Epidemiology of Chikungunya Virus Infection in the Philippines

1. Introduction and Clinical Significance

An arthritogenic mosquito-borne alphavirus belonging to the Togaviridae family, CHIKV was first isolated during the 1952-1953 Newala epidemic in southern Tanzania [1]. The outbreak was characterized by high attack rates and severe arthralgia, and the virus was successfully recovered from both affected patients and Aedes aegypti mosquitoes, establishing its role as the causative agent and confirming mosquito-borne transmission [2].

However, historical evidence suggests that CHIKV caused outbreaks long before its formal identification. Affected individuals were recognized primarily by their clinical presentation, predominated by debilitating arthralgia. This gave rise to descriptive local terms such as “knuckle fever,” first reported in 1779 in present day Jakarta, reflecting the prominent involvement of peripheral small joints. Similar clinical patterns were later documented during the same year in Cairo (1779), as well as in later major urban epidemics across India, most notably in Calcutta in 1824 [3].

In the Philippines, evidence of CHIKV circulation dates to investigations of the epidemic hemorrhagic fevers conducted in the 1950s. During a major 1956 outbreak of hemorrhagic fever in Manila and surrounding Luzon, Dr. William M. Hammon first suspected the presence of chikungunya because some children suffered from unusually severe crippling joint pain[4]. However, the virus itself was not isolated until 1965 by Dr. L. E. Campos and his team[5]. This timeline confirmed earlier suspicions regarding the earlier co-circulation of these two pathogens; historical reviews by virologists like Scott B. Halstead verified that during this period, dengue viruses were responsible for severe hemorrhagic cases while chikungunya co-circulated as the cause of a clinically similar but milder febrile illness in the region[6] Consistent with these early descriptions, a 1968 case series from the Philippines involving 41 serologically confirmed patients reported a predominantly mild febrile illness characterized by rash and arthralgia (66%), without severe complications [6]. Sporadic infections were later identified among US Peace Corps volunteers in 1985–1986, suggesting a low transmission setting in the Philippines[7].

We conducted a systematic review of human CHIKV studies in the Philippines to summarize the epidemiological patterns and outbreak dynamics that may inform public health policies and disease control strategies.

2. Methodology

This systematic literature review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines to address the following question: “What is the epidemiology of CHIKV infection among individuals residing in the Philippines, including incidence, prevalence, seroprevalence, outbreak characteristics, attack rates, and disease burden in both laboratory-confirmed and clinically suspected cases. The study protocol was registered in the International Prospective Register of Systematic Reviews (PROSPERO) (protocol CRD42024544293).

2.1. Information Sources

We systematically searched PubMed and Embase covering literature available up to March 2026. Additionally, we included grey literature from the Philippine Department of Health (DOH) Epidemiology Bureau, specifically the national chikungunya surveillance dataset (2018–March 2026) obtained directly from the DOH.

2.2. Search Strategy

The strategy combined keywords and MeSH terms as follows: (chikungunya[Title/Abstract] OR CHIKV[Title/Abstract] OR "chikungunya virus"[MeSH] OR "chikungunya fever"[Title/Abstract] OR CHIK[Title/Abstract]) AND (Philippine*[Title/Abstract] OR Philippines[Title/Abstract] OR Manila[Title/Abstract] OR Cebu[Title/Abstract] OR Davao[Title/Abstract] OR Laguna[Title/Abstract] OR Mindanao[Title/Abstract] OR Luzon[Title/Abstract] OR Quezon[Title/Abstract] OR "Metro Manila"[Title/Abstract] OR Calabarzon[Title/Abstract] OR Visayas[Title/Abstract]) AND (epidemiolog*[Title/Abstract] OR outbreak*[Title/Abstract] OR incidence[Title/Abstract] OR prevalence[Title/Abstract] OR seroprevalen*[Title/Abstract] OR surveillan*[Title/Abstract] OR burden[Title/Abstract] OR "attack rate"[Title/Abstract] OR coinfect*[Title/Abstract] OR transmission[Title/Abstract] OR seroepidemiolog*[Title/Abstract] OR "disease burden"[Title/Abstract]).

