Global Prevalence and Distribution of H9 Subtype of Avian Influenza Viruses in Wild Birds: Literature Review with Meta-Analysis

Abstract

Background:

As a natural accelerator of highly pathogenic avian influenza in wild birds, the H9 subtype of avian influenza poses a substantial threat to both humans and the poultry industry. A comprehensive meta-analysis is necessary to assess the current status of the global H9 outbreak. In this research, a literature review and meta-analysis are presented on the surveillance studies of the H9 subtype of avian influenza in wild birds worldwide up to 2024.

Methods:

A comprehensive search strategy was employed, utilizing the China Science and Technology Journal Database, China National Knowledge Infrastructure, PubMed, Google Scholar, and Scientific Direct databases. The exclusion criteria for this study included duplicate studies, reviews, other host studies, as well as research with inconsistent or insufficient data. An analysis was conducted on data obtained from a total of 31 publications. The rate-conversion analyses were conducted using a random-effects model in the “meta” package of the “R” software, with the PFT method implemented.

Results:

In the meta-analysis, the prevalence of wild bird H9 avian influenza virus (AIV) was found to be 0.02% (193 out of 365,972). Statistically significant higher prevalences of wild bird influenza A virus were observed in Norway and South Africa (0.87%, 21/2417 and 0.44%, 10/1155, respectively) in comparison with other regions. Within the Anseriformes family, the prevalence rate was much greater (0.17%, 80 out of 90,014) compared with other species. In addition, we performed subgroup analyses that included geographical variables. These assessments showed a higher prevalence of H9 in wild birds in cold regions (0.08%, 30/100,691).

Conclusion:

In summary, our results suggest that the occurrence of H9 AIV in avian populations differs among different geographical areas and species. Therefore, it is necessary to conduct further surveillance on the prevalence of AIV in wild birds to guide the creation of strong and efficient regulatory strategies targeted at eradicating the transmission of AIV across different species.

Background

Comprising members of the Influenza A family, the avian influenza virus (AIV) is an RNA virus of the Orthomyxoviridae family (Spackman, 2008). Influenza A is classified into subtypes based on hemagglutinin (H) and neuraminidase (N) glycoprotein. Subtypes 16H subtypes and 9N have been found in the natural bird reservoir (Fouchier et al., 2005). Together, these subtypes create several influenza viruses including H5N8, H7N9, and H9N2. Pathogenicity further divides AIV into low pathogenic avian influenza (LPAI) and highly pathogenic avian influenza (HPAI) (Peacock et al., 2019), with LPAI derived from various hemagglutinin subtypes and H5 and H7 subtypes largely driving HPAI (Spackman, 2008).

H9 is an LPAI subtype (Lloren et al., 2017). Originally found in American turkeys in 1966 (Homme and Easterday, 1970) and was subsequently isolated in wild birds across continental Europe during the subsequent decades (Shortridge, 1992). The H9 subtype is prevalent in poultry populations in some countries in Asia, the Middle East, and Africa (Fusaro et al., 2024). Unlike the highly pathogenic H5 and H7 subtypes of AIVs, H9 viruses have not been prioritized by disease control agencies in many countries (Zhang et al., 2023). Although the H9 subtype has received little attention due to its perceived low pathogenicity, the threat of H9 avian influenza should not be underestimated. When the host is infected with multiple AIVs at the same time, the internal genes of the virus can reassort, resulting in a new influenza virus with pandemic potential (Wen et al., 2024). In addition, the A/H9N2 strains pose an increasing threat to poultry and humans. Mixed infections caused by the A/H9N2 virus are associated with bacteria and other viruses and cause significant economic losses to the poultry industry in the form of reduced egg production and increased mortality (Bi et al., 2022; Fusaro et al., 2024). These economic losses have led some countries to adopt poultry vaccination strategies, which have further promoted the genetic diversity of H9 AIV (Zhang et al., 2023). Indeed, not only birds, multiple human infections with H9N2 have been reported in recent years (Quan et al., 2019), and as of December 1, 2023, H9N2 has been linked to 128 human infections, 90% of which have been reported in China (Fusaro et al., 2024). More than half of the cases are from 2020 to 2023. Human H9N2 infections are usually mild and only one death from the virus has been reported, possibly due to underlying health conditions (Peacock et al., 2019). The vast majority of these human cases have been confirmed to have had contact with poultry and no evidence of human-to-human transmission has been reported (Cáceres et al., 2021).

Migratory birds and wild birds are thought to be major hosts and carriers of AIV including the H9 subtype (Webster et al., 1992). Wild birds play a significant role in the spread of avian influenza from the poultry-wild bird-poultry interface. Many wild birds have been shown to carry virus, which they then pass on to other birds (Wang et al., 2012). In South Africa, the European spectrum of the H9N2 virus has been isolated from farmed ostriches. It appears that these cases most likely represent dead-end spillover events from wild migratory birds (Peacock et al., 2019). Another important factor thought to be responsible for the spread of H9N2 is live bird markets (LBMs), where high densities of poultry species provide ideal conditions for the spread of the virus, and LBMs were identified early on as a major source of AIV outbreaks in mainland China and Hong Kong in the late 1990s (Peacock et al., 2019). The World Health Organization (WHO) and national health agencies recommend that surveillance should be strengthened in poultry farming intensive areas where the virus is highly prevalent to prevent further spread of the virus. Many nations have started monitoring avian influenza A, including wild bird populations with the H9 subtype (Ge et al., 2018). Accurately describing the epidemiology of H9 subtype AIV in wild bird populations is difficult, nevertheless, given the sometimes scattered and difficult to compare data generated from these studies.

Meta-analysis provides a strong statistical approach for combining data from several research, therefore generating more accurate estimates of the pathogen’s epidemiological traits (Hernandez et al., 2020). Our aim in this meta-analysis is to methodically review the current body of research on the global prevalence and distribution of avian influenza A subtype H9 in wild birds. We will also examine the factors that contribute to the spread of H9 subtypes AIV in wild bird populations.

This study will offer a more complete knowledge of the global prevalence of H9 AIV in wild birds by summarizing available data. It will also help to pinpoint important knowledge gaps that must be filled in order to enhance control and preventative plans for this important pathogen. In addition, this study will provide a more comprehensive understanding by assessing the effects of geographical region, season, bird species and detection strategies on the incidence of H9 AIV. Eventually, the results will guide the creation of sensible preventive and control strategies for the H9 subtype AIV.

