Epidemiological investigation and risk factors of Peste des Petits Ruminants (PPR) in goats of Northeastern Bangladesh

Highlights

• •PPR remains endemic in northeastern Bangladesh goats.
• •Seroprevalence (35%) exceeds molecular prevalence (16.6%).
• •Herd-level exposure reached 75.6% across sampled households.
• •Supply water and free-ranging grazing associated to PPRV infections.

Abstract

Peste des Petits Ruminants (PPR) remains a major constraint to small ruminant health and rural livelihoods in Bangladesh. This cross-sectional study investigated the molecular and serological prevalence and associated risk factors of PPR among 368 unvaccinated goats sampled from eight upazilas in northeastern Bangladesh between April and July 2024. Competitive ELISA and RT-PCR targeting the N and F genes were used to detect PPR-specific antibodies and molecular detection, respectively. The seroprevalence was 35.05% (95% CI: 30.18–40.17), while the molecular prevalence was 16.6% (95% CI: 12.92–20.78). At the household-level (n = 90), the molecular prevalence was 41.11% (95% CI: 30.81–51.98), while the seroprevalence was 75.56% (95% CI: 65.29–84.07). Among clinical signs, coughing was strongly associated with PPR virus infection. Multivariable logistic regression identified two potential key factors of PPR infection: water source and grazing system. Goats drinking supply water showed higher odds of PPR positivity compared to those using pond water (aOR = 2.19; 95% CI: 0.97–5.13), although this association was not statistically significant (p > 0.05) and should be considered as borderline. In contrast, goats managed under free-ranging systems had significantly higher odds of PPR positivity compared to gidding grazing systems (aOR = 3.64; 95% CI: 1.15–16.13). These findings underscore the role of management practices and environmental exposure in PPR transmission. Targeted improvements in water management, grazing practices, and surveillance are essential for Bangladesh to achieve its goal of PPR eradication by 2030.

1. Introduction

Goats are integral to the agricultural economy of Bangladesh, contributing to meat, milk, and skin production, and providing income to rural households, particularly smallholders and marginal farmers. The country has approximately 27.1 million goats, primarily of the Black Bengal breed, noted for high fertility and meat quality (DLS, 2024). Goats provide essential supplementary income to low-resource households and play a critical role in supporting rural livelihoods. However, Peste des petits ruminants (PPR) poses a major threat to this sector. The highly contagious viral disease can lead to severe outbreaks, resulting in substantial economic losses. Globally, PPR causes an estimated 2.1 billion USD in annual losses (OIE-FAO, 2015), and in Bangladesh, losses may reach 25 million USD annually (Islam et al., 2013), with morbidity, mortality, and case fatality rates reported at 75%, 59%, and 78%, respectively (Chowdhury et al., 2014).

PPR, commonly known as “goat plague,” is a viral disease of domestic and wild small ruminants first reported in Côte d’Ivoire in 1942 (Baazizi et al., 2017). The causative agent, PPR virus (PPRV), is a single-stranded RNA virus of the Morbillivirus genus (Couacy-Hymann et al., 2007a), closely related to rinderpest virus, canine distemper virus, measles virus, and various marine mammal morbilliviruses (Khalafalla et al., 2010). The disease primarily affects sheep and goats and is endemic across much of Asia and Africa, spreading to >70 countries by 2016 (Balamurugan et al., 2014; Zhuravlyova et al., 2020). Bangladesh reported its first outbreak in 1993 (Islam et al., 2001), and the disease remains endemic, with transmission occurring through respiratory secretions, ocular discharge, feces, contaminated feed or water, and direct contact (Banyard et al., 2010). Clinical signs include fever, ocular–nasal discharge, pneumonia, stomatitis, mucosal ulceration, and severe diarrhea (Balamurugan et al., 2014; Jaisree et al., 2018). Seroprevalence studies show considerable variation across PPR-endemic regions, including 82.60% in Nepal, 46.3% in India, 28.29% in Pakistan, 68.98% in Ethiopia, and 27.3% in Uganda (Acharya et al., 2018; Balamurugan et al., 2020; Baloch et al., 2021; Dubie et al., 2022; Nkamwesiga et al., 2023). In Bangladesh, earlier research reported a 21% seroprevalence in 2008 (Sarker & Islam, 2011), while a recent pooled estimate suggested a prevalence of 31.02% (Munibullah et al., 2022). Although multiple studies have described PPR epidemiology at the national level (Nabi et al., 2018; Rahman et al., 2018), there is limited district-level information for northeastern Bangladesh, despite evidence of a 35% seroprevalence in Sylhet (Siddiqui et al., 2023).