2.3. Eligibility Criteria

Studies were eligible for inclusion if they were conducted in human populations in the Philippines and reported original epidemiological data or secondary analyses of surveillance datasets on CHIKV infection, including incidence, prevalence, seroprevalence, outbreak characteristics, surveillance findings, or disease burden. Eligible study designs included cross-sectional studies, cohort studies, case series, surveillance reports, outbreak investigations, and conference proceedings with epidemiological elements, provided they contributed unique human data not captured in full publications.

2.4. Selection Process

Two independent reviewers (N.A.J. and H.A.I.) screened titles and abstracts based on the eligibility criteria using Rayyan software, classifying studies as included, excluded, or uncertain. Studies with uncertain relevance were retained to ensure that all potentially eligible articles were considered. Full-text articles were subsequently retrieved and independently assessed for eligibility. Disagreements were resolved through discussion, with a third reviewer (P.C.) involved when necessary. The same process was applied during full-text screening.

Only articles published in English were included. Studies were excluded if they were reviews, editorials, commentaries, or opinion pieces; if they reported only entomological or laboratory-based findings without human epidemiological outcomes; or if they were not specific to the Philippines. Multi-country studies were excluded unless Philippines-specific data could be extracted.

2.5. Data Collection Process

Two independent reviewers (N.A.J. and H.A.I.) extracted data using a standardized form. A second reviewer independently verified the extracted data to ensure completeness and accuracy. Any discrepancies were resolved through discussion, with a third reviewer (P.C.) involved if necessary.

2.5. Data Items

The data extraction form captured key study characteristics, including author, publication year, study design, location, population, sample size, diagnostic methods, and epidemiological outcomes such as incidence, prevalence, seroprevalence, outbreak characteristics, attack rates, and disease burden, as well as main findings.

3. Results

A total of 133 records were identified through database searching (PubMed: 39; Embase: 94). After removal of duplicates and screening of titles and abstracts, 19 unique records were selected for full-text assessment. Three conference abstracts (Yoon 2014, Sy 2015, Sy 2012) were excluded because each was later published as a full paper. Sixteen studies met the inclusion criteria and were included in the systematic review (Figure 1).

3.1. Study Characteristics

The included studies comprised cross-sectional studies, prospective cohort studies, case series, and outbreak investigations, and one multi country retrospective seroprevalence study from which Philippines specific data were extracted. Studies were conducted across Cebu, Manila, Laguna, and Mindanao and other regions (Table 1).

Study periods ranged from the 1950s to recent years, capturing both historical and contemporary epidemiological patterns of CHIKV infection.

Sample sizes varied substantially, from small outbreak investigations to large community-based cohorts and hospital based seroprevalence studies. Diagnostic approaches included serological assays (IgM and IgG ELISA), plaque or focus reduction neutralization tests (PRNT/FRNT), and molecular methods such as reverse transcription polymerase chain reaction (RT-PCR). Several studies analyzed dengue-suspected cases retrospectively, reflecting diagnostic overlap between arboviral infections.

Several studies analyzed dengue-suspected cases retrospectively, reflecting diagnostic overlap between arboviral infections. One hospital series tested dengue-suspected inpatients at San Lazaro Hospital and St. Luke’s Medical Center in 1999 with RT-PCR, antigen sandwich ELISA, IgM-IFA, IgM-capture ELISA, and IgG ELISA[8]. A separate 2001 multi-country serosurvey contributed 160 Philippine sera tested by IgM-capture ELISA, IgG ELISA, and CHIKV FRNT[9].

The studies reported incidence, seroprevalnce, outbreak characteristics, attack rates, and disease burden. Considerable heterogeneity in study design, populations, and outcome reporting precluded quantitative meta-analysis.

3.2. Epidemiological Findings

The included studies cannot be pooled quantitatively. They can be compared on five axes: when transmission ocurred, where it clustered, how immunity and incidence co-evolved, how diagnostic method shaped the apparent burden, and which viral lineages circulated.