Methods

This study followed the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Protocols (Page et al., 2021) for conducting a systematic review and meta-analysis (Supplementary Table S1).

Search strategy

A systematic literature search was conducted to identify studies concerning the occurrence and spread of the H9 subtype of AIV in wild birds. PubMed, Science Direct, China National Knowledge Infrastructure, Google Scholar, and China Science and Technology Journal Database were searched for published articles in English or Chinese. Included Chinese articles should have at least an English abstract. The last search time was January 6, 2025. Keywords included, “AIV,” “wild bird,” “migratory bird,” “prevalence,” and “surveillance.”

Inclusion and exclusion criteria

The inclusion criteria for the articles meeting this study were as follows: (1) the articles must be investigations or surveillance studies evaluating the occurrence and/or transmission of H9 subtype influenza virus in wild birds. (2) The study object must be wild birds. (3) The total number of wild birds investigated and the number of positives in the article must be provided. (4) Data must come from active or passive collection of field populations.

The articles were excluded if any of the following cases applied: (1) The full text is not available. (2) No data are provided in the article. (3) The article contains incomplete data or only provides the infection rate. (4) The article contains duplicate data. (5) The article contains conflicting data. (6) The article contains a sample size of less than 25. (7) The article contains mixed samples. (Samples from different birds were mixed and individual samples were not tested after the mixed samples tested positive.)

Data extraction

Two independent reviewers extracted and recorded the following data from the eligible literature: publication year, sampling period, sample location, detection method, season, country, continent, country classification, sample type, wild bird species, thorough geographic and climatic information, total number of samples, and the number of AIV-positive cases of subtype H9 were extracted from the eligible studies. The extracted data were summarized into a Word Processing System (WPS) table for analysis. Any disagreements need to be discussed and resolved by the two reviewers.

Quality evaluation

The data were extracted from the selected articles using a standardized data collection form, in accordance with the purpose and inclusion criteria of this study. The data scoring criteria were as follows: (1) clear sampling location, (2) sample size greater than 50, (3) clear detection species, (4) clear detection season, and (5) inclusion five geographic and climatic risk factors (longitude, latitude, altitude, precipitation, temperature). In accordance with the aforementioned scoring criteria, each item is assigned a value of one point, with the total score for each item subsequently added to the total score for the article.

Statistical analysis

The data collected were analyzed using the “meta” package (release 4.12) of the R software (release 4.1.2) (Viechtbauer, 2010). Four data transformation methods were tested to obtain a closer approximation to a Gaussian distribution: the original exchange rate (PRAW), the logarithmic transformation (PLN), the logit transformation (PLOGIT), the sine transformation (PAS), and the double sine transformation (PFT). The PFT was selected based on the criterion that the W value was close to 1 (W = 0.86002; Table 1). The different transformations of the normal and arcsine exchange rates for the normal distribution were examined by choosing the appropriate effect model (Ni et al., 2020).

Table 1.

Normal Distribution Tests for Normal Rates and Different Transitions

Conversion form	W	p
PRAW	0.58479	3.401e–08
PLN	NaN	NA
PLOGIT	NaN	NA
PAS	0.85169	0.0005506
PFT	0.86002	0.000843

Bold text indicates the final selected/adopted data.

“PRAW”: raw exchange rate; PLN: log conversion. “PLOGIT”: logit transformation.

“PAS”: arcsine transformation; “PFT”: double arcsine transformation.

“NaN”: meaningless number; NA*: data are missing.

The Cochran-Q statistic and I² statistics yielded significant heterogeneity. If the I² value was >50%, the p value was <0.05, whereas if the I² value was <50% and the p value was >0.1, the heterogeneity between the individual studies was visualized by means of forest plots. A random-effects model was used because the assessment results showed significant heterogeneity (I² ≥ 90.8%). The presence of publication bias was determined by the symmetry of the funnel plot, and the asymmetry of the funnel plot was assessed by visual observation and was verified using Egger’s test and trim-fill test. In addition, a sensitivity analysis was performed for the assessment of the reliability of the results of the meta-analysis.

A statistical analysis was conducted to determine the prevalence of AIV H9. Subgroup analyses were then employed to explore sources of heterogeneity among the included studies and to predict factors contributing to heterogeneity. Factors assessed included sampling time, sample type, assay method, species, country, season, and score. Furthermore, to investigate whether geographical factors are associated with sources of heterogeneity, factors including latitude, altitude, precipitation, and temperature have been evaluated.

Results

Search results and qualified publications

Originally found were 4512 studies by searching five databases using a mix of pertinent keywords. Excluded were 1625 duplicate studies and 2887 irrelevant articles upon evaluation of the titles and abstracts. After a full-text review, 89 articles were eliminated as they did not fit the screening criteria. Therefore, a final collection of 31 articles were fit for inclusion (Abolnik et al., 2022; Barkhasbaatar et al., 2023; El Zowalaty et al., 2022; Espano et al., 2023; Ferro et al., 2010; Fuller et al., 2015; Ge et al., 2018; Gomes et al., 2023; Hansbro et al., 2010; Henriques et al., 2011; Kang et al., 2010; Kang et al., 2024; Karmacharya et al., 2015; Kim et al., 2019; Krauss et al., 2013; Lee et al., 2017; Muzyka et al., 2012; Na et al., 2022; Nam et al., 2021; Ni et al., 2015; Piaggio et al., 2012; Ramey et al., 2016; Ruiz et al., 2023; Shi et al., 2014; Tang et al., 2020; Tønnessen et al., 2013; Verhagen et al., 2017; Wang et al., 2021; Wang et al., 2023; Yang, 2020; Yao et al., 2022) (Fig. 1; Table 2).

FIG. 1.

Flow diagram of literature search and selection.

Table 2.

