Skip to main content
NCBI home page
Emerging Microbes & Infections logoLink to Emerging Microbes & Infections
. 2026 Jan 16;15(1):2614742. doi: 10.1080/22221751.2026.2614742

Global molecular epidemiology of Mycoplasma pneumoniae and its association with macrolide resistance and disease severity: a systematic review and meta-analysis

Xinyue Zhu a,b,c,d,e,*, Yuyi Tang a,b,c,d,e,*, Karen Spruyt f, Janne Estill g,h, Ivan D Florez i,j,k, Yasser Sami Amer l,m,n,o, Akihiko Ozaki p,q,r, Yaolong Chen b,c,e,s,t,u,v,✉, Enmei Liu a,b,c,d,e,CONTACT
PMCID: PMC12818305  PMID: 41544187

ABSTRACT

The global molecular epidemiology of Mycoplasma pneumoniae (MP) and its associations with macrolide resistance and disease severity remain unclear. Studies reporting MP genotypes distribution by P1 typing, multiple locus variable number tandem repeat analysis (MLVA), or multilocus sequence typing analysis (MLST) among MP-infected patients were included. Data quality was assessed using the Joanna Briggs Institute tool and the Newcastle-Ottawa scale. Random-effects models were used to calculate pooled proportions of each MP genotype, with subgroup analyses by geographic region, age group, sex, time period, specimen type, and test assay. The proportions of macrolide-resistant MP and severe MP pneumonia cases were also evaluated. A total of 116 studies met the criteria. Predominant genotypes were P1-1 by P1 typing and 4572 by MLVA in the Western Pacific region and European region, and ST3 by MLST in the Western Pacific region. The proportions of P1-1 and 4572 were higher in the Western Pacific region than in other regions, and in children than in adults, whereas P1-2 and 3562 were opposite. Genotype dominance cycled between P1-1 and P1-2, and between 4572 and 3662. Macrolide resistance rates were highest in P1-1, 4572, and ST3 genotypes. Based on the currently available data, no association was detected between P1 genotypes and disease severity, although the limited sample size restricted the statistical power of this analysis. This comprehensive analysis elucidates the global molecular epidemiology of MP, highlights its clinical implications, and underscores the need for ongoing molecular surveillance to guide management and control strategies.

KEYWORDS: Molecular epidemiology, Mycoplasma pneumoniae, macrolide resistance, disease severity, genotype

Introduction

Mycoplasma pneumoniae (MP) is a leading cause of community-acquired pneumonia in children, responsible for 10%−40% of cases [1]. Despite substantial changes in respiratory pathogen epidemiology associated with the coronavirus disease 2019 pandemic, MP has remained among the primary causative pathogens of paediatric acute respiratory tract infections [2]. MP genotyping can be performed using several molecular methods, including P1 typing [3], multiple locus variable number tandem repeat analysis (MLVA) [4], and multilocus sequence typing analysis (MLST) (Supplementary Figure 1) [5]. P1 typing differentiates MP based on sequence variation within repetitive regions (RepMP2/3 and RepMP4) of the P1 gene, classifying strains primarily into P1-1, P1-2, and their variants [6]. MLVA is based on the variation in the copy number of tandemly repeated sequences across multiple genomic loci [4]. Based on an amended MLVA scheme using 4 variable-number tandem-repeat (VNTR) loci [7], genotypes 4572, 3562 and 3662 are identified as the most prevalent [8,9]. MLST relies on the polymorphisms within housekeeping genes [5], with sequence type (ST) 3 frequently identified as major sequence type in many regions [10].

Although previous studies have described the proportions of MP genotypes, substantial variability exists across geographic regions and age groups [11–13]. In addition, the temporal trends of MP genotype distribution remain unclear. Comprehensive knowledge of the epidemiological patterns of MP genotypes is essential for developing effective surveillance strategies and understanding potential clinical implications. However, a global synthesis of MP molecular epidemiology is still lacking.

Macrolides are the recommended first-line treatment for MP infections, but since 2000, macrolide-resistant Mycoplasma pneumoniae (MRMP) strains have emerged worldwide [14]. In China, the proportion of MRMP was as high as 93.02% in May 2024 [15]. While most infections are mild and self-limiting, MP can occasionally cause severe pneumonia and extrapulmonary complications [16,17]. Although prior research has characterized host–pathogen interactions in MP infections, evidence linking MP genotypes to macrolide resistance or disease severity remains limited and inconclusive [18–22].

This study aims to systematically review and meta-analyse the global molecular epidemiology of MP genotypes, focusing on P1 typing, MLVA, and MLST methods. Additionally, we evaluated the association of different genotypes with macrolide resistance and disease severity.