3.2.1. Transmission Dynamics

Evidence from historical virus isolation and serological studies, reconstructive modelling, prospective cohorts, and contemporary surveillance suggest a broadly consistent pattern of periodic CHIKV transmission in the Philippines. CHIKV was recovered during investigations of epidemic hemorrhagic fever in the 1950s and was isolated again in the mid-1960s[4,5]

By the late 1960s–mid-1970s, CHIKV neutralizing antibody was already present on several islands: 13/58 adults sampled in 1968–1970 (22%), 8/58 in Tagudin, Luzon (14%), 18/101 in Marinduque (18%), and 20/67 in Atlas, Cebu (30%). No Ross River neutralizing antibody was detected at these sites[10]. Age-specific rates rose with age in Cebu and Mindanao, consistent with repeated rather than single-wave exposure. Subsequent reports documented sporadic infections among Peace Corps volunteers in 1986 and a community outbreak in Cavite in 1996. A 1999 hospital series from Metro Manila found no CHIKV RNA in 315 dengue-suspected sera, but antigen sandwich ELISA high-titer results clustered in August–December, consistent with rainy-season transmission in the interval between the 1996 Cavite outbreak and the 2011–2013 epidemic[8]. A retrospective serosurvey of dengue-suspected patients’ sample in the Philippines in 2001 found CHIKV neutralization confirmed infection in 26.8% of specimens, indicating substantial prior or recent exposure in the interval between in the 1996 Cavite outbreak and the 2011-2013 epidemic[9]. Active fever surveillance among 300 children aged 2-14 years in San Pablo and Cebu in 2010-2011 detected CHIKV IgM in 35 children, an incidence density of 21.3 per 100 person years (95% CI 15.3-29.6)[11].

The reconstructed waves in Cebu were concentrated in 1968–1969 (about 18% of remaining susceptibles; ~135,000 infections), 1986 (37%; ~190,000), and 1993 (14%; ~32,000), in addition to the directly observed 2012–2013 wave (106 of 853 participants seroconverted; 17% of susceptible)[12]. At enrolment in 2012, 241 of 853 participants (28%) were PRNT-positive, but none of 294 children younger than 14 years had detectable antibody, consistent with negligible circulation during the preceding 14 years [12]. Large outbreaks were followed by low level transmission.

Findings from the 2012–2014 prospective Cebu cohort provide additional support for this pattern [13]. During the first year of surveillance, overall incidence was 12.32 infections per 100 person years, increasing to 16.9 per 100 person-years among participants who were seronegative for neutralizing antibodies. In the second year, incidence declined to 2.84 infections per 100 person-years and seroprevalence increased from 28% to 42%.

3.2.2. Spatial Pattern: Introduction Through Mindanao, and Local Transmission

The 2011–2013 epidemic was first detected in Mindanao before being recognized in other parts of the Philippines [14]. All available sequences from other parts of the country during this period belonged to the Asian genotype, where East/Central/South African (ECSA) Indian Ocean lineage viruses were detected only in Davao del Sur and Davao Oriental[14]. Phylogeographic analysis favored an introduction from Indonesia. Full genome analysis of 2012 isolates placed acquisition of nsP3-D483N in Compostela Valley and subsequent E1-P397L and E3-Q19R, consistent with export after the Laguna-associated substitutions[15]. A single 2012 isolate from Davao (CK12-455) belonged to the Indian Ocean Lineage of the ECSA genotype[15] . Following introduction, transmission appeared to be spatially clustered at a local level. local. In Cebu, Participants who seroconverted in 2012–2013 were 2.2 times as likely to live within 50 m of another seroconverter (95% CI 1.5–3.1) and 1.7 times as likely to live within 100 m (95% CI 1.2–2.3); clustering remained significant out to 230 m[12]. In Concepcion Village, San Pablo City (Laguna), 98 clinically defined cases occurred between 23 June and 6 August 2012 (village attack rate 1.2%; 98/7,881). Fifty-eight of 84 tested sera (69%) were CHIKV IgM-positive. One patient was hospitalized and none died. In an unmatched case-control analysis, illness was associated with a chikungunya case in the household (aOR 6.2, 95% CI 3.0–12.9) and haphazard garbage disposal (aOR 2.7, 95% CI 1.4–5.4). House and Breteau indices were 27% and 28, above national dengue-programme thresholds, with larvae predominantly Aedes aegypti [16].