Northeastern Bangladesh is distinct from other regions due to its haor wetlands, hilly landscapes, high small-ruminant density, and cross-border livestock movement, all of which may influence local transmission dynamics. However, most national studies either include very small numbers of samples from Sylhet or pool the region with other districts, limiting their ability to detect localized patterns of risk. This underrepresentation creates a clear regional knowledge gap, particularly regarding husbandry practices, environmental exposure, and management factors unique to this area. Epidemiological studies have identified various factors associated with PPR transmission, including animal movement, grazing pattern, herd mixing, seasonality, limited veterinary services, and low vaccination coverage (Nkamwesiga et al., 2023; Cao et al., 2018; Gelana et al., 2020). Socio-economic conditions and management practices further shape disease risk (Dubie et al., 2022).

Given these interacting ecological and management factors, a region-specific epidemiological investigation is essential to identify locally associated factors influencing PPR transmission. This study is particularly timely in light of Bangladesh’s commitment to the Global Strategy for the Control and Eradication of PPR, which aims for global eradication by 2030 (FAO-WOAH, 2022). Accordingly, this study sought to address the regional evidence gap by: (1) estimating goat- and household-level serological and molecular prevalence of PPR in northeastern Bangladesh; and (2) identifying animal-level risk factors associated with PPRV infection.

2. Materials and methods

2.1. Ethical statement

This study was performed in accordance with the ethical standards and guidelines set by the Ethical Review Committee at Sylhet Agricultural University, Sylhet-3100, Bangladesh, approved the research protocol #AUP2023012. All protocols involving animals were reviewed and approved to ensure ethical conduct throughout the study. Where applicable, informed consent was obtained from relevant stakeholders, and efforts were made to minimize harm and distress to all subjects involved in the research.

2.2. Study area

The study was conducted in the Sylhet division of northeastern Bangladesh, comprising four districts: Sylhet, Moulvibazar, Habiganj, and Sunamganj. Two upazilas were randomly selected from each district—Golapganj and Gowainghat (Sylhet), Jagannathpur and Chhatak (Sunamganj), Moulvibazar Sadar and Rajnagar (Moulvibazar), and Habiganj Sadar and Nabiganj (Habiganj). Within each upazila, three villages were selected using simple random sampling. Sylhet division lies between 23°58′–25°12′N and 90°56′–92°30′E, characterized by floodplains, river basins, and hill tracts connected to the Meghalaya and Khasia ranges. All sampling sites and households were georeferenced using a handheld GPS and mapped in ArcMap 10.8 (Fig. 1).

Fig. 1.

This map illustrates the distribution of goat households in the northeastern region of Bangladesh, specifically in the Sylhet division. Highlighted areas represent the selected upazilas for the study, with blue shading indicating the study area. Black triangles mark the locations of specific households where samples were collected.

2.3. Study design and sampling strategy

A cross-sectional epidemiological study was conducted from April to July 2024 to estimate the serological and molecular prevalence of PPR in goats and to identify associated risk factors. The study focused on goats that had not received PPR vaccination in the past 12 months to prevent vaccine-induced antibodies from affecting the serological results. A multi-stage random selection process was used to obtain a representative sample from the study region, which included districts, upazilas, villages, households, and goats. Initially, districts were randomly selected, followed by the random selection of upazilas within each district and three villages within each upazila. Households were then randomly chosen in each village, and representative goats were randomly selected from each participating household. The minimum required sample size was calculated using Thrusfield’s formula based on an expected seroprevalence of 35% from a recent Sylhet study (Thrusfield, 2018), resulting in a target of 350 animals; however, to increase statistical precision and ensure proportional representation across regions, a total of 368 goats were sampled from 90 households. From each selected goat, two types of specimens were collected: a blood sample for serological testing and a nasal swab for molecular detection of PPRV (Fig. 2).

Fig. 2.

Overview of the sampling design for assessing PPRV seroprevalence and viral detection in goats from northeastern Bangladesh.

2.4. Data collection

Data were collected at both the household and individual animal levels using a structured, pre-tested questionnaire administered through face-to-face interviews. Household-level information included housing conditions, drainage and ventilation quality, cleaning frequency, grazing practices, water sources, and potential wildlife exposure. At the animal level, data were recorded for each goat’s unique identification number, age, sex, breed, observable clinical signs, and relevant management or biosecurity practices. All questionnaire responses were entered electronically in real time, checked for completeness, and later merged with laboratory results for statistical analysis.

*Gidding grazing system: A controlled grazing practice where goats are tethered to a fixed point and allowed to graze within a limited radius, reducing movement and contact with other goats.