Methods

Search strategy, data sources, and inclusion and exclusion criteria

This systematic review was registered on PROSPERO (registration no. CRD42024630059) and conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [23]. The PRISMA 2020 checklist and its extension for abstracts were used to ensure comprehensive reporting (Supplementary Table 1).

We searched the following electronic databases: PubMed, Embase, Web of Science, Cochrane Library, China National Knowledge Infrastructure, China Biology Medicine disc, and Wanfang Data. Additionally, we screened the reference lists of included articles manually to identify further eligible studies. The search covered all records from inception to December 2, 2024, and was restricted to publications in English or Chinese, which constitute the majority of relevant studies in this field. The search terms included “Mycoplasma pneumoniae,” “subtype,” “genotype,” “molecular,” and their derivatives (Supplementary Table 2).

We included cross-sectional, cohort, and case–control studies that reported MP genotypes using P1 typing, MLVA, or MLST in patients with MP infection and provided extractable genotype frequency data. We excluded reviews, conference abstracts, theses, dissertations, and articles with irretrievable full texts or incomplete statistics. To minimize the influence of unstable extreme values on pooled estimates, studies with a sample size less than 10 were excluded [24].

Study selection, data extraction and quality assessment

Two investigators independently screened all identified articles by title and abstract according to the selection criteria. Full texts of potentially eligible studies were retrieved and evaluated. EndNote 21 software was used to manage references. Two investigators independently extracted data into a pre-defined excel spreadsheet, including the following variables: first author; publication year; time period; geographic region; population (children and/or adults); number of MP infections successfully genotyped; distribution of MP genotypes; genotype distribution by geographic region, age group, sex (male or female), time period, specimen type, and test assay; number of MRMP and macrolide-sensitive MP for each genotype; MRMP genotype distribution by geographic region; distribution of 23S rRNA mutations by genotype in MRMP strains; number of severe Mycoplasma pneumoniae pneumonia (SMPP) and general Mycoplasma pneumoniae pneumonia for each genotype. MRMP was defined based on either (1) detection of macrolide resistance-associated mutations in domain V of the 23S rRNA gene; or (2) determination of minimum inhibitory concentrations for macrolides using the broth microdilution method [25]. SMPP was defined as MPP with at least one of the following clinical features [26]: (1) poor general condition; (2) high fever; (3) increased respiratory rate; (4) cyanosis or dyspnoea; (5) multilobar or ≥2/3 pulmonary infiltration; (6) transcutaneous oxygen saturation ≤92% in room air; or (7) presence of extrapulmonary complications.

Two investigators independently assessed study quality using the Joanna Briggs Institute (JBI) critical appraisal tool for cross-sectional studies reporting prevalence data, and the Newcastle-Ottawa scale (NOS) for cohort studies [27,28]. For JBI, each “yes” response scored 1, while “no,” “not clear,” or “not applicable” responses scored 0; these values were then added up to determine the total score. For NOS, a maximum score of 9 points could be obtained, with 1 point awarded for each item in the Selection and Outcome domains and up to 2 points in the Comparability domain.

Data analysis

We calculated the proportions of each MP genotype by dividing the number of cases of that genotype by the total number of successfully genotyped patients. Results were visualized using forest plots. A meta-analysis was conducted using the DerSimonian-Laird random-effect model due to anticipated heterogeneity across studies. The pooled proportion and 95% confidence intervals (CIs) were computed using Freeman-Tukey double arcsine transformation to stabilize variance in proportions. Heterogeneity was assessed using the I2 statistic.

We conducted subgroup analyses, sensitivity analyses, and meta-regression to explore sources of heterogeneity and presented the results with tables and line graphs. If sufficient data were available, subgroup analyses were stratified by geographic region, age group, sex, time period, specimen type, and test assay. Geographic regions were defined according to the World Health Organization classification [29]. Sensitivity analyses were performed in 3 ways: (1) excluding 5% of studies with highest and 5% with lowest MP genotype proportion estimates; (2) omitting one study at a time; and (3) sequentially excluding subgroups. Meta-regression analyses were conducted using univariate models for each covariate. Publication bias was assessed through visual inspection of funnel plots, and quantified using Egger’s test.

We also calculated the proportion of MRMP by dividing the number of MRMP cases by the number of MP strains that were both genotyped and tested for macrolide resistance. We further calculated the proportion of 23S rRNA mutations by dividing the number of a certain mutation type by the number of MRMP strains. Similarly, the proportion of SMPP was calculated based on the number of MP strains genotyped and assessed for disease severity. Chi-square tests were used to compare proportions across genotypes with statistical significance set at P < 0.05.