Studies in the Analysis

Study ID	Published year	Country	Test method	Event	Positive rate (%)
Hansbro et al. (2010)	2010	Australia	rRT-PCR*	16/21,858	0.07%
Kang et al. (2010)	2010	Korea	RT-PCR*	9/28,214	0.03%
Ferro et al. (2010)	2010	United States	rRT-PCR	0/5363	0.00%
Henriques et al. (2011)	2011	Portugal	RT-PCR	9/5691	0.16%
Muzyka et al. (2012)	2012	Ukraine	RT-PCR	3/3857	0.08%
Piaggio et al. (2012)	2012	United States	rRT-PCR	6/75,532	0.01%
Krauss et al. (2013)	2013	Canada	RT-PCR	2/1240	0.16%
Tønnessen et al. (2013)	2013	Norway	rRT-PCR	21/2417	0.87%
Shi et al. (2014)	2014	China	rRT-PCR	22/4365	0.50%
Fuller et al. (2015)	2015	Benin	rRT-PCR	0/830	0.00%
Karmacharya et al. (2015)	2015	Nepal	rRT-PCR	1/1811	0.06%
Ni et al. (2015)	2015	China	rRT-PCR	0/34	0.00%
Ramey et al. (2016)	2016	United States	RT-PCR	1/1129	0.09%
Lee et al. (2017)	2017	Korea	RT-PCR	7/31,146	0.02%
Verhagen et al. (2017)	2017	Netherlands	RT-PCR	5/68,637	0.01%
Ge et al. (2018)	2018	China	RT-PCR	8/4534	0.18%
Kim et al. (2019)	2019	Korea	rRT-PCR	2/11,145	0.02%
Tang et al. (2020)	2020	China	rRT-PCR	9/5112	0.18%
Yang (2020)	2020	China	RT-PCR	1/9185	0.01%
Nam et al. (2021)	2021	Korea	RT-PCR	3/7942	0.04%
Wang et al. (2021)	2021	China	RT-PCR	11/8594	0.13%
Abolnik et al. (2022)	2022	South Africa	rRT-PCR	10/1155	0.87%
El Zowalaty et al. (2022)	2022	South Africa	rRT-PCR	0/48	0.00%
Yao et al. (2022)	2022	China	RT-PCR	29/10,910	0.27%
Na et al. (2022)	2022	Korea	rRT-PCR	3//7590	0.04%
Barkhasbaatar et al. (2023)	2023	Mongolia	RT-PCR	0/10,222	0.00%
Espano et al. (2023)	2023	Korea	RT-PCR	0/6758	0.00%
Gomes et al. (2023)	2023	NA	rRT-PCR	0/254	0.00%
Ruiz et al. (2023)	2023	Chile	rRT-PCR	2/4349	0.05%
Wang et al. (2023)	2023	China	RT-PCR	13/15,899	0.08%
Kang et al. (2024)	2024	Mongolia	rRT-PCR	0/10,149	0.00%

rRT-PCR*: real-time reverse transcription polymerase chain reaction.

RT-PCR: reverse transcription polymerase chain reaction.

Study characteristics

Of the 31 studies that met the criteria for meta-analysis of the prevalence of H9 AIV infection in wild birds, 13 trials reported sampling periods prior to 2013, 13 trials reported sampling periods between 2013 and 2018, and 7 trials reported sampling periods greater than 2018. Sample types included feces (21 trials), cloacal swabs (7 trials), cloacal and oropharyngeal swabs (4 trials), and oropharyngeal swabs (1 trials). Sixteen studies used real-time reverse transcription polymerase chain reaction (rRT-PCR) as the detection method and 15 studies used RT-PCR. Because the number of bird species was scattered and excessive, this study grouped bird species into Anseriformes (11 trials), Charadriiformes (5 trials), and others (28 trials). Eight trials reported that the sampling season was summer, 14 trials reported that the sampling season was winter, and another 12 studies did not report the sampling season. Fourteen studies were conducted in advanced economies, 16 studies were conducted in developing economies, and 1 study was conducted in Antarctica and was not applicable to national development situations. In addition, according to study scores, 1 study scored low (2 points), 22 studies scored medium (3–4 points), and 8 studies scored high (5 points).

Subgroup analyses revealed that 26 trials were conducted in the northern hemisphere and 4 trials were conducted in the southern hemisphere. Twenty-four trials were conducted in the eastern hemisphere and six trials were conducted in the western hemisphere. The altitudes reported in the trials were categorized as <200 m in 22 instances, 200–1000 m in 3 instances, and >1000 m in 5 instances. There were 10 trials classified as semiarid, 12 trials classified as humid, and 7 trials classified as hyper-humid. Two trials indicated hot regions, whereas warm regions were reported in 22 trials, and cold regions were reported in 6 trials. Thirty studies provided the required geographic details, while the remaining study lacked specific sampling site information.

Publication bias and sensitivity analysis of the publications

The forest plot exposed very significant variation among the selected research (Fig. 2). The funnel plot shows that the points representing the data are concentrated at the top of the funnel and are skewed to one side. Points concentrated at the top of the funnel usually represent studies with larger sample sizes. These studies tend to be more likely to be published, whereas studies with small sample sizes and nonsignificant effects may not be published, suggesting that there may be a publication bias that overestimates the overall effect size (Fig. 3).

FIG. 2.

Forest map of the global H9 subtype of avian influenza viruses.

FIG. 3.

Funnel plot with pseudo 95% CI for publication bias test. CI, confidence interval.

We then ran a trim-and-fill study and an Egger’s test to assess this more closely. The results of Egger’s test confirmed the existence of publication bias in the selected papers (p = 0.0011 and p ≤ 0.05; Fig. 4). An iterative approach was used to estimate the number of missing studies in order to test the hypothesis that publication bias adds to the asymmetry in the funnel plot (Duval and Tweedie, 2000). These 12 extra dummy studies were then included into a fresh meta-analysis, producing an adjusted p value of 0.000 (Fig. 5). The sensitivity research showed that no one study had any effect on the whole data, thereby verifying the dependability and strength of our results (Fig. 6).

FIG. 4.

Egger’s test for publication bias.

FIG. 5.

Shear complement graph and pseudo 95% CI publication bias test.

FIG. 6.

Sensitivity analysis.