2.5. Serological examination (cELISA)

The serum samples were tested for PPR viral antibodies using the commercial cELISA kit, ID Screen® PPR Competition (IDvet Innovative Diagnostics, Grabels, France), following the manufacturer’s OIE-recommended protocol. The cELISA assay has a specificity of 99.4% and a sensitivity of 94.5% (Libeau et al., 1995). Optical density (OD) values were measured at 450 nm using a Benchmark Scientific 96-well microplate absorbence reader (Libeau et al., 1995). Each assay plate included the manufacturer-provided positive and negative controls to validate test performance, and all procedures were carried out following standard contamination prevention and quality-assurance practices.

2.6. RNA extraction

Total RNA was extracted from nasal swab samples using a commercial viral RNA extraction kit (AddBio Inc. Ltd., Korea). The procedure included lysis, binding of nucleic acids to a silica membrane, washing to remove impurities, and elution in RNase-free water. Extracted RNA was stored at –20°C until RT-PCR analysis.

2.7. RT-PCR amplification of N and F genes

RT-PCR was performed to detect Peste des Petits Ruminants Virus (PPRV) by targeting both the Nucleoprotein (N) and Fusion (F) genes. Two sets of primers were used: NP3 (forward: 5’-GTC TCG GAA ATC GCC TCA CAG ACT-3’) and NP4 (reverse: 5’-CCT CCT CCT GGT CCT CCA GAA TCT-3’) for the N gene (351 bp), and F1b (forward: 5’-AGT ACA AAA GAT TGC TGA TCA CAGT-3’) and F2d (reverse: 5’-GGG TCT CGA AGG CTA GGC CCG AAA TA-3’) for the F gene (448 bp) (Ishag et al., 2023; Nabi et al., 2018). Each 20 µL PCR reaction contained 10 µL of master mix, 1 µL each of forward and reverse primers, 5 µL of extracted RNA, and 3 µL of nuclease-free water. A known PPRV-positive RNA sample was included as a positive control, while nuclease-free water served as a negative (no-template) control to monitor contamination. The thermal cycling conditions included reverse transcription at 50°C for 30 min, initial denaturation at 95°C for 10 min, followed by 40 cycles of denaturation at 95°C for 30 s, annealing at 60°C for 30 s, and extension at 72°C for 1 min. A final extension was carried out at 72°C for 5 min, and the reactions were held at 4°C. PCR products were subjected to 1.5% agarose gel electrophoresis and visualized under UV light. Distinct bands at 351 bp and 448 bp indicated successful amplification of the PPRV N and F genes, respectively, in positive samples, while the negative control showed no amplification.

2.8. Case definition

For prevalence estimation, serological and molecular results were reported separately to distinguish prior exposure (cELISA) from active infection (RT-PCR targeting the N or F gene). For the risk factor analysis, a combined case definition was applied whereby a goat was classified as PPR-positive if it tested positive by either cELISA or RT-PCR. This approach was used to capture overall PPRV exposure and circulation within this endemic population, reflecting both current infection and evidence of past viral exposure relevant to transmission dynamics and eradication planning.

2.9. Statistical analysis

Statistical analyses were conducted in R 4.5.2. Clinical signs including nasal discharge, ocular discharge, fever, diarrhoea, ulcer, coughing, and stoma lesions were coded as binary variables and PPRV positivity served as the primary outcome. Initial associations between each clinical sign and PPRV positivity status were evaluated using χ² tests. Univariable odds ratios (ORs) with 95% confidence intervals (CIs) were calculated from 2 × 2 contingency tables; because several signs were infrequent and produced sparse-data bias, ORs and CIs were also estimated using Firth-corrected logistic regression.

Beyond clinical signs, associations between explanatory variables and PPR positivity were first assessed using univariable fixed-effects logistic regression (glm, binomial family). For each predictor, odds ratios (ORs), 95% confidence intervals (CIs), and Wald p-values were calculated, along with an overall likelihood-ratio (LR) p-value for the variable. Variables with an overall LR p-value ≤ 0.20 in the univariable screening were considered candidates for the multivariable model to avoid prematurely excluding potential confounders. All selected variables were then included in an initial multivariable logistic regression model, followed by a backward elimination procedure. Variables with overall LR p-values ≤ 0.05 were retained in the final model. The final model are presented as adjusted odds ratios (AOR) with 95% confidence intervals, providing estimates of the strength and precision of associations between management factors and PPR positivity. Potential confounding was assessed by monitoring changes in regression coefficients; variables that caused a ≥ 20% change in another coefficient were retained in the model. Multicollinearity among predictors was evaluated using variance inflation factors (VIF < 5). Although goats were sampled from multiple households and villages, only 3–5 animals were sampled per household, resulting in small within-cluster sample sizes. Preliminary evaluation indicated very low intraclass correlation (ICC), and therefore random-effects models were not considered appropriate. Instead, upazila was included as a fixed categorical variable to account for geographic variation. Model adequacy was evaluated using DHARMa residual diagnostics to assess dispersion, outliers, and zero inflation. Additional checks included the Hosmer–Lemeshow goodness-of-fit test and influence diagnostics (Cook’s distance). Model discrimination was assessed using the area under the receiver operating characteristic curve (AUC). Statistical significance was defined as p < 0.05.