All statistical analyses were performed using R software (version 4.4.2), utilizing the Metaprop and Metareg functions for meta-analyses and meta-regression, respectively.

Results

Study characteristics

The study selection process (Figure 1) and study characteristics (Supplementary Table 3) are presented. In total, 116 studies from 25 countries were included in the quantitative analysis, including 6 multinational studies that were stratified by country, yielding 133 country-level records (Figure 2). Apart from 4 multi-regional studies, most studies were conducted in the Western Pacific region (78 studies, 67.2%), followed by the European region (23 studies, 19.8%), the American region (8 studies, 6.9%), the Eastern Mediterranean region (1 study, 0.9%), the African region (1 study, 0.9%), and the South-east Asian region (1 study, 0.9%). Of the 116 studies, 67 focused exclusively on children, 34 included both children and adults, and 4 focused only on adults; 11 studies did not report participants’ ages. In the diagnosis of MP infection, polymerase chain reaction (PCR) was employed in 72 studies, culture in 14 studies, and both methods were used in 18 studies. Regarding specimen types, throat swabs (TS) were the most commonly used (28 studies), followed by nasopharyngeal aspirates (NPA; 7 studies), oropharyngeal swabs (OPS; 7 studies), bronchoalveolar lavage fluid (BALF; 6 studies), nasopharyngeal swabs (NPS; 5 studies), and bronchial aspirates (BA; 4 studies).

Figure 1.

Figure 1.

Open in a new tab

Flow diagram of study selection. Abbreviations: CNKI, China National Knowledge Infrastructure; CBM, China Biology Medicine disc.

Figure 2.

Figure 2.

Open in a new tab

Geographic distribution of country-level records included in the meta-analysis. Data are shown at the country level.

The total numbers of cases successfully genotyped according to P1 typing, MLVA, and MLST were 16,852, 9268, and 1433, respectively. Quality assessment using the JBI and NOS tools showed that most studies were of moderate to high methodological quality. The JBI scores varied from 4 to 8 (101 studies) and NOS scores from 6 to 8 (15 studies) (Supplementary Table 4).

Pooled proportion of different MP genotypes

Globally, among the 97 studies using P1 typing, the pooled proportion of P1-1 genotype was 75.4% (95% CI 71.0–79.5), followed by P1-2 at 24.7% (95% CI 20.5–29.1), and P1-2v at 3.8% (95% CI 1.7–6.4). Based on 55 studies using MLVA, the most common genotype was 4572 (61.8% [95% CI 55.4–67.9]), followed by 3562 (22.8% [95% CI 18.1–27.9]), 3662 (2.8% [95% CI 1.3–4.8]), and other MLVA types (5.7% [95% CI 4.1–7.6]). In the 10 studies using MLST, the ST3 genotype was the most prevalent (62.3% [95% CI 52.6–71.5]), followed by ST14 (13.9% [95% CI 7.8–21.3]), and other MLST types (20.7% [95% CI 12.2–30.8]) (Figure 3; Supplementary Figure 2).

Figure 3.

Figure 3.

Open in a new tab

Forest plot of the proportion of different MP genotypes. * P1-2 includes P1-2v.

Distribution of MP genotypes across subgroups

To further characterize the epidemiological patterns of MP genotypes, subgroup analyses were performed by geographic region, age group, sex, time period, specimen type, and test assay (Table 1; Supplementary Table 5 and Supplementary Figures 3–10).

Table 1.

Distribution of MP genotypes across different geographic regions, age groups, and sexes

