic search) in order to generate the database with all relevant articles related to our topic listed in WoS (full search term available in Table S15) •

VBDs (DEN, CHIK) and their synonyms and related insect vectors [AND]. •
Names of the countries in the HKH region or names of territories of countries, as well as river and mountain areas in the HKH region (given by the International Centre for Integrated Mountain Development (ICIMOD) [16]), and their synonyms were added.
Inclusion criteria for selecting studies to our final database were 1.
Epidemiological studies dealing with both spatial and temporal distribution of DENV or CHIK.

2.
Studies conducted in the HKH region countries (as defined by the International Centre for Integrated Mountain Development (ICIMOD) [16]).
All articles not matching the inclusion criteria and other non-peer-reviewed articles were excluded (Table S3). For final eligibility, all selected and rejected articles were verified by a second person. Subsequently, the final database was reported ( Figure 1). All included articles were downloaded and analyzed based on the objective of this study. All articles not matching the inclusion criteria and other non-peer-reviewed articles were excluded (Table S3). For final eligibility, all selected and rejected articles were verified by a second person. Subsequently, the final database was reported ( Figure 1). All included articles were downloaded and analyzed based on the objective of this study.

Spatiotemporal Distribution of DEN and CHIK
The final database (Table S2)was analyzed by the publication year and the countries of origin to build up time-and country-specific bibliometric figures. For the systematic review, the database was sorted by DEN, CHIK, and their vectors (Table S4): 1.
DEN/CHIK and Aedes  Table 2. Reported period of outbreak/cases and reported peak month of outbreak per country. Included countries for analysis are India, Nepal, and Bhutan. Pakistan, Afghanistan, Myanmar, and Bangladesh were excluded because data for analysis was not present for these countries. In total, 66 articles were qualitatively analyzed by the year of publication, the study location, the study period, the disease reported month, and the peak month for disease outbreak (Tables 1 and 2). Among those 66 articles, 61 articles giving the exact number of DEN/CHIK cases (Table S5) were quantitatively analyzed for the number of reported and confirmed DENV/CHIK cases and, finally, the spatiotemporal distribution of DEN and CHIK observations in HKH from 1951 to 2020.

Bibliometric Description of Database
As shown in the PRISMA flow diagram (Figure 1), the screening and eligibility check of 490 initially searched records (Table S1) resulted in a database with 66 original articles (Table S2), that fully met the inclusion criteria. Out of these 66 selected articles, 83.33% dealt with the investigation of DEN (n = 55), 10.61% with CHIK (n = 7), and 6.06% with the investigation of both DEN and CHIK diseases (n = 4). As shown in Table 1 (Table S14).

Number of Reported DEN and CHIK Cases in the HKH Region
In total, 72,715 DEN cases were reported from all HKH countries from 1995 to 2019, from which 52.50% cases were clinically confirmed (Table S5). For CHIK, 2334 suspected cases were reported throughout the HKH region, of which 14.40% were clinically confirmed (Table S5). The clinical confirmation of both DEN and CHIK was done by serological tests using enzyme-linked immune-sorbent assays (ELISAs) to capture for nonstructural protein 1 antigen (NS1), immunoglobulin M (IgM), and immunoglobulin G (IgG) antibodies.
The highest number of clinically confirmed DEN cases within the HKH region was reported from Nepal (56.83%) (Table S11),followed by India (41.91%) (Table S10), whereas a low number of DEN cases within the HKH region was reported from Bhutan (0.58%) (Table S7), followed by the HKH countries of Pakistan (0.36%) (Table S6), Bangladesh (0.26%) (Table S9), and Afghanistan (0.06%) (Table S8). Two articles dealt with the coinfection of DEN and other diseases: one article reported a coinfection of DEN with scrub typhus [33], and another article reported the coinfection of DEN, malaria, and scrub typhus [36].

Spatiotemporal Distribution of DEN in HKH Countries
In Nepal, DEN was first reported in a Japanese volunteer working in Nepal in the year 2004 [69]. Local transmission of DEN was confirmed in the lowland areas of 11 districts (Bardiya, Banke, Dang, Salyan, Sindhuli, Birjung, Parsa, Rupandehi, Jhapa, Kapilbastu, Dhading) with a circulation of all four serotypes during the first DEN outbreak in August to November 2006 [75,76]. Afterward, DEN cases were reported from the western lowland region of Nepal in 2007 [61]. In 2008 and 2009, DEN was expanded geographically to western, far-western, and central Nepal [59]. DEN outbreaks were continuously reported from 2006 onwards and spread to 32 districts located in the lowlands, hills, and highland regions of Nepal out of Nepal's total 75 districts [19,23]. A large DEN outbreak was reported in 2019 from 68 districts of Nepal, with more than 10,000 reported cases and six deaths [15]. Similar to Nepal, DEN in Bhutan was reported not earlier than 2004, and then again in 2005, 2006 [63], 2013, and 2014 [45].