Subgroup analysis results

The subgroup analysis performed throughout the study period showed that the prevalence rate in the rRT-PCR subgroup was 0.09% (95% confidence interval [CI]: 0.02–0.21%), whereas the prevalence rate reported by RT-PCR as a test was 0.06% (95% CI: 0.02–0.10%). In addition, the rate of H9 detection in cloacal swabs was 0.17% (95% CI: 0.02–0.49%), surpassing the rates of 0.00% (95% CI: 0.00–3.95) recorded in oropharyngeal swabs, 0.06% (95% CI:0.02–0.11) in fecal samples, and 0.04% (95% CI: 0.00–0.16) in cloacal and oropharyngeal swabs. An examination of subgroups based on species showed that Anseriformes had the greatest prevalence rate of 0.17% (95% CI: 0.05–0.36%). A greater seasonal prevalence of the disease was seen in summer (0.19%, 95% CI: 0.00–0.98%) compared with winter (0.10%, 95% CI: 0.02–0.21%). The prevalence of H9 AIVs in wild birds was shown to be 0.07% (95% CI: 0.01–0.18) in the group before 2013, 0.04% (95% CI: 0.01–0.08) in the group from 2013 to 2018, and 0.07% (95% CI: 0.00–0.27) in the group after 2018 (Table 3). The data demonstrate a declining pattern, succeeded by an ascending one. A greater prevalence rate of 0.05% (95% CI: 0.01–0.12) was seen in the advanced economies group compared with the developing economies group (0.11%, 95% CI: 0.03–0.27) for the national circumstances subgroup. Wildfowl AIV had a statistically significant higher cumulative prevalence in Norway and South Africa (0.87% and 0.44%, respectively) compared with other regions (Table 4).

Table 3.

Results of Subgroup Analysis

Variable	Category	No. studies	No. examined	No. positive	% (95% CI*)	Heterogeneity			Univariate meta-regression
Variable	Category	No. studies	No. examined	No. positive	% (95% CI*)	χ2	p value	I² (%)	p value*	Coefficient (95% CI)
Detection method	rRT-PCR	16	152,012	92	0.09% (0.02–0.21)	196.28	<0.01	92.4	<0.0001	0.0294 (0.0166–0.0422)
	RT-PCR	15	213,960	101	0.06% (0.02–0.10)	142.72	<0.01	90.2
Sample type	Fecal	21	270,967	138	0.06% (0.02–0.11)	216.26	<0.01	90.8	0.2022	0.0208 (−0.0112–0.0527)
	Cloacal swab	7	19,273	40	0.17% (0.02–0.49)	60.80	<0.01	90.1
	Cloacal and oropharyngeal swab	4	75,708	15	0.04% (0.00–0.16)	23.27	<0.01	87.1
	Oropharyngeal swab	1	24	0	0.00% (0.00–3.95)	0.00	—	—
Species	Anseriformes	11	90,014	80	0.17% (0.05–0.36)	186.40	<0.01	94.6	<0.0001	0.0404 (0.0256–0.0544)
	Charadriiformes	5	17,687	4	0.01% (0.00–0.04)	7.68	0.10	47.9
	Other birds	28	255,595	108	0.03% (0.01–0.07)	173.49	<0.01	84.4
Season	Summer	8	20,205	28	0.19% (0.00–0.98)	59.44	<0.01	88.2	0.0138	0.0381 (0.0078–0.0684)
	Winter	14	79,475	101	0.10% (0.02–0.21)	189.57	<0.01	93.1
Quality	Low	1	254	0	0.00% (0.00–0.38)	0.00	—	—	<0.0001	0.0382 (0.0223–0.0542)
	Medium	22	314,027	118	0.05% (0.02–0.09)	195.15	<0.01	89.2
	High	8	51,691	75	0.15% (0.03–0.35)	93.86	<0.01	92.5
Sample year	Before 2013	13	232,973	96	0.07% (0.01–0.18)	182.64	<0.01	93.4	0.0059	0.0188 (0.0054–0.0322)
	2013–2018	13	86,515	41	0.04% (0.01–0.08)	51.10	<0.01	76.5
	After 2018	7	39,444	28	0.07% (0.00–0.27)	46.17	<0.01	87.0
National circumstances	Advanced economies	14	274,664	84	0.05% (0.01–0.12)	124.74	<0.01	89.6	0.0010	0.0228 (0.0092–0.0364)
	Developing economies	16	90,130	109	0.11% (0.03–0.27)	175.32	<0.01	91.4

Bold text indicates statistically significant values (P < 0.05).

CI^*, confidence interval; –^*not applicable; p value: ^* p ≤ 0.05 is statistically significant.

Quality: High: 5 points; Medium: 3 or 4; Low: 2.

Table 4.

Concentrated Prevalence Rate of H9 Subtype of Avian Influenza Viruses in Different Countries

Variable	Category	No. studies	No. examined	No. positive	% (95% CI^*)
Variable	Category	No. studies	No. examined	No. positive	Country	Australia	1	21,858	16	0.07% (0.04–0.11)
	Benin	1	830	0	0.00% (0.00–0.21)
	Canada	1	1240	2	0.16% (0.00–0.49)
	Chile	1	4349	2	0.05% (0.00–0.14)
	China	8	58,633	93	0.03% (0.00–0.14)
	Korea	6	82,795	24	0.02% (0.01–0.04)
	Mongolia	2	20,371	0	0.00% (0.00–0.01)
	Nepal	1	1811	1	0.06% (0.00–0.24)
	Netherlands	1	68,637	5	0.01% (0.00–0.02)
	Norway	1	2417	21	0.87% (0.53–1.28)
	Portugal	1	5691	9	0.16% (0.07–0.28)
	South Africa	2	1213	10	0.44% (0.07–1.02)
	Ukraine	1	3857	3	0.08% (0.01–0.20)
	United States	3	82,024	7	0.00% (0.00–0.00)

CI, confidence interval.

In relation to geographical variables, the prevalence rates were seen to be greater in the southern hemisphere (0.04%, 95% CI: 0.00–0.53) compared with the northern hemisphere (0.02%, 95% CI: 0.00–0.07). Moreover, the observed prevalence rates were greater in regions with elevations between 200 and 1,000 m (0.04%, 95% CI: 0.00–0.16) compared with regions with elevations above 1,000 m (0.00%, 95% CI: 0.00–0.32) and below 200 m (0.03%, 95% CI: 0.00–0.09). Concerning temperature, no cases of H9 subtype prevalence were found in hot regions, while cold and warm regions had prevalence rates of 0.08% (95% CI: 0.00–0.33) and 0.01% (95% CI: 0.00–0.06), respectively (Table 5).