　 P1 MLVA MLST
Variable P1-1 P1-2* P1-2v 4572 3562 3662 Other ST3 ST14 Other
Geographic region                  
Western Pacific                  
 Proportion (%) 83.1 16.9 1.6 71.0 16.9 0.0 5.4 62.3 13.9 20.7
 (95% CI) (78.2–87.5) (12.5–21.9) (0.2–3.8) (62.5–78.9) (10.7–24.1) (0.0–0.0) (2.8–8.6) (52.6–71.5) (7.8–21.3) (12.2–30.8)
 I2 (%) 97.5 97.5 97.4 97.4 97.3 36.1 93.8 92.2 92.0 94.4
No. of studies 68 68 68 33 33 33 33 10 10 10
Sample sizes 8345/11079 2740/11079 789/11079 3725/5194 1074/5194 23/5194 372/5194 867/1433 221/1433 345/1433
 Europe                    
 Proportion (%) 60.2 39.8 7.3 45.3 31.8 10.5 5.6 NA NA NA
 (95% CI) (53.9–66.4) (33.6–46.1) (1.5–16.3) (36.9–53.8) (23.2–41.1) (6.7–15.0) (3.8–7.8) NA NA NA
 I2 (%) 93.6 93.6 98.8 93.0 94.6 87.6 64.4 NA NA NA
 No. of studies 21 21 21 13 13 13 13 NA NA NA
Sample sizes 2888/4774 1886/4774 446/4774 1343/2621 725/2621  410/2621 143/2621 NA NA NA
 Americas                    
 Proportion (%) 46.8 53.2 23.5 52.2 29.4 11.7 3.4 NA NA NA
 (95% CI) (29.7–64.3) (35.7–70.3) (5.3–48.5) (39.9–64.4) (20.5–39.0) (6.5–18.1) (1.5–5.9) NA NA NA
 I2 (%) 95.9 95.9 98.2 91.9 87.8 83.3 56.6 NA NA NA
 No. of studies 9 9 9 8 8 8 8 NA NA NA
 Sample sizes 483/988 505/988 347/988 507/1028 333/1028 133/1028 55/1028 NA NA NA
Age group                    
 Children                    
 Proportion (%) 81.5 18.6 3.1 69.1 18.8 0.5 5.6 62.8 13.9 19.2
 (95% CI) (76.5–86.0) (14.0–23.6) (1.0–6.1) (61.1–76.6) (12.8–25.6) (0.0–1.6) (3.3–8.5) (50.3–74.5) (6.5–23.4) (8.5–32.6)
 I2 (%) 97.3 97.3 97.7 97.2 97.0 89.4 92.9 93.0 92.4 95.3
 No. of studies 71 71 71 37 37 37 37 9 9 9
 Sample sizes 8085/10588 2504/10588 829/10588 3929/5585 1136/5585 153/5585 367/5585 550/930 132/930 248/930
 0∼4 years                    
 Proportion (%) 74.6 25.5 4.2 64.8 17.2 0.1 3.0 NA NA NA
 (95% CI) (55.7–90.4) (9.6–44.3) (0.0–19.3) (57.4–71.9) (8.5–27.4) (0.0–6.0) (0.0–14.3) NA NA NA
 I2 (%) 87.1 87.1 89.7 0.0 21.1 43.3 57.4 NA NA NA
 No. of studies 13 13 13 8 8 8 8 NA NA NA
 Sample sizes 332/443 111/443 22/443 158/254 52/254 25/254 19/254 NA NA NA
 5∼9 years                    
 Proportion (%) 74.5 25.5 10.3 59.9 28.5 0.7 2.4 NA NA NA
 (95% CI) (51.8–92.6) (7.4–48.2) (0.0–39.2) (32.1–85.2) (5.0–58.9) (0.0–10.8) (0.0–10.3) NA NA NA
 I2 (%) 92.9 92.9 96.6 86.9 89.3 72.2 47.0 NA NA NA
 No. of studies 11 11 11 6 6 6 6 NA NA NA
 Sample sizes 339/517 178/517 64/517 226/370 76/370 53/370 15/370 NA NA NA
 10∼17 years                    
 Proportion (%) 62.3 37.7 15.2 59.7 21.7 0.5 2.0 NA NA NA
 (95% CI) (37.0–85.2) (14.8–63.0) (0.0–54.5) (28.1–88.2) (3.1–47.0) (0.0–8.3) (0.0–6.5) NA NA NA
 I2 (%) 87.3 87.3 95.7 87.6 81.3 50.4 4.6 NA NA NA
 No. of studies 10 10 10 7 7 7 7 NA NA NA
 Sample sizes 156/268 112/268 46/268 122/223 55/223 28/223 18/223 NA NA NA
 Adults                    
 Proportion (%) 53.6 46.4 5.2 51.2 25.5 4.4 5.2 30.0 19.6 32.2
 (95% CI) (42.9–64.2) (35.8–57.1) (0.1–14.2) (40.3–62.1) (17.7–34.0) (0.1–12.5) (2.2–9.1) (0.0–91.2) (5.2–38.2) (0.6–76.3)
 I2 (%) 86.6 86.6 92.6 75.1 63.4 83.1 33.7 77.2 0.0 55.3
 No. of studies 18 18 18 13 13 13 13 2 2 2
 Sample sizes 536/973 437/973 55/973 323/566 142/566 56/566 45/566 16/31 7/31 8/31
 18∼49 years                    
 Proportion (%) 59.3 40.7 16.4 44.9 8.0 13.0 11.4 NA NA NA
 (95% CI) (36.3–80.8) (19.2–63.8) (0.0–55.7) (11.2–80.8) (0.0–23.3) (6.7–20.3) (0.0–38.8) NA NA NA
 I2 (%) 61.0 61.0 88.8 76.3 29.6 0.0 65.3 NA NA NA
 No. of studies 7 7 7 5 5 5 5 NA NA NA
 Sample sizes 112/186 74/186 10/186 96/169 28/169 32/169 13/169 NA NA NA
 ≥ 50 years                    
 Proportion (%) 37.0 63.0 4.0 29.2 56.5 4.8 0.3 NA NA NA
 (95% CI) (23.5–51.2) (48.9–76.5) (0.0–33.0) (6.8–56.6) (18.1–91.7) (0.0–20.5) (0.0–23.2) NA NA NA
 I2 (%) 0.0 0.0 77.2 21.4 55.1 0.0 46.5 NA NA NA
 No. of studies 7 7 7 4 4 4 4 NA NA NA
 Sample sizes 32/80 48/80 4/80 13/34 15/34 4/34 2/34 NA NA NA
Sex                    
 Male                    
 Proportion (%) 77.2 22.9 0.7 65.8 25.1 2.3 1.8 NA NA NA
 (95% CI) (66.6–86.4) (13.6–33.4) (0.0–6.2) (36.1–90.3) (4.4–54.0) (0.0–12.2) (0.0–6.0) NA NA NA
 I2 (%) 93.1 93.1 95.2 97.2 97.2 92.8 71.8 NA NA NA
 No. of studies 15 15 15 6 6 6 6 NA NA NA
 Sample sizes 921/1313 392/1313 55/1313 306/494 115/494 53/494 20/494 NA NA NA
 Female                    
 Proportion (%) 78.1 22.0 0.3 63.0 24.7 1.9 3.6 NA NA NA
 (95% CI) (67.9–87.0) (13.2–32.2) (0.0–3.9) (34.5–87.5) (7.1–48.0) (0.0–8.3) (0.5–8.6) NA NA NA
 I2 (%) 92.3 92.3 91.7 96.9 95.6 85.7 71.4 NA NA NA
 No. of studies 15 15 15 6 6 6 6 NA NA NA
 Sample sizes 909/1268 360/1268 34/1268 287/450 114/450 29/450 20/450 NA NA NA
Open in a new tab