3. Results

3.1. Exploratory data description

A total of 368 goats were sampled from eight upazilas in northeastern Bangladesh. Most goats were Black Bengal (89.4%) and female (63.9%). Age groups included ≤6 months (20.1%), 7–17 months (31.8%), and ≥18 months (48.1%). Over half were reared on soiled floors (52.4%), under moderate ventilation (56.0%), and most farms lacked footbaths (94.6%). Water sources included tube-wells (67.1%), ponds (23.1%), and supply water (9.8%). Free-range grazing was common (90.2%), and wildlife contact occurred in 69% of herds. Seasonal disease variation was reported by 59.8% of respondents. These variables were subsequently analyzed to assess their association with PPR infection (Table 1).

Table 1.

Serological and molecular positivity of caprine PPR in the Northeastern Region of Bangladesh.

Variable	Total (n)	Molecular positive			Seropositive
		Positive (n)	Molecular prevalence (%)	95% CI	Positive (n)	Seroprevalence (%)	95% CI
Overall	368	61	16.6	12.92 - 20.78	129	35.05	30.18–40.17
Herd Level	90	37	41.11	30.81% – 51.98	68	75.56	65.29 – 84.07
Upazila
Chhatak	49	11	22.45	11.77 - 36.62	22	44.90	30.67- 59.77
Golapganj	49	14	28.57	16.58 - 43.26	11	22.45	11.77–36.62
Gowainghat	51	15	29.41	17.49 - 43.83	17	33.33	20.76–47.92
Habiganj Sadar	50	2	4.00	0.49 - 13.71	23	46.00	31.81–60.67
Jagannathpur	49	6	12.24	4.63 - 24.77	14	28.57	16.58–43.26
Moulvibazar Sadar	48	10	20.83	10.47 - 34.99	17	35.42	22.16–50.54
Nabiganj	23	2	8.70	1.07 - 28.04	5	21.73	07.46–43.70
Rajnagar	49	1	2.04	0.05 - 10.85	20	40.82	26.99- 55.78
Floor type
Cemented	116	17	14.66	8.78 - 22.42	36	31.03	22.77–40.29
Slatted	59	8	13.56	6.04 - 24.98	27	45.76	32.71–59.24
Soiled	193	36	18.65	13.42 - 24.88	66	34.20	27.53–41.35
Ventilation
Good	8	3	37.50	8.52 - 75.51	2	25.00	03.18–65.08
Moderate	206	34	16.50	11.71 - 22.29	78	37.86	31.21–44.86
Poor	154	24	15.58	10.25 - 22.3	49	31.82	24.55–39.79
Clean animals regularly
No	237	35	14.77	10.51 - 19.94	84	35.44	29.35–41.89
Yes	131	26	19.85	13.39 - 27.71	45	34.35	26.27–43.14
Drainage
No	209	30	14.35	9.9 - 19.85	72	34.45	28.03–41.31
Yes	159	31	19.50	13.65 - 26.52	57	35.85	28.40–43.82
Water sources
Pond	85	12	14.12	7.51 - 23.36	15	17.65	10.22–27.43
Supply water	36	10	27.78	14.2 - 45.19	16	44.44	27.93–61.90
Tube-well	247	39	15.79	11.48 - 20.95	78	31.58	25.83–37.77
Grazing
Confinement	20	2	10.00	1.23 - 31.7	5	25.00	08.65–49.10
Free ranging	332	58	17.47	13.54 - 21.99	122	37.88	32.56- 43.43
Gidding	16	1	6.25	0.16 - 30.23	2	12.50	1.55–38.34
Wildlife contact
No	114	19	16.67	10.34 - 24.8	38	33.33	24.78–42.77
Yes	254	42	16.54	12.18 - 21.69	91	35.83	29.92–42.05
Footbath
No	348	58	16.67	12.91 - 21.01	123	35.34	30.32- 40.61
Yes	20	3	15.00	3.21 - 37.89	6	30.00	11.89–54.27
Seasonal variation
No	148	26	17.57	11.81 - 24.67	49	33.11	25.59–41.30
Yes	220	35	15.91	11.34 - 21.42	80	36.36	30.00–43.09
Breed
Black Bengal	329	59	17.93	13.94 - 22.51	116	35.26	30.09–40.68
Jamunapari	39	2	5.13	0.63 - 17.32	13	33.33	19.08–50.21
Age
≤ 6 months	74	14	18.92	10.75 - 29.7	16	26.66	16.07–39.66
7–17 months	117	15	12.82	7.36 - 20.26	23	32.85	22.09–45.12
≥ 18 months	177	32	18.08	12.71 - 24.55	90	37.81	31.63–44.30
Gender
Female	235	37	15.74	11.33 - 21.04	86	36.60	30.42–43.10
Male	133	24	18.05	11.92 - 25.65	43	32.33	24.48–40.98