Table 5.

A Subgroup Analysis of the Prevalence of H9 by Geographical Location

Variable	Category	No. studies	No. examined	No. positive	% (95% CI*)	Heterogeneity			Univariate meta-regression
Variable	Category	No. studies	No. examined	No. positive	% (95% CI*)	χ2	p value	I²(%)	p value *	Coefficient (95% CI)
Hemispheres	Southern Hemisphere	4	27,410	28	0.04% (0.00–0.53)	21.33	<0.01	85.9	<0.0001	0.0279 (0.0190–0.0368)
	Northern Hemisphere	26	338,308	165	0.02% (0.00–0.07)	290.88	<0.01	91.4
	Eastern Hemisphere	24	272,412	173	0.02% (0.00–0.07)	278.24	<0.01	91.7	<0.0001	0.0308 (0.0213–0.0404)
	Western Hemisphere	6	93,306	20	0.04% (0.00–0.12)	31.99	<0.01	84.4
Altitude	<200 m	22	243,701	158	0.03% (0.00–0.09)	241.87	<0.01	91.3	<0.0001	0.0304 (0.0204–0.0404)
	200–1000 m	3	98,632	24	0.04% (0.00–0.16)	27.27	<0.01	92.7
	>1000 m	5	23,385	11	0.00% (0.00–0.32)	38.81	<0.01	89.7
Precipitation	Semi-arid	10	209,680	56	0.05% (0.00–0.15)	92.00	<0.01	90.2	<0.0001	0.0334 (0.0193–0.0475)
	Humid	12	63,201	90	0.03% (0.00–0.12)	98.80	<0.01	88.9
	Hyper-humid	7	88,488	45	0.00% (0.00–0.12)	58.14	<0.01	89.7
Temp	Cold regions	6	100,691	30	0.08% (0.00–0.33)	79.61	<0.01	93.7	0.0020	0.0298 (0.0109–0.0408)
	Warm regions	22	258,834	163	0.01% (0.00–0.06)	214.28	<0.01	90.2
	Hot regions	2	6193	0	0.00% (0.00–0.02)	0.32	0.57	0.0

Bold text indicates statistically significant values (P < 0.05).

CI*, confidence interval; p value*, p ≤ 0.05 is statistically significant.

Discussion

Data from 365,972 wild birds from 31 articles were evaluated in this meta-analysis, resulting in prevalence of 0.02% (193/365,972). The results showed that H9 subtype avian influenza was widely distributed around the world, and the highest detection rate was 0.87% (95% CI: 0.53–1.28) in Norway and 0.44% (95% CI: 0.07–1.02) in South Africa. This coincides with the trend of circulating H9 in African poultry (Fusaro et al., 2024), which may be due to the less intensive rearing of African poultry and more opportunities for exposure to wild birds, resulting in virus spillover. Norway is situated on major migratory routes for migratory birds and has a large number of wetlands and coastal areas that provide rich habitats and food sources for wild birds. This may create favorable conditions for the concentration of wild bird populations and the spread and circulation of virus. Geographically, the prevalence of AIV at an altitude of 200–1000 m was significantly higher than that at an altitude of more than 1000 m. Areas of 200–1000 m above sea level are often a confluence of wetlands, forests, and farmland, and many waterfowl and landbirds tend to be active at this altitude, with important correlations between viral activity and wild bird ecology.

We observed the highest prevalence in the cold region classification, which is consistent with previous findings that cold regions favor the preservation of avian influenza viral RNA and longer maintenance of infection (Meng et al., 2022; Sarah et al., 2018). Several investigations have proven that AIVs exhibit increased survival rates at colder temperatures (Zhang et al., 2014). For instance, many viruses maintain their activity for almost 200 days when exposed to a temperature of 4°C (Stallknecht et al., 1990). Moreover, there have been documented occurrences of AIV outbreaks subsequent to a decrease in temperature (Liu et al., 2007). Investigations conducted by Liu et al. have also verified that the occurrence of avian influenza increases when the temperature decreases (Liu et al., 2018).

Our study also found that the positive rate of studies using rRT-PCR was higher than that of RT-PCR, and the detection sensitivity of rRT-PCR was speculated to be higher. rRT-PCR enables real-time monitoring of changes in the fluorescence signal during amplification, allowing precise quantification of the target RNA. This real-time monitoring allows even small amounts of early target RNA to be detected.

Moreover, the design of rRT-PCR usually uses fluorescent probes or dyes, which makes it more sensitive for the detection of low-abundance RNA and can detect less starting template. However, the heterogeneity among the studies was high, and there are still many potential factors for heterogeneity waiting to be explored.

Regarding the sampling time, we found that the positive rate increased significantly after 2018. In China, H9N2 has gradually become the main subtype of AIV in China’s live poultry markets since 2016 (Bi et al., 2020), which may have contributed to the increased infection rate of H9 subtype avian influenza in wild birds. The prevalence of H9 has increased over time, raising concerns that the H9 subtype of avian influenza could become a pandemic. Routine influenza surveillance in China during 2019–2020 showed that the hemagglutinin genes of H9N2 viruses did not cluster together, suggesting recent evolution in the genomes of these H9N2 viruses (Bi et al., 2022; Zhang et al., 2023). The pandemic potential of H9N2 is reflected in the continuous emergence of human infectious influenza viruses with H9N2 internal genes (Wen et al., 2024), demonstrating the urgent need to prevent the generation of new human infectious AIVs at their source or to control the ability of migratory birds to spread viruses across states.

In addition, it is also possible that vaccination may have promoted the evolution of H9 virus and contributed to the increased infection rate in recent years. China (Zhang et al., 2008), Israel (Banet-Noach et al., 2007), South Korea (Lee and Song, 2013), and Morocco (El Houadfi et al., 2016) began to use vaccination as the main means to avoid huge economic losses caused by H9N2 virus infection. Widespread vaccination also contributed to antigenic drift of the H9 virus (Yan et al., 2022), which also contributed to the poor protective efficacy of the vaccine in the immunized chicken flocks (Okamatsu et al., 2008; Sun et al., 2012).