Abbreviation: NA = not applicable. * P1-2 includes P1-2v. Sample sizes are the number of cases of a certain genotype / the total number of successfully genotyped patients.

With regard to geographic distribution, the proportions of P1-1 (83.1% [95% CI 78.2–87.5]) and 4572 (71.0% [95% CI 62.5–78.9]) were highest in the Western Pacific region compared to Europe and the Americas. In contrast, P1-2 (53.2% [95% CI 35.7–70.3]) was more prevalent in the American region. 3562 and 3662 genotypes were mainly observed in Europe and the Americas, where their proportions were higher than those in the Western Pacific region. For MLST, all 10 studies were conducted in the Western Pacific region (Table 1 and Supplementary Figure 3).

As for age groups, genotypes P1-1, 4572, and ST3 were more prevalent in children (81.5% [95% CI 76.5–86.0], 69.1% [95% CI 61.1–76.6], 62.8% [95% CI 50.3–74.5], respectively) than in adults (53.6% [95% CI 42.9–64.2], 51.2% [95% CI 40.3–62.1], 30.0% [95% CI 0.0–91.2], respectively). In contrast, genotypes P1-2, 3562, 3662 and ST14 were more common in adults. Specifically, P1-1 and 4572 were most common in children under 5 years (74.6% [95% CI 55.7–90.4] and 64.8% [95% CI 57.4–71.9], respectively), whereas P1-2 (63.0% [95% CI 48.9–76.5]) and 3562 (56.5% [95% CI 18.1–91.7]) were predominant in adults aged 50 years and above. No major differences in genotype proportion were observed between sexes (Table 1 and Supplementary Figures 4–6).

Globally, we observed a symmetrical switch between genotypes P1-1 and P1-2, as well as between 4572 and 3662 (Figure 4 and Supplementary Figure 7). Before 2010, these pairs of genotypes exhibited shifts in predominance approximately every 10 years. In recent years, the proportion of P1-1, although remaining relatively high, showed a decreasing trend, declining from 80.0% (95% CI 73.5–85.8) between 2010 and 2014 to 67.8% (95% CI 45.1–86.9) between 2020 and 2024. At the same time, the proportion of P1-2, while still less prevalent, showed an increasing trend, rising from 20.1% (95% CI 14.2–26.5) to 32.2% (95% CI 13.1–54.9). Similarly, the proportion of 4572 declined from >60% during 2010–2019 to 47.7% (95% CI 12.7–84.1) during 2020–2024, whereas 3562 rose from <25% to 44.9% (95% CI 12.4–80.0). The proportion of 3662 stayed on a consistently low level (<3%) since 2010. ST3 remained consistently predominant throughout the observed time periods (∼65%) in the Western Pacific region, while ST14 remained below 20%.