More recently, DOH surveillance data showed marked regional variation, with the largest number of reported cases in Cavite, Laguna, Batangas, Rizal, and Quezon (CALABARZON) n=2,619, Mindoro, Marinduque, Romblon, and Palawan (MIMAROPA) n=2,285, and the Cordillera Administrative Region (CAR) n=1,309. This distribution suggested that reported CHIKV transmission geographically restricted to regions as shown in (Figure 2).

All age groups were affected, with adults aged 25 [7]-34 years accounted for the largest number of proportions of reported cases, followed by children aged 5 to 14 years and adults aged over 60 years. In Davao de Oro in 2019-2020, CHIKV IgM positive febrile cases were unevenly distributed across four provincial hospital’s, with higher proportion at Laak and Pantukan than at Montevisa [17].

3.2.3. Immunity, Incidence and the Under Recognized Burden of Infection

In Cebu in 2012, baseline neutralizing antibody seroprevalence was 28% [18,19]. During the first year of active surveillance, subclinical incidence (10.03 per 100 person-years) exceeded symptomatic incidence (2.19 per 100 person-years), only 19% of infections were symptomatic [18]. Detectable preexisting neutralizing antibody was associated with no symptomatic infections in that year[18]. The ratio of subclinical to symptomatic infection was 4:6:1 overall and increased with age, from 2:1 in children aged >6 months-5 years to 12:1 in adults older than 50 years. A baseline PRNT titer >10 was associated with 100% protection from symptomatic infection (95% CI 46.1-100)[18]. When both surveillance years were analyzed together, 25 symptomatic infections and 104 subclinical seroconversions occurred among participants with no detectable pre-year neutralizing antibody (615 in year 1; 444 in year 2). Among those with a pre-year PRNT₈₀ titer ≥1:10 there were no symptomatic infections and one subclinical seroconversion (relative risk 0 for symptomatic infection and 0.018 for subclinical seroconversion). Neutralizing antibody remained detectable through two years of follow-up [19].

Among dengue-suspected inpatients in Manila and Quezon City in 1999, Inoue et al. detected no CHIKV by RT-PCR (0/315)[8]. Serological reactivity was higher and assay-dependent: IgM-IFA 15.3% (90/590), IgM-capture ELISA 13.7% (47/343), IgG ELISA 18.5% (91/492), and antigen sandwich ELISA high-titer 4.1% (26/640). Dual dengue–CHIK reactivity was 7.5% (44/590) in the IFA-linked subset and 5.5% (19/343) in the IgM-capture ELISA subset; the authors estimated that 7–8% of suspected DF/DHF cases were reactive for CHIKV only[8].

Additionally, dengue suspected sera collected in the Philippines in 2001, CHIKV IgG ELISA positivity was 57.5% and neutralization confirmed infections was 26.8%[9]. During the 2012-2013 epidemic, 11 of 112 patients (10%) with a clinical diagnosis of dengue at a tertiary hospital in Quezon City were CHIKV RT-PCR positive, and none were coinfected with dengue [20]. In four provincial hospitals in Davao de Oro in 2019–2020, CHIKV IgM was detected in 7.07% as a single infection and in a further 7.33% together with dengue markers (any CHIKV IgM, 14.4%)[17].

In contrast to hospital convenience samples, a prospective pediatric cohort in San Pablo and Cebu in 2010–2011 found CHIKV IgM incidence of 21.3 per 100 person-years among children aged 2–14 years[11].

DOH recorded 12,662 cases over eight years, of which only 5.7% were laboratory-confirmed, and only three deaths (CFR<0.03%).

3.2.4. Diagnostic Method as a Source of Apparent Heterogeneity

Differences in diagnostic methods may contribute to the heterogeneity observed across studies. An early Metro Manila hospital study combined RT-PCR after C6/36 inoculation with antigen ELISA on culture fluid, IgM-IFA, IgM-capture ELISA, and IgG ELISA, but without PRNT or FRNT [8]. RT-PCR was uniformly negative, while IgM-IFA (15.3%) and IgM-capture ELISA (13.7%) were numerically similar but not statistically associated[8].