3.2. Serological and molecular detection of PPRV in goats

Out of 368 goats tested across eight upazilas in northeastern Bangladesh, the molecular prevalence of PPR was 16.6% (95% CI: 12.92–20.78), while seroprevalence was higher at 35.05% (95% CI: 30.18–40.17) (Table 1). Molecular prevalence was highest in Gowainghat (29.41%), Golapganj (28.57%), and Chhatak (22.45%), and lowest in Rajnagar (2.04%). Seroprevalence was highest in Habiganj Sadar (46.0%), Chhatak (44.9%), and Rajnagar (40.82%), and lowest in Nabiganj (21.73%). Among the 368 nasal samples, 32 (8.7%) were positive for the F gene (95% CI: 6.2–12.0%) and 36 (9.8%) were positive for the N gene (95% CI: 7.1–13.2%) (Figs. 3 and 4). At the household or herd level (n = 90), the molecular prevalence was 41.11% (37/90; 95% CI: 30.81–51.98), while the seroprevalence was 75.56% (68/90; 95% CI: 65.29–84.07). Goats housed on soiled floors had a higher molecular prevalence (18.65%) compared to cemented (14.66%) or slatted floors (13.56%). Goats using supply water showed the highest molecular (27.78%) and seroprevalence (44.44%) compared to those using tube-wells or ponds. Free-ranging goats had higher molecular (17.47%) and seroprevalence (37.88%) than confined goats.

Fig. 3.

Prevalence of positive samples for F and N genes with 95% confidence intervals.

Fig. 4.

Amplification PPRV genes. Panel A shows amplification of the N gene, where samples (1–12) produced the expected 351 bp band. Panel B shows amplification of the F gene, with samples (1–6) yielding the expected 448 bp band. A positive control (PC) displays the corresponding bands for each gene, while the negative control (NC) shows no amplification. M = DNA ladders in both panels provide molecular size references.

3.3. Potential risk factors of PPRV

The forest plot highlights the clinical signs most strongly associated with PPR infection (Fig. 5). Coughing and ocular discharge emerged as the clearest indicators of viral detection, each showing odds ratios greater than 1 with comparatively narrow confidence intervals. Diarrhea showed a modest but uncertain increase in odds. In contrast, fever, ulcer formation, stoma lesions, and nasal discharge exhibited very wide confidence intervals, reflecting limited precision due to low frequencies and providing little statistical evidence of association with PPRV positivity.

Fig. 5.

Odds ratios (ORs) and 95% confidence intervals for clinical signs associated with PPRV infection. The vertical dashed line marks an OR of 1.0, indicating no association.

Risk factor analysis was performed using fixed-effects logistic regression to identify management and animal-level predictors associated with PPR infection. Several predictors showed weak or borderline associations; only water source and grazing system met the predefined criteria for entry into the final model. In univariable analysis (Table 2), most animal-level variables—including age, sex, breed, ventilation, drainage, floor type, seasonal variation, wildlife contact, and routine cleaning—did not show significant associations with PPRV positivity (p > 0.05). Geographic location (upazila) also did not demonstrate evidence of effect (p = 0.351), although Nabiganj showed lower odds compared with Chhatak (OR = 0.31; 95% CI: 0.10–0.89; p = 0.035).

Table 2.

Univariable and multivariable logistic regression analysis was conducted to examine potential risk factors associated with PPR positivity at the individual animal level.