Figure 4.

Figure 4.

Open in a new tab

Distribution of MP genotypes across different time periods. (A) Distribution of P1 genotypes across different time periods. (B) Distribution of MLVA genotypes across different time periods. (C) Distribution of MLST genotypes across different time periods. * P1-2 includes P1-2v.

When stratified by WHO regions, we can still observe the trend of the alternating dominance between the two major pairs of genotypes (Figure 4 and Supplementary Figure 8). However, the direction and timing of shifts diverged by regions. The proportion of P1-1 decreased in the Western Pacific region since 2005 and in the American region since 2010; however, it showed an opposite trend in the European region, increasing since 2000. Similarly, 4572 declined in the Western Pacific since 2015 and in the Americas since 2010, while increased in Europe since 2000.

Moreover, we performed subgroup analyses by specimen type and test assay (Supplementary Table 5, and Supplementary Figures 9 and 10). The prevalence of P1-1 and 4572 was higher in NPA, BA and BALF. Conversely, P1-2 and 3562 were most frequently detected in OPS. Regarding test assay, the use of both PCR and culture was associated with an increased proportion of P1-1 and 4572, while PCR alone yielded the highest proportion of P1-2 and 3562.

Sensitivity analysis, meta-regression analysis, and publication bias

Sensitivity analyses showed that excluding extreme studies (highest 5% and lowest 5%), any single study, or most subgroups did not substantially alter pooled proportion estimates (Supplementary Tables 6 and 7 and Supplementary Figure 11). However, the exclusion of Western Pacific region or children subgroup significantly affected proportion estimates, suggesting that these two subgroups were the primary driver of observed heterogeneity. Specifically, removing either the Western Pacific region or the children subgroup resulted in decreased proportions of P1-1 and 4572, whereas the proportions of P1-2, 3562, and 3662 increased. In addition, exclusion of the children subgroup led to a decrease in ST3 and an increase in ST14. These trends were consistent with the findings from the subgroup analyses.

To quantify the independent contribution of these factors, we performed meta-regression analyses. The results confirmed that geographic region and age group were statistically significant sources of heterogeneity (Supplementary Table 8). P1-1 and 4572 were more predominant in the Western Pacific region, while P1-2, 3562, and 3662 were more common in the American region and European region. P1-1 (coefficient –0.31; 95%CI: –0.45, –0.17) and 4572 (coefficient –0.21; 95%CI: –0.38, –0.04) showed a significant negative association with age (higher prevalence in children), whereas P1-2 (coefficient 0.31; 95%CI: 0.17, 0.45) and 3662 (coefficient 0.17; 95%CI: 0.07, 0.28) showed a significant positive association (higher prevalence in adults).

Regarding publication bias, funnel plots were generally symmetrical (Supplementary Figure 12), and Egger’s test indicated no significant publication bias (P > 0.05 for all genotypes).

Macrolide resistance and disease severity by genotype

Based on 64 studies reporting MRMP cases by genotype, MRMP was most frequently observed in P1-1 (72.8% [95% CI 55.6–87.1]), 4572 (67.6% [95% CI 48.8–84.0]), and ST3 (76.7% [95% CI 52.0–94.5]) genotypes (Supplementary Figures 13 and 14), suggesting that these genotypes are commonly associated with macrolide resistance. When stratified by WHO regions, a similar genotype-resistance pattern was observed, with MRMP remaining more prevalent in P1-1 and 4572 genotypes across regions. However, the absolute proportions of MRMP in these genotypes were substantially higher in the Western Pacific region than in Europe and the Americas (Supplementary Table 9 and Supplementary Figure 15).

We further examined the distribution of 23S rRNA mutations by genotype in MRMP strains. The A2063G mutation predominated in all genotypes, with no significant difference in mutation pattern between genotypes (Supplementary Table 10 and Supplementary Figure 16).

Due to the currently available data, SMPP proportion was analysed only for P1-1 and P1-2 across 5 studies. SMPP accounted for 45.8% (95% CI 28.9–63.1) of P1-1 cases and 48.5% (95% CI 18.0–79.5) of P1-2 cases, with no statistically significant difference observed (P = 0.58) (Supplementary Figure 17). Definitive associations remain elusive owing to the small number of studies and inconsistent definitions of SMPP.

Discussion

In this systematic review and meta-analysis, we provide a comprehensive synthesis of the global molecular epidemiology of MP and its association with macrolide resistance. Our findings confirm that the predominant genotypes globally are P1-1 (P1 typing), 4572 (MLVA), and ST3 (MLST). Genotype distributions varied by geographic region, age group, time period, specimen type and test assay. Predominant genotypes were P1-1 by P1 typing and 4572 by MLVA in most regions, and ST3 by MLST in the Western Pacific region. Notably, a cyclical type shift was observed over time between P1-1 and P1-2, and for the first time between 4572 and 3662. In addition, P1-1, 4572, and ST3 genotypes were associated with MRMP strains.