Consistent with past studies (Dugan et al., 2008), the systematic review found Anseriformes as the main natural hosts and transmitters of H9 avian influenza. Our study showed that Anseriformes had far more AIV frequency than non-Anseriformes, implying that Anseriformes are more likely to be avian influenza vulnerable. Researchers should thus give the frequency of AIVs and antibodies in this population top priority. Anseriformes exhibited a greater incidence of AIVs in comparison with nonwaterfowl, therefore providing additional evidence for the notion that Anseriformes possess greater sensitivity than other wild birds.

In the remaining subgroup analyses, the prevalence was significantly higher in summer than in winter, which was not consistent with the conclusions of many studies. We considered that the reasons for this were the large heterogeneity between groups and the fact that many studies did not report the sampling season, which may affect the reliability of this subgroup analysis. In addition, our finding that H9 is more prevalent in wild birds in developing economies than in developed economies is also puzzling, for which we have yet to find a plausible explanation. Sample type was not significantly associated with H9 AIV infection in wild birds (p ≥ 0.05), so it will not be discussed.

Acknowledging specific constraints in this meta-analysis is paramount. First and foremost, the rather limited sample size could affect the dependability and consistency of the findings. The limited generalizability of the small sample size could so influence the results. Despite our efforts to collect the literature on the prevalence of H9 in wild birds, there are still many risk factors that have not been reported in the literature, which affect the accuracy of the results to a certain extent. Furthermore, comprehensive knowledge on H9 strains in wild bird populations was lacking, hence preventing the building of a phylogenetic tree and so compromising our knowledge of strain distribution over several geographical areas. The literature incorporated in this meta-analysis primarily originates from the Northern Hemisphere, which introduces a possible regional bias that could direct the conclusions toward conditions specific to the Northern Hemisphere.

Conclusion

To summarize, this study conducted a comprehensive meta-analysis to assess the worldwide occurrence of H9 AIV. H9 AIV has shown a rising prevalence over time. To mitigate the risk of future viral outbreaks, it is advisable to implement enhanced immunization programs and maintain ongoing surveillance.

Footnotes

Authors’ Contributions

W.-X.T.: Data curation, methodology, supervision, and writing—review and editing. S.-Y.Q. and X.Y.: Writing—review and editing. X.-M.L.: Data curation, resources, and software. J.-H.L. and H.C.: Data curation, resources, and software. J.-J. and H.-T.S.: Data curation, methodology, and visualization. Q.Z.: Conceptualization, supervision, and funding acquisition.

Data Availability

Data will be made available on request.

Author Disclosure Statement

The authors declare that they have no competing interests

Funding Information

This work was supported by the National Key Research and Development Program of China (2022YFC2303803).

Supplementary Material

Supplementary Table S1

References

Abolnik

, Phiri

, van der Zel

, et al. Wild Bird Surveillance in the Gauteng Province of South Africa during the High-Risk Period for Highly Pathogenic Avian Influenza Virus Introduction. Viruses, 2022; 14(9); doi: 10.3390/v14092027

Banet-Noach

, Perk

, Simanov

, et al. H9N2 influenza viruses from Israeli poultry: A five-year outbreak. Avian Dis, 2007; 51(Suppl 1):290–296; doi: 10.1637/7590-040206r1.1

Barkhasbaatar

, Gilbert

, Fine

, et al. Ecological characterization of 175 low-pathogenicity avian influenza viruses isolated from wild birds in Mongolia, 2009-2013 and 2016-2018. Vet Med Sci, 2023; 9(6):2676–2685; doi: 10.1002/vms3.1281

, Li

, et al. Dominant subtype switch in avian influenza viruses during 2016-2019 in China. Nat Commun, 2020; 11(1):5909; doi: 10.1038/s41467-020-19671-3

, Li

, Shi

. The time is now: A call to contain H9N2 avian influenza viruses. Lancet Microbe, 2022; 3(11):e804–e805; doi: 10.1016/s2666-5247(22)00232-4

Cáceres

, Rajao

, Perez

. Airborne transmission of avian origin H9N2 Influenza A Viruses in Mammals. Viruses, 2021; 13(10):1919.

Dugan

, Chen

, Spiro

, et al. The evolutionary genetics and emergence of avian influenza viruses in wild birds. PLoS Pathog, 2008; 4(5):e1000076; doi: 10.1371/journal.ppat.1000076

Duval

, Tweedie

. Trim and Fill: A Simple Funnel-Plot–Based Method of Testing and Adjusting for Publication Bias in Meta-Analysis. Biometrics, 2000; 56(2):455–463; doi: 10.1111/j.0006-341X.2000.00455.x

El Houadfi

, Fellahi

, Nassik

, et al. First outbreaks and phylogenetic analyses of avian influenza H9N2 viruses isolated from poultry flocks in Morocco. Virol J, 2016; 13(1):140; doi: 10.1186/s12985-016-0596-1

10.

El Zowalaty

, DeBeauchmp

, Jeevan

, et al. Molecular detection of influenza A viruses and H5 subtype among migratory Amur falcons (Falco amurensis) and captive birds of prey. Transbound Emerg Dis, 2022; 69(2):369–377; doi: 10.1111/tbed.13988

11.

Espano

, Shim

, Song

, et al. Surveillance of avian influenza viruses from 2014 to 2018 in South Korea. Sci Rep, 2023; 13(1):8410; doi: 10.1038/s41598-023-35365-4

12.

Ferro

, Budke

, Peterson

, et al. Multiyear surveillance for avian influenza virus in waterfowl from wintering grounds, Texas coast, USA. Emerg Infect Dis, 2010; 16(8):1224–1230; doi: 10.3201/eid1608.091864

13.

Fouchier

, Munster

, Wallensten

, et al. Characterization of a novel influenza A virus hemagglutinin subtype (H16) obtained from black-headed gulls. J Virol, 2005; 79(5):2814–2822; doi: 10.1128/jvi.79.5.2814-2822.2005

14.

Fuller

, Ducatez

, Njabo

, et al. Avian influenza surveillance in Central and West Africa, 2010-2014. Epidemiol Infect, 2015; 143(10):2205–2212; doi: 10.1017/S0950268814003586

15.