		Univariable Analysis				Multivariable analysis
Variables	Characteristic	OR	95% CI	p-value	p-value	OR	95% CI	p-value	p-value
Upazila	Chhatak	Ref.			0.351
	Nabiganj	0.312	0.098- 0.889	0.035
	Golapganj	0.782	0.351- 1.729	0.544
	Gowainghat	1.076	0.489- 2.373	0.853
	Habiganj Sadar	0.753	0.339- 1.658	0.482
	Jagannathpur	0.560	0.248- 1.244	0.157
	Moulvibazar Sadar	0.813	0.364- 1.805	0.612
	Rajnagar	0.663	0.296- 1.467	0.312
Floor type	Cemented	Ref.			0.18
	Slatted	1.798	0.957- 3.408	0.069
	Soiled	1.187	0.746- 1.896	0.469
Clean animals regularly	No	Ref.			0.535
	Yes	1.144	0.746- 1.757	0.535
Footbath	No	Ref.			0.932
	Yes	0.961	0.378- 2.380	0.932
Drainage	No	Ref.			0.140
	Yes	1.365	0.902- 2.069	0.140
Ventilation	Good	Ref.			0.476
	Moderate	0.943	0.217- 4.086	0.935
	Poor	0.730	0.167- 3.190	0.664
Breed	Black Bengal	Ref.		0.186	0.180
	Jamunapari	0.628	0.308- 1.235
Age	≤ 6 months	Ref.			0.729
	≥ 18 months	1.212	0.703- 2.103	0.489
	7-17 months	1.050	0.584- 1.894	0.870
Gender	Female	Ref.			0.814
	Male	0.95	0.618- 1.455	0.814
Seasonal variation	No	Ref.			0.664
	Yes	0.911	0.600- 1.385	0.664
Water sources	Pond	Ref.			0.019	Ref.			0.032
	Supply water	2.146	0.966- 4.962	0.065		2.187	0.971- 5.134	0.063
	Tube-well	0.780	0.475- 1.281	0.326		0.809	0.489- 1.337	0.407
Wild animal contact	No	Ref.			0.936
	Yes	1.018	0.653- 1.589	0.936
Grazing*	Gidding*	Ref.			0.034	Ref.			0.041
	Free ranging	3.982	1.255- 17.60	0.033		3.642	1.145- 16.133	0.047
	Confinement	2.333	0.521- 12.72	0.285		1.987	0.431- 11.049	0.394

Two management factors emerged as notable potential associations of infection. Water source was significantly associated with PPRV positivity (p = 0.019), with goats using supply water having more than twice the odds of infection compared with those using pond water (OR = 2.15; 95% CI: 0.97–4.96). Likewise, the grazing system demonstrated a significant overall effect (p = 0.034), with goats under free-ranging management showing markedly higher odds of PPR positivity (OR = 3.98; 95% CI: 1.26–17.60) compared with those in the Gidding practices. Based on the screening criteria, only water source and grazing system were retained in the multivariable logistic regression model (Table 2). After adjustment, both factors remained statistically significant. Goats supplied water had higher adjusted odds of PPR positivity (aOR = 2.19; 95% CI: 0.97–5.13; p = 0.063) compared with those drinking from ponds, while goats managed under free-ranging systems exhibited substantially higher adjusted odds (aOR = 3.64; 95% CI: 1.15–16.13; p = 0.047) compared to the Gidding grazing system. Model diagnostics confirmed that the final logistic regression model was appropriate for the data (Supplementary Fig. 1). The Hosmer–Lemeshow test indicated no evidence of lack of fit (χ² = 7.67, df = 8, p = 0.47), suggesting that the final logistic regression model demonstrated adequate goodness of fit. The ROC curve indicated moderate discrimination (AUC = 0.649). DHARMa residual diagnostics showed no evidence of overdispersion (p = 0.906), no outliers, and no zero inflation (p = 1.00), with QQ plots indicating good alignment between observed and simulated residuals. Influence diagnostics demonstrated that no individual observations exerted undue influence on model estimates, as Cook’s distance values were all below the threshold. Overall, the diagnostic outputs support the adequacy and reliability of the fitted model.

4. Discussion

This study provides focused epidemiological evidence of PPRV circulation in northeastern Bangladesh, complementing the national-level patterns described in the Introduction. The seroprevalence of 35.05% reflects past exposure, while the PCR positivity of 16.6% indicates ongoing viral circulation and subclinical infections. This combination of high antibody prevalence with moderate molecular detection supports silent viral maintenance typical of endemic settings, consistent with previous reports from Bangladesh and neighboring countries (Rahman et al., 2018; Nabi et al., 2018; Chowdhury et al., 2022; Rahman et al., 2023). The observed seroprevalence aligns with recent findings from unvaccinated goats in Sylhet (Siddiqui et al., 2023), exceeds earlier nationwide estimates of 8.70% obtained using cELISA (Islam et al., 2016), and closely matches seroprevalence reported in Assam (27.28%) across the border (Devi et al., 2016). At a broader scale, meta-analytic estimates of 38.34% for Asia and 31.02% for Bangladesh (Ahaduzzaman, 2020) place our results within the expected regional range, reinforcing that northeastern Bangladesh remains a PPRV hotspot. Given the limited prior molecular data from this region, our findings fill a key knowledge gap and support evidence that PPRV can circulate subclinically, sustaining endemicity (Rahman et al., 2023). The substantially higher household-level or herd-level seroprevalence (75.56%) compared with molecular prevalence (41.11%) suggests that PPRV has been circulating widely in the area over time, with many herds showing evidence of past exposure while a smaller proportion had active infection detected at the time of sampling.