Regional disparities were apparent. The proportions of P1-1 and 4572 were notably higher in the Western Pacific region. In fact, P1-1 and 4572 were the predominant genotypes in most regions except that P1-2 was more prevalent in the American region. These differences may be influenced by geographic, climatic, and sociocultural factors. Additionally, the lack of data in the American region before 2010 and after 2019 may also influenced the result. Besides, the apparent global predominance of MLST ST3 was entirely attributable to the fact that all available MLST data originate from the Western Pacific region. Therefore, it is essential that future studies incorporate MLST data from other regions to accurately determine its true global distribution.

A cyclical genotype replacement of P1-1 and P1-2, previously described in Japan with an approximate 10-year periodicity, was corroborated by our findings [30]. Notably, we also observed a similar type shift phenomenon between 4572 and 3662 that to our knowledge has not been reported before. Some longitudinal studies have shown that a shift in the dominance of P1 type may be related to the occurrence of outbreaks [31,32]. For example, in Korea, the predominant genotype shifted from P1-2 to P1-1 around 2006, and notably, a nationwide outbreak occurred during the same year [33,34]. Collectively, these studies provide support for the theory that genotype dominance may be shaped by population-level immunity. As herd immunity accumulates against a dominant strain, selective pressure may facilitate emergence of alternate genotypes, contributing to cyclical epidemics [32]. Of note, paediatric populations exhibited higher proportions of P1-1, 4572, and ST3, whereas P1-2, 3562, 3662, and ST14 were more common among adults. These differences may reflect age-related immunological variations.

Our analysis identified a strong association between MRMP and P1-1, 4572, and ST3. These genotypes were also predominant in the Western Pacific region and among child populations with documented high MRMP proportion [35]. This suggests that sustained dominance of a genotype under macrolide pressure may foster the emergence of resistance [36]. Conversely, if genotypes such as P1-2 or 3562 were to predominate over time, it could be hypothesized that they might similarly acquire resistance [37]. Longitudinal molecular surveillance is essential to monitor such evolutionary dynamics. Given that our findings indicated variation in specific mutation sites is unlikely to explain the observed differences in MRMP prevalence across genotypes, further basic experimental research is needed to explore the molecular mechanisms driving these associations.

Based on the currently available data, we found no significant difference in the proportion of SMPP between P1-1 and P1-2 genotypes. Several studies similarly reported that the proportion of SMPP was not associated with MLVA genotypes [20,38]. A recent article also detected no link between MLST genotypes and severity of illness [39]. These findings were consistent with previous studies suggesting that clinical severity is more closely linked to pathogen load than to genotype [40]. However, definitive associations remain elusive due to heterogeneity in laboratory criteria used to assess severity [22,41]. The limited number of studies and inconsistent definitions of SMPP across cohorts further constrain interpretation [26,42]. These preliminary findings require validation in future prospective studies employing standardized, objective severity classifications.

Interestingly, our subgroup analyses revealed similar distribution patterns and resistance rates between the genotype profiles of P1-1/4572/ST3 and P1-2/3562/ST14, suggesting a potential co-evolution or linkage across typing methods. This observation is supported by previous studies demonstrating overlapping genotype associations [10,34,37,43].

The findings of our review highlight the critical need for integrating molecular surveillance of MP into routine clinical and public health practice, particularly in paediatric care settings. Given the high burden of MRMP in paediatric respiratory infections and the relationship between MRMP and genotypes, antimicrobial stewardship programmes must utilize local genotype surveillance to guide antibiotic prescribing, reduce resistance, and protect the efficacy of macrolides for future use. Moreover, periodic changes in dominant genotypes may coincide with epidemic surges. Public health laboratories and paediatric hospitals should be resourced to conduct P1 typing and MLVA, as these data can serve as early indicators for MP outbreaks.

Besides, these findings highlight the importance of revising paediatric pneumonia practice guidelines to reflect molecular epidemiology insights. Clinical practice guidelines should no longer treat MP as a homogeneous entity; rather, they should incorporate risk stratification by genotype, especially when dealing with severe or treatment-refractory cases. As shown in this review, genotypes like P1-1 and 4572 are not only more prevalent in children but are also more likely to harbour resistance mutations. Updating evidence-based national and regional pneumonia treatment guidelines and protocols to reflect these insights can improve clinical outcomes, reduce unnecessary antimicrobial exposure, and promote evidence-based care delivery.