Fusaro

, Pu

, Zhou

, et al.; International H9 Evolution Consortium. Proposal for a Global Classification and Nomenclature System for A/H9 Influenza Viruses. Emerg Infect Dis, 2024; 30(8):1–13; doi: 10.3201/eid3008.231176

16.

, Yao

, Chai

, et al. A 627K variant in the PB2 protein of H9 subtype influenza virus in wild birds. Influenza Other Respir Viruses, 2018; 12(6):728–741; doi: 10.1111/irv.12592

17.

Gomes

, Prado

, Degrave

, et al. Active surveillance for influenza virus and coronavirus infection in Antarctic birds and mammals in environmental fecal samples, South Shetland Islands. An Acad Bras Cienc, 2023; 95(Suppl 3):e20230741; doi: 10.1590/0001-3765202320230741

18.

Hansbro

, Warner

, Tracey

, et al. Surveillance and analysis of avian influenza viruses, Australia. Emerg Infect Dis, 2010; 16(12):1896–1904; doi: 10.3201/eid1612.100776

19.

Henriques

, Fagulha

, Barros

, et al. Multiyear surveillance of influenza A virus in wild birds in Portugal. Avian Pathol, 2011; 40(6):597–602; doi: 10.1080/03079457.2011.618943

20.

Hernandez

, Marti

, Roman

. Meta-Analysis. Chest, 2020; 158(1S):S97–S102; doi: 10.1016/j.chest.2020.03.003

21.

Homme

, Easterday

. Avian influenza virus infections. I. Characteristics of influenza A-turkey-Wisconsin-1966 virus. Avian Dis, 1970; 14(1):66–74.

22.

Kang

, Jeong

, Kim

, et al. Surveillance of avian influenza virus in wild bird fecal samples from South Korea, 2003-2008. J Wildl Dis, 2010; 46(3):878–888; doi: 10.7589/0090-3558-46.3.878

23.

Kang

, Tseren Ochir

, Heo

, et al. Surveillance and Genetic Analysis of Low-Pathogenicity Avian Influenza Viruses Isolated from Feces of Wild Birds in Mongolia, 2021 to 2023. Animals (Basel), 2024; 14(7); doi: 10.3390/ani14071105

24.

Karmacharya

, Manandhar

, Sharma

, et al. Surveillance of Influenza A Virus and Its Subtypes in Migratory Wild Birds of Nepal. PLoS One, 2015; 10(7):e0133035; doi: 10.1371/journal.pone.0133035

25.

Kim

, Kim

, Son

, et al. The difference of detection rate of avian influenza virus in the wild bird surveillance using various methods. J Vet Sci, 2019; 20(5):e56; doi: 10.4142/jvs.2019.20.e56

26.

Krauss

, Pryor

, Raven

, et al. Respiratory tract versus cloacal sampling of migratory ducks for influenza A viruses: Are both ends relevant? Influenza Other Respir Viruses, 2013; 7(1):93–96; doi: 10.1111/j.1750-2659.2012.00359.x

27.

Lee

, Song

. H9N2 avian influenza virus in Korea: Evolution and vaccination. Clin Exp Vaccine Res, 2013; 2(1):26–33; doi: 10.7774/cevr.2013.2.1.26

28.

Lee

, Kang

, Song

, et al. Surveillance of avian influenza viruses in South Korea between 2012 and 2014. Virol J, 2017; 14(1):54; doi: 10.1186/s12985-017-0711-y

29.

Liu

, Lin

, Chen

, et al. Temperature drops and the onset of severe avian influenza A H5N1 virus outbreaks. PLoS One, 2007; 2(2):e191; doi: 10.1371/journal.pone.0000191

30.

Liu

, Kang

, Zhang

, et al. Independent and interactive effects of ambient temperature and absolute humidity on the risks of avian influenza A(H7N9) infection in China. Sci Total Environ, 2018; 619–620:1358–1365; doi: 10.1016/j.scitotenv.2017.11.226

31.

Lloren

KKS

, Lee

, Kwon

, et al. Molecular Markers for Interspecies Transmission of Avian Influenza Viruses in Mammalian Hosts. Int J Mol Sci, 2017; 18(12); doi: 10.3390/ijms18122706

32.

Meng

, Zhang

, Ma

, et al. Persistence of avian influenza virus (H9N2) on plastic surface. Sci Total Environ, 2022; 834:155355; doi: 10.1016/j.scitotenv.2022.155355

33.

Muzyka

, Pantin-Jackwood

, Spackman

, et al. Avian influenza virus wild bird surveillance in the Azov and Black Sea regions of Ukraine (2010-2011). Avian Dis, 2012; 56(Suppl 4):1010–1016; doi: 10.1637/10157-040912-ResNote.1

34.

, Chae

, Park

, et al. Avian influenza virus surveillance in wild bird in South Korea from 2019 to 2022. Korean J Vet Serv, 2022; 45(4):285–292; doi: 10.7853/kjvs.2022.45.4.285

35.

Nam

, Espano

, Song

, et al. Surveillance of avian influenza viruses from 2009 to 2013 in South Korea. Sci Rep, 2021; 11(1):23991; doi: 10.1038/s41598-021-03353-1

36.

, Gong

, Zhao

, et al. Prevalence of Haemophilus parasuis”Glaesserella parasuis” in pigs in China: A systematic review and meta-analysis. Prev Vet Med, 2020; 182:105083; doi: 10.1016/j.prevetmed.2020.105083

37.

, He

, Hu

, et al. Investigation of avian influenza virus in poultry and wild birds due to novel avian-origin influenza A(H10N8) in Nanchang City, China. Microbes Infect, 2015; 17(1):48–53; doi: 10.1016/j.micinf.2014.09.007

38.

Okamatsu

, Sakoda

, Kishida

, et al. Antigenic structure of the hemagglutinin of H9N2 influenza viruses. Arch Virol, 2008; 153(12):2189–2195; doi: 10.1007/s00705-008-0243-2

39.

Page

, McKenzie

, Bossuyt

, et al. The PRISMA 2020 statement: An updated guideline for reporting systematic reviews. Syst Rev, 2021; 10(1):89; doi: 10.1186/s13643-021-01626-4

40.

Peacock

THP

, James

, Sealy

, et al. A Global Perspective on H9N2 Avian Influenza Virus. Viruses, 2019; 11(7); doi: 10.3390/v11070620

41.