Spatial variation was observed, with Habiganj Sadar showing the highest seroprevalence and Gowainghat demonstrating the highest PCR positivity. Such heterogeneity may reflect differences in informal animal husbandry practices, movement networks, or recent introduction events rather than intrinsic geographic risk (Gao et al., 2021). Previous studies similarly report that management conditions and movement patterns often drive micro-level variation more than administrative boundaries. This is consistent with studies showing that individual and management-level factors often have a greater influence than geography in endemic regions (Rahman et al., 2023).

Age-related differences were observed, with goats older than 18 months showing a higher PPR seroprevalence and molecular positivity compared to younger animals. An important age-related factor considered was maternal immunity, which declines sharply after birth, dropping below protective levels by three to five months of age (Awa et al., 2002; Balamurugan et al., 2012). Young goats in endemic regions often experience early-life exposure that induces protective immunity, even in the absence of vaccination. As a result, many animals in the 6–18-month group may already have immunity from earlier subclinical infections. In contrast, older goats (>18 months) accumulate multiple exposure opportunities over time, especially through frequent trading, communal grazing, movement, and breeding-related contacts, all of which increase viral transmission risk. Several epidemiological studies have reported higher PPR seropositivity among adult goats, attributing this pattern to prolonged environmental exposure and greater mobility within production systems (Nkamwesiga et al., 2023; Torsson et al., 2017).

Clinical indicators offer further insight into the clinical expression of PPRV in endemic settings. Coughing and ocular discharge showed the strongest associations with PPR positivity, with odds ratios >1 and comparatively narrower confidence intervals, suggesting these signs may be more reliable indicators of viral shedding (Jones et al., 2020). Subclinical infection is increasingly observed in endemic settings and may reflect partial immunity from prior exposure, low-dose infection, or early-stage viral shedding that does not progress to overt clinical signs. Because PPRV is transmitted primarily through respiratory droplets and is shed at high levels in ocular and nasal secretions during the early infectious period, the stronger associations observed for coughing and ocular discharge in our study align with established transmission pathways and expected patterns of viral shedding (Couacy-Hymann et al., 2007; Parida et al., 2019; Taylor, 1984). Given that PPRV is shed most intensely through ocular and nasal secretions, the elevated odds observed for coughing and ocular discharge in our study are consistent with the biological expectation of where and when viral RNA is most detectable. This pattern reflects common challenges in PPR epidemiology, where symptom frequency is influenced by immunity, prior exposure, and virus lineage (Rahman et al., 2023).

Risk factor analysis identified potential associations with water source and grazing system; however, these patterns must be interpreted with caution. Although the data indicate measurable exposure and viral detection, the wide confidence intervals and the absence of key management variables such as herd size, stocking density, movement patterns, and hygiene suggest that unmeasured confounding may have influenced the observed effects. Goats accessing supply water source showed higher odds of PPR positivity. This may be due to communal or shared water points facilitating indirect transmission, as multiple households congregate in these areas, and viral contamination of surfaces or water sources has been reported in similar settings (O'Brien & Xagoraraki, 2019). Shared resources such as feed and water points may facilitate interactions among animals, increasing the potential for PPRV transmission within and between herds (Abubakar et al., 2016; Selim et al., 2025). Nonetheless, because our study did not include environmental or water-source sampling, we cannot determine whether supply water acted as a mechanical exposure point or played any direct role in PPRV transmission. A similar finding was observed in Sudan, India, and Mongolia, where animals sharing watering sources were significantly associated with PPR infection (Saeed et al., 2018; Ullah et al., 2022). Shared water points are recognized as important transmission pathways for PPRV in smallholder settings. Similarly, free-ranging goats were more likely to pose a potential risk for PPR than confined animals. This pattern is consistent with prior studies showing that free-ranging management increases inter-household contact, grazing overlap, and exposure to contaminated fomites or shared environments, thereby elevating PPRV transmission risk (Bwihangane et al., 2016; Mdetele et al., 2021). This finding supports earlier studies that identified free-grazing systems as a major risk factor for the spread of PPRV (Akwongo et al., 2022). However, unmeasured factors such as larger flock sizes, other household contact, variable nutritional status, or differences in owner management and herd-mixing practices may also influence infection risk and cannot be excluded.