Early multi-country work used plaque-reduction neutralization assays for seroprevalence among populations in Luzon, Marinduque, Cebu, and Mindanao [10]. A Later multicounty hospital serosurvey applied IgM capture ELISA, IgG ELISA and FRNT₈₀ [9]. Later hospital and outbreak investigations utilized IgM-capture ELISA, sometimes with RT-PCR restricted to IgM-negative acute samples. During 2011–2013, 2,891 of 5,729 outbreak-surveillance sera were IgM-positive. Among 382 IgM-negative specimens subsequently tested by RT-PCR, 131 were positive for CHIKV [14].

3.2.5. Molecular Epidemiology

Molecular studies provide additional evidence of multiple CHIKV lineages circulating during the 2011–2013 epidemic. The Asian-genotype sequences showed high nucleotide similarity (>99%) and clustered with Indonesian strains, consistent with an introduction into the Philippines before 2010[14]. An independent Japanese series of imported infections identified three laboratory confirmed cases acquired in the Philippines in 2012-2013. The isolates recovered from a viremic traveler in 2012 belonged to the Asian genotype and clustered with contemporaneous Pacific and Caribbean strains, consistent with dissemination of the Cosmopolitan Asian lineage from the Philippines[21]. Two E1 sequences from Manila patients in 2012-2013 also belonged to the Asian genotype and clustered with viruses subsequently reported from Micronesia and the Caribbean[20]. Through the Western and South Pacific [21] with the three ECSA viruses belonged to the Indian Ocean lineage, carried the E1-A226V substitution, which has been associated with enhanced transmission by Aedes albopictus. These viruses were closely related to a 2010 Indonesian virus[14,22]. Independent analyses of isolates collected in 2012 likewise supported the emergence of a cosmopolitan Asian lineage in the Philippines rather than importation of Indian Ocean ECSA as the principal driver of the outbreak [15].

4. Discussion

CHIKV transmission in the Philippines is best understood as episodic and focal process, characterized by outbreaks separated by prolonged inter-epidemic intervals during which population susceptibility gradually accumulates. Our systematic review identified chikungunya outbreaks as early as 1956 in Manila and surrounding Luzon, with subsequent episodic events documented in the late 1960s, 1980s, 1990s, and a large nationwide outbreak in 2011-2013. Reconstructive serology identified four major waves and an approximately 17-year inter-epidemic interval, while longitudinal observations from Cebu documented the decline of transmission as neutralizing antibody prevalence approached 40% [12,13]. More recent DOH surveillance data from 2018-2026 show a similar pattern, with recurrent peaks exceeding 2,900 reported cases in 2018, 2023 and 2024, interspersed with substantially lower-case numbers in intervening years. Together, these findings support a recurrent cycle in which outbreaks infect only a proportion of susceptible individuals, leaving sufficient population susceptibility to permit subsequent re-emergence[13].

The apparent size and frequency of these outbreaks are likely substantially underestimated by routine surveillance. Community based serological studies indicate that a large proportion of CHIKV infections are subclinical or sufficiently mild that affected individuals do not seek care[18]. Hospital-based studies, including a 1999 Metro Manila series of dengue-suspected inpatients[8], preferentially capture patients presenting with dengue-like illness and may therefore combine true CHIKV infections with cases that would otherwise be attributed to dengue. Diagnostic limitations further contribute to under-recognition for example, IgM ELISA testing among outbreak suspects may miss patients presenting early during viremia and may be affected by serological cross-reactivity [19].

The epidemiological and genetic evidence further suggests that the Philippines is connected to regional CHIKV transmission networks rather than being characterized by a single persistent endemic lineage. During the 2011-2013 outbreak, the Asian genotype predominated, while the ECSA were detected sporadically and were reported primarily in Mindanao[14,15]. Asian genotypes viruses related to Indonesian strains and ECSA viruses carrying the E1-A226V mutation were identified in Mindanao during. 2011-2013 outbreak, before Asian-genotypes viruses became more widely disturbed nationally[15,22,23]. These observations are consistent with repeated viral introductions and dissemination through regional human movement.

However, molecular surveillance after 2013 remains limited, making it difficult to determine whether ECSA viruses remained geographically restricted to southern Mindanao or whether subsequent local evolution and lineage replacement occurred. The limited availability of contemporary whole-genome data is an important gap in understanding the molecular epidemiology of CHIKV in the Philippines.