Spatiotemporal Distribution of CHIK in HKH Countries
A study conducted in the HKH region of India from 2014 to 2017 [17] shows that CHIK infection occurs throughout the whole year. Other studies from the HKH region of India showed that CHIK disease peaks in September, October, and November [17,27,46]. In Nepal, CHIK cases were reported from August to November [44] and in March, May, and June [49]. However, no study from Nepal mentioned a peak month for CHIK infection. It has to be noted that CHIK cases are reported from the hilly and lowland regions of Nepal [30,44,49]. In Bhutan, CHIK cases were reported to occur in July [52], whereas, in Myanmar, no peak season of CHIK occurrence was documented [73].
month of June to December from Nepal [47,49,50,59,60,62,63,65]. In Bhutan, DEN was reported to start in May [45]. Most DEN cases in India and Nepal were found in September, October, or November [28]. However, one article from India [57] showed that the DEN outbreak peaks in December. Other literature from Nepal reports August as DEN peak month [34]. From Bhutan, July and September were reported as peak months for DEN outbreaks [45]. Deducted from this sparse information, it appears that the DEN outbreak period and the DEN peak month for DEN outbreaks vary only slightly within the HKH region. The articles related to the HKH countries of Pakistan, Afghanistan, Myanmar, and Bangladesh do not mention DEN outbreak periods and peak month for the disease outbreak.

Discussion
The present study determines the spatial and temporal distribution of 72,715 DEN and 2334 CHIK cases in the HKH region, as reported in the scientific literature from 1952 to 2020. The studies included in our database reveal that DEN occurs in HKH areas of at least seven out of eight HKH countries and CHIK in at least four out of eight HKH countries. DEN fever emerged in the HKH region in 1951, thus, twenty years earlier than CHIK fever. An increase of reported DEN and CHIK cases in the HKH region was observed from 2004 onwards.
In the HKH region, DENV was first recognized in tea gardens of Northern Assam, India, in 1951 [74]. According to the literature, a large number of states (eight states and two provinces) in the Himalayan and sub-Himalayan region of India have been affected by DEN after 2005 onwards [18,20,21,24,25,27,28,33,36,39,40,42,50,51,54,56,57,60,62,65,71,72,74] The virus circulating in India in the 1950s, causing mild diseases, was replaced or evolved into genotypes with bigger virulence and transmissibility [82]. The movement of the human population from DEN-endemic areas of India to nonendemic areas might have significantly contributed to the outbreak of DEN in new areas and more frequent DEN epidemics after 2005 [57].
The DEN cases in Pakistan were earlier reported among employees of a construction contractor at the power generation plant in Baluchistan in 1995 [70], but no further DEN cases were reported in this HKH country until 2007. Lack of proper surveillance systems or complexity in laboratory diagnosis due to infection with other diseases like malaria, typhoid, and hepatitis might cause an underreporting of cases [83]. In the meantime, frequently reported DEN incidence in this region from 2007 onwards [48] might be driven, in part, by increasing human mobility, particularly in areas with climatic suitability for the mosquito vector [84]. A mobile-phone-based study conducted in Pakistan shows the mobility of infected travelers from endemic regions to all other parts of the country during the outbreak in 2013 [85]. In Nepal as well, the improvement in social and infrastructure development such as the expansion of roads, rapid urbanization, and increase in trade and business opportunities after the end of decade-long armed conflict (1996)(1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006) has increased the mobility of people within and from neighboring countries [86]. This fact might be the cause for the rapid expansion of DENV in this HKH country, supported by entomological investigations that show the presence of Ae. albopictus in Nepal in 1956 already [87]. However, the primary vector for DEN Ae. aegypti was reported for the first time in 2006 from the lowlands of Nepal [76], later in 2009 from Kathmandu, which is a hilly region of Nepal [88], and reports of Ae. aegypti and Ae. albopictus from at least 2000 m above sea level [89]. In parallel, cases of DEN were reported from Terai (lowland) and hill and mountain regions (68 out of 77 districts of Nepal) in 2019 [15]. Although the recent study in Afghanistan [90] reported the first locally acquired DENV cases in 2019, our study shows that dengue antibodies were detected earlier, in between 2010 and 2011, among Afghan National Army recruits in Afghanistan. However, DEN was distributed more frequently or rapidly in the HKH countries of India, Nepal, Pakistan, and Bhutan. China is the only HKH country with no reported cases of DEN in the HKH region.
The distribution of CHIK was observed in Myanmar [73], India [17,26,27,31,46,60], Nepal [30,44,49], and in Bhutan [52]. Although CHIK was observed quite early in Myanmar in 1973 and 1974, it was most frequently reported in India. However, no clear conclusions can be drawn for CHIK due to the low number of articles in our database and due to the unavailability of local governmental reports.
The period of outbreaks of DEN and CHIK varies only slightly within the HKH region. Most of studies from India [18,24,40,51,56] and Nepal [19,23,29,32,41,43,61,64] show almost consistent seasonality of DEN outbreaks, lasting from June to December. The majority of the studies from India and Nepal reported maximum DEN cases in September, October, and November in the postmonsoon period [18][19][20][23][24][25][27][28][29][39][40][41][42][43]54,61,62]. Our findings concerning a seasonal distribution of DEN in HKH countries are similar to the studies on Brazil [91] and Thailand [92]. These findings are further supported by entomological studies conducted in Nepal [89] and in Assam, India [93], that show a high abundance of vectors Ae. aegypti and Ae. albopictus at the end of the monsoon and postmonsoon season (September to November) compared to the winter season. The peak for DEN transmission during the postmonsoon season might be due to the most favorable weather conditions, including moderate rainfall, mild mean temperature, and optimum temperature range, which help vectors to breed, survive, and reproduce [89]. Temperature fluctuations also influence DEN infection in mosquitos [94]. Accordingly, adult Ae. aegypti live longer and, thus, were more likely to become infected under moderate temperature fluctuations. Moderate temperature fluctuations are typical during high DEN transmission season, whereas large temperature fluctuations favor a low dengue transmission season [94]. In Ae. albopictus, a temperature fluctuation from 28 • -23 • -18 • C showed a probability of lower DEN transmission (in regard to virus titer in the salivery glands) than a constant 28 • C temperature [95]. Accordingly, vector activity and, thus, virus distribution are linked since temperatures influence, especially during the DEN season, virus transmission of the mosquitos.
The scientific reports of DEN and CHIK outbreaks in the HKH region have gradually increased from 2004/2005 until 2020 (Table S14). The reported incidence rates show an increase in disease burden over time ( Figure 3). Accordingly, this increasing trend indicates the expansion of these disease vectors in the HKH region. The distributional shifts and increasing abundance of those disease vectors are also directly or indirectly influenced by weather variables such as temperature, humidity, and precipitation [57]. The Intergovernmental Panel on Climate Change (IPCC) also concludes that anthropogenic climate change, in particular, the changing temperature and precipitation patterns, has already altered the distribution of vector-borne diseases worldwide [96]. Nevertheless, the impacts of climate change on the distribution of DEN vectors should always be considered in the context of multiple social, behavioral, economic, environmental, and health system factors. For instance, human mobility from DEN endemic areas to non-DEN-endemic areas is vital for determining the changes in exposure and susceptibility to the DEN virus in the face of climate variability and change [97]. Furthermore, problems with water scarcity and, consequently, human behavior can even increase the breeding opportunities for DEN vectors [86]. The dry climate could force local communities to store water in containers, which ultimately increases the breeding sites for Ae. aegypti and Ae. albopictus [89,98].