This study provides region-specific epidemiological data that can inform future investigations. While the identified associations highlight potential management risks, particularly communal water use and free-ranging systems, further longitudinal and environmental studies will be necessary to substantiate these pathways before policy recommendations can be firmly established. Future research should incorporate environmental sampling, socio-economic factors, and cross-border livestock movement to better contextualize PPRV transmission within the PPR eradication framework.

4.1. Limitations

This study has several limitations that should be considered when interpreting the findings. First, information on management practices, herd history, and clinical signs was collected through farmer interviews, which may introduce recall bias or reporting bias, particularly regarding vaccination history, previous disease events, and hygiene practices. Second, the study was conducted using a cross-sectional design, which provides a snapshot of infection and exposure at a single time point. As a result, temporal relationships between exposure variables and infection cannot be established, and the observed associations should not be interpreted as causal relationships. Third, although goats were sampled from multiple households and villages, only 3–5 animals were sampled per household, which limited the ability to model hierarchical clustering at the household or village level using mixed-effects models. Consequently, a fixed-effects logistic regression approach was used. Fourth, sampling was conducted during a single seasonal period (April–July 2024), which may not capture potential seasonal variation in PPRV transmission across different climatic periods such as monsoon, winter, or dry seasons. Fifth, several potentially important epidemiological variables including herd size, animal movement patterns, market contact, and environmental contamination of shared water sources were not measured in this study. The absence of these variables introduces the possibility of unmeasured confounding, which may influence the observed associations. Finally, although both serological (cELISA) and molecular (RT-PCR) methods were used to detect exposure and infection, environmental sampling of water sources or shared grazing areas was not performed. Therefore, the study cannot determine whether these environmental factors directly contributed to viral transmission. Despite these limitations, the study provides valuable region-specific epidemiological insights into PPRV exposure and circulation in northeastern Bangladesh, contributing important evidence for future surveillance and control strategies.

5. Conclusion

This study confirms that peste des petits ruminants (PPR) remains firmly endemic in northeastern Bangladesh, with substantial evidence of both past exposure (35.05%) and active infection (16.6%); notably, at the household-level or herd level (n = 90), the molecular prevalence was 41.11% (37/90) and the seroprevalence was 75.56% (68/90), indicating widespread circulation across herds. Although management-related exposures (e.g., shared or supply water sources and free-ranging grazing) were statistically associated with PPRV infection, these husbandry practices largely reflect structural constraints in smallholder production systems and are unlikely to be easily modified at scale. Therefore, while such factors may help identify higher-risk settings and transmission opportunities, they should not be viewed as primary standalone interventions for disease control. Instead, our findings reinforce that achieving and sustaining high vaccination coverage remains the only realistic pathway toward PPR elimination and eradication. The identified risk factors are most valuable for improving risk-based vaccination strategies, optimizing the geographic and seasonal prioritization of campaigns, and guiding post-vaccination monitoring and surveillance to detect gaps in immunity and ongoing circulation. Strengthening resource mobilization for effective vaccination delivery, ensuring cold-chain integrity, increasing community engagement to reach underserved herds, and implementing systematic post-vaccination evaluation will be essential to reduce transmission and advance Bangladesh’s national PPR eradication plan in alignment with the global 2030 target.

Data availability statement

All data generated and analyzed in this study are included in the main manuscript.

Ethical statement

This study was performed in accordance with the ethical standards and guidelines set by the Ethical Review Committee at Sylhet Agricultural University, Sylhet-3100, Bangladesh, approved the research protocol #AUP2023012. All protocols involving animals were reviewed and approved to ensure ethical conduct throughout the study. Where applicable, informed consent was obtained from relevant stakeholders, and efforts were made to minimize harm and distress to all subjects involved in the research.

CRediT authorship contribution statement

Mohammed Abdul Kahir: Writing – review & editing, Writing – original draft, Methodology, Investigation, Formal analysis, Data curation. Tajul Islam Mamun: Writing – review & editing, Writing – original draft, Software, Investigation, Formal analysis, Data curation. Md. Abu Saeed: Writing – review & editing, Writing – original draft, Investigation, Data curation. Md. Khademul Islam: Writing – review & editing, Writing – original draft, Formal analysis, Data curation. Md. Irtija Ahsan: Writing – review & editing, Writing – original draft, Validation, Formal analysis, Data curation. Sharmin Akter: Writing – review & editing, Writing – original draft, Validation, Formal analysis. Suman Paul: Writing – review & editing, Writing – original draft, Validation, Supervision, Formal analysis. Md Bashir Uddin: Writing – review & editing, Writing – original draft, Supervision, Project administration, Methodology, Formal analysis. Syed Sayeem Uddin Ahmed: Writing – review & editing, Writing – original draft, Supervision, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.