Although outbreaks can involve substantial numbers of infections, mortality appears to be uncommon. Only isolated fatal cases associated with atypical complications, including vasculitis and renal failure, have been documented[23]. While large cohort and surveillance studies have reported no deaths[13,17,18]. Recent DOH surveillance recorded three deaths among 12,662 reported cases, corresponding to a very low reported case fatality ratioCRF Thus, mortality is unlikely to represent the principal public health burden of chikungunya in the Philippines. Instead, the major consequences include prolonged musculoskeletal symptoms in a subset of symptomatic patients, diagnostic confusion with dengue, and the healthcare and public health resources required during recurrent outbreaks (Add Ref. persistence symptoms). Community attack rates may be relatively modest for example, an attack rate of 1.2% was reported during the 2012 outbreak in Laguna. Nevertheless, strong household clustering indicates that transmission can become highly focal within affected communities [16].

Environmental and behavioral factors appear to contribute to this focal transmission. Household clustering and factors such as inadequate waste segregation, water storage, and other conditions that provide opportunities for Aedes mosquito breeding have been associated with transmission [16,24]. These findings suggest that control strategies should complement national surveillance with targeted, community level interventions in areas where transmission is detected or where environmental conditions favor focal amplification. Between outbreaks, sustained community engagement and vector control activities may therefore be more appropriate than relying exclusively on broad, centralized emergency responses initiated after case numbers have already increased.

Several operational implications follow from these findings. First, repeated serological surveys using neutralization-based assays in the same populations at regular intervals could provide a more informative measure of population susceptibility and help anticipate areas at risk of future outbreaks than passive case counts alone. Second, surveillance of acute febrile illness should consider CHIKV and dengue concurrently, particularly in settings where clinical differentiation is difficult, with RT-PCR or other molecular testing prioritized during the first week of illness when IgM may remain negative. Third, molecular surveillance should be strengthened in areas that may serve as points of viral introduction or amplification, particularly southern Mindanao, and should be intensified when passive surveillance indicates an increase in suspected or confirmed cases. Such an approach could improve early recognition of re-emergence and provide information on the lineage responsible for subsequent outbreaks.

This review has several limitations. Restriction to English-language and open-access articles may have introduced selection bias and excluded relevant local-language or paywalled studies. The search was limited to two major bibliographic databases and available national surveillance data and therefore may not have captured all provincial outbreak reports. Considerable heterogeneity in study populations, study designs, case definitions, and diagnostic methods limited comparability and precluded quantitative meta-analysis. Older serological studies, particularly those without neutralization testing, are also vulnerable to cross-reactivity with other alphaviruses and arboviruses. Furthermore, many of the highest-quality longitudinal data originate from Cebu and may not be representative of other regions, particularly Mindanao and geographically distinct areas such as the Cordillera Administrative Region. Hospital-based studies may also underestimate community-level transmission because they preferentially capture symptomatic individuals who seek healthcare. Finally, the scarcity of contemporary genomic data limits conclusions regarding the geographic distribution, persistence, and evolution of circulating CHIKV lineages.

Despite these limitations, the available evidence consistently supports an episodic, focal pattern of CHIKV transmission in the Philippines, with substantial inter-epidemic accumulation of susceptible individuals. The persistence of this pattern over several decades suggests that future outbreaks are likely, although their timing, geographic distribution, magnitude, and dominant viral lineage remain uncertain. Improving population-based Sero surveillance, integrating CHIKV testing into acute febrile illness surveillance, strengthening molecular surveillance, and maintaining targeted community-level vector control will be important for detecting and responding to future re-emergence before transmission becomes widespread.

5. Conclusion

Chikungunya virus transmission in the Philippines is characterized by episodic and focal outbreaks separated by prolonged inter-epidemic intervals, during which population susceptibility accumulates and facilitates recurrent re-emergence. The true burden is likely substantially underestimated by routine surveillance because of subclinical infections, healthcare-seeking behavior, diagnostic limitations, and clinical overlap with dengue. Strengthening integrated dengue–CHIKV surveillance, repeated serological assessment of population susceptibility, molecular monitoring of circulating lineages, and targeted community-level vector control will be important for improving preparedness and reducing the impact of future outbreaks