Piaggio

, Shriner

, VanDalen

, et al. Molecular surveillance of low pathogenic avian influenza viruses in wild birds across the United States: Inferences from the hemagglutinin gene. PLoS One, 2012; 7(12):e50834; doi: 10.1371/journal.pone.0050834

42.

Quan

, Wang

, Zhang

, et al. Avian influenza a viruses among occupationally exposed populations, China, 2014-2016. Emerg Infect Dis, 2019; 25(12):2215–2225; doi: 10.3201/eid2512.190261

43.

Ramey

, Pearce

, Reeves

, et al. Surveillance for Eurasian-origin and intercontinental reassortant highly pathogenic influenza A viruses in Alaska, spring and summer 2015. Virol J, 2016; 13:55; doi: 10.1186/s12985-016-0511-9

44.

Ruiz

, Galdames

, Baumberger

, et al. Remote Sensing and Ecological Variables Related to Influenza A Prevalence and Subtype Diversity in Wild Birds in the Lluta Wetland of Northern Chile. Viruses, 2023; 15(6); doi: 10.3390/v15061241

45.

Sarah

, Holly

, Jonathan

, et al. The cloacal microbiome of five wild duck species varies by species and influenza a virus infection status. 2018; 3; doi: 10.1128/msphere.00382-00318; doi: 10.1128/msphere.00382-18

46.

Shi

, Gao

, Zhu

, et al. Investigation of avian influenza infections in wild birds, poultry and humans in Eastern Dongting Lake, China. PLoS One, 2014; 9(4):e95685; doi: 10.1371/journal.pone.0095685

47.

Shortridge

. Pandemic influenza: A zoonosis? Semin Respir Infect, 1992; 7(1):11–25.

48.

Spackman

. A brief introduction to the avian influenza virus. Methods Mol Biol, 2008; 436:1–6; doi: 10.1007/978-1-59745-279-3_1

49.

Stallknecht

, Kearney

, Shane

, et al. Effects of pH, temperature, and salinity on persistence of avian influenza viruses in water. Avian Dis, 1990; 34(2):412–418.

50.

Sun

, Pu

, Fan

, et al. Evaluation of the protective efficacy of a commercial vaccine against different antigenic groups of H9N2 influenza viruses in chickens. Vet Microbiol, 2012; 156(1–2):193–199; doi: 10.1016/j.vetmic.2011.10.003

51.

Tang

, Tang

, Li

, et al. Avian Influenza Virus Prevalence and Subtype Diversity in Wild Birds in Shanghai, China, 2016-2018. Viruses, 2020; 12(9); doi: 10.3390/v12091031

52.

Tønnessen

, Kristoffersen

, Jonassen

, et al. Molecular and epidemiological characterization of avian influenza viruses from gulls and dabbling ducks in Norway. Virol J, 2013; 10:112; doi: 10.1186/1743-422x-10-112

53.

Verhagen

, Lexmond

, Vuong

, et al. Discordant detection of avian influenza virus subtypes in time and space between poultry and wild birds; Towards improvement of surveillance programs. PLoS One, 2017; 12(3):e0173470; doi: 10.1371/journal.pone.0173470

54.

Viechtbauer

. Conducting meta-analyses in R with the metafor package. J Stat Soft, 2010; 36(3):1–48; doi: 10.1863/7/jss.v036.i03

55.

Wang

, Chen

. Complete genome sequence of an H9N2 avian influenza virus isolated from egret in Lake Dongting wetland. J Virol, 2012; 86(21):11939; doi: 10.1128/jvi.02042-12

56.

Wang

, Li

, Xiong

, et al. Ecology of avian influenza viruses in migratory birds wintering within the Yangtze River wetlands. Sci Bull (Beijing), 2021; 66(19):2014–2024; doi: 10.1016/j.scib.2021.03.023

57.

Wang

, Wang

, Zhang

, et al. Prevalence, evolution, replication and transmission of H3N8 avian influenza viruses isolated from migratory birds in eastern China from 2017 to 2021. Emerg Microbes Infect, 2023; 12(1):2184178; doi: 10.1080/22221751.2023.2184178

58.

Webster

, Bean

, Gorman

, et al. Evolution and ecology of influenza A viruses. Microbiol Rev, 1992; 56(1):152–179; doi: 10.1128/mr.56.1.152-179.1992

59.

Wen

, Yan

, Chen

, et al. Recent H9N2 avian influenza virus lost hemagglutination activity due to a K141N substitution in hemagglutinin. J Virol, 2024; 98(4):e0024824; doi: 10.1128/jvi.00248-24

60.

Yan

, Cui

, Engelsma

, et al. Molecular and Antigenic Characterization of Avian H9N2 Viruses in Southern China. Microbiol Spectr, 2022; 10(1):e0082221; doi: 10.1128/spectrum.00822-21

61.

Yang

. Surveillance of Wild Bird Influenza Combined with Sentinels in Hubei Province(in Chinese). 2020; doi: 10.2700/9/d.cnki.gdblu.2020.000661

62.

Yao

, Zheng

, Xiong

, et al. Genetic and Pathogenic Characterization of Avian Influenza Virus in Migratory Birds between 2015 and 2019 in Central China. Microbiol Spectr, 2022; 10(4):e0165222; doi: 10.1128/spectrum.01652-22

63.

Zhang

, Huang

, Liao

, et al. H9N2 avian influenza viruses: Challenges and the way forward. Lancet Microbe, 2023; 4(2):e70–e71; doi: 10.1016/s2666-5247(22)00305-6

64.

Zhang

, Tang

, Liu

, et al. Characterization of H9N2 influenza viruses isolated from vaccinated flocks in an integrated broiler chicken operation in eastern China during a 5 year period (1998–2002). J Gen Virol, 2008; 89(Pt 12):3102–3112; doi: 10.1099/vir.0.2008/005652-0

65.

Zhang

, Chen

, et al. Evaluating the impact of environmental temperature on global highly pathogenic avian influenza (HPAI) H5N1 outbreaks in domestic poultry. Int J Environ Res Public Health, 2014; 11(6):6388–6399; doi: 10.3390/ijerph110606388

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.06 MB

0.00 MB