Limitations of the Study
We have calculated the total number of reported and confirmed DEN and CHIK cases from each HKH country, but we have certainly failed to report the true number of DEN and CHIK cases from each HKH country. More than one study was conducted in same period, especially in India [17,18,20,21,[24][25][26][27][28]36,39,40,42,50,51,54,56,57,60] and in Nepal [19,22,23,30,34,37,38,41,43,44,47,49,53,55,58,59,61,64,66,67,69]. This might have duplicated the DEN or CHIK cases in our database. In addition, we extracted the data only from English publications, which could account for unexpected deficits in scientific articles from China, Afghanistan, Bangladesh, and Myanmar in our database. Correspondingly, the underreporting of DEN or CHIK cases from these countries may be based on a lack of research in those particular VBDs or a poor DEN/CHIK surveillance system in the HKH areas of these countries. Additionally, the lack of an appropriate surveillance system for the differentiation of DEN/CHIK, a lack of appropriate diagnosis tools leading to a misdiagnosis of DEN/CHIK with malaria and other vector-borne diseases, and unfamiliarity of health workers with the epidemiology of DEN/CHIK may negatively bias the number of reported cases.

Conclusions
DEN and CHIK viruses' expansion is widespread in the HKH region and reveals an increasing trend of infection. DEN and CHIK are highly seasonal. DEN starts with the onset of the monsoon (July in India, June in Nepal) and with the onset of spring (May in Bhutan), and CHIK occurs throughout the whole year. Similarly, both DEN and CHIK peak in the postmonsoon season, i.e., September, October, and November. Temperature and climatic changes have a high impact on the distribution of mosquitoes and viruses. Therefore, an advanced understanding of the spatiotemporal distribution of DEN and CHIK in connection with the rapid changes in climate, infrastructure, social mobility, human behavior, and vector distribution in the HKH region is important for the improvement of DEN and CHIK disease control and vector control management. For this, outbreak preparedness and response, more climate change data, modeling, and supranational studies are needed. Further, DEN and CHIK diagnostics and reporting processes need improvement. The increasing prevalence of both diseases in more provinces of the HKH region requires the need for global coordination of response efforts across national borders.