This study has several limitations. First, significant heterogeneity existed across included studies, partially attributable to differences in geographic regions and population characteristics. However, residual unexplained heterogeneity is likely to remain. Second, the limited and methodologically inconsistent data on genotype-disease severity associations preclude firm conclusions. Findings in this area should be considered preliminary and warrant validation through future, specifically designed studies. Third, the languages of the included studies were limited to English and Chinese. Although these two languages account for the vast majority of published MP genotyping studies, some potentially relevant studies may have been omitted and should be incorporated in future updates of this review. Fourth, although the search was limited to studies published before December 2024, an updated search through August 2025 indicated that more recent studies were consistent with our findings. Future research should incorporate these updates to further refine the conclusions.

To our knowledge, this is the first comprehensive meta-analysis of the global molecular epidemiology of MP. We identified predominant genotypes and described temporal and geographic trends, as well as associations with macrolide resistance. These findings underscore the need for ongoing molecular surveillance and provide a foundation for future studies exploring the clinical implications of MP genotypes.

Authors’ contributions

XZ and YT contributed equally and served as co-first authors. EL and YC served as co-corresponding authors. Concept and design: YT; Acquisition, analysis, or interpretation of data: XZ, YT; Drafting of the manuscript: XZ, YT, EL, YC; Critical revision of the manuscript for important intellectual content: KS, JE, IDF, YSA, AO; Statistical analysis: XZ, YT; Obtained funding: EL; Administrative, technical, or material support: EL, YC. All authors reviewed the manuscript, and agreed with the decision to submit for publication.

Supplementary Material

Supplemental Material
TEMI_A_2614742_SM4587.docx (2.7MB, docx)

Funding Statement

This work was supported by the National Key Research and Development Program of China [2022YFC2704900].

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

This meta-analysis did not require the collection of new data, but rather the analysis of previously published data. Extracted and aggregated data from published studies are available on reasonable request to the corresponding author.

Supplemental Material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/22221751.2026.2614742.

References

  • 1.Cillóniz C, Torres A, Niederman M, et al. Community-acquired pneumonia related to intracellular pathogens. Intensive Care Med. 2016;42(9):1374–1386. doi: 10.1007/s00134-016-4394-4 [DOI] [PubMed] [Google Scholar]
  • 2.Tang Y, Dang X, Lv M, et al. Changes in the prevalence of respiratory pathogens in children due to the COVID-19 pandemic: a systematic review and meta-analysis. J Infect. 2023;86(2):154–225. doi: 10.1016/j.jinf.2022.11.023 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Kenri T, Taniguchi R, Sasaki Y, et al. Identification of a new variable sequence in the P1 cytadhesin gene of Mycoplasma pneumoniae: evidence for the generation of antigenic variation by DNA recombination between repetitive sequences. Infect Immun. 1999;67(9):4557–4562. doi: 10.1128/IAI.67.9.4557-4562.1999 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Dégrange S, Cazanave C, Charron A, et al. Development of multiple-locus variable-number tandem-repeat analysis for molecular typing of Mycoplasma pneumoniae. J Clin Microbiol. 2009;47(4):914–923. doi: 10.1128/JCM.01935-08 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Brown RJ, Holden MT, Spiller OB, et al. Development of a multilocus sequence typing scheme for molecular typing of Mycoplasma pneumoniae. J Clin Microbiol. 2015;53(10):3195–3203. doi: 10.1128/JCM.01301-15 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Dumke R, Catrein I, Herrmann R, et al. Preference, adaptation and survival of Mycoplasma pneumoniae subtypes in an animal model. Int J Med Microbiol. 2004;294(2-3):149–155. doi: 10.1016/j.ijmm.2004.06.020 [DOI] [PubMed] [Google Scholar]
  • 7.Sun H, Xue G, Yan C, et al. Multiple-locus variable-number tandem-repeat analysis of Mycoplasma pneumoniae clinical specimens and proposal for amendment of MLVA nomenclature. PLoS One. 2013;8(5):e64607. doi: 10.1371/journal.pone.0064607 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Diaz MH, Benitez AJ, Winchell JM.. Investigations of Mycoplasma pneumoniae infections in the United States: trends in molecular typing and macrolide resistance from 2006 to 2013. J Clin Microbiol. 2015;53(1):124–130. doi: 10.1128/JCM.02597-14 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Gullsby K, Olsen B, Bondeson K.. Molecular typing of Mycoplasma pneumoniae strains in Sweden from 1996 to 2017 and the emergence of a new P1 Cytadhesin gene, Variant 2e. J Clin Microbiol. 2019;57(6):e00049-19. doi: 10.1128/JCM.00049-19 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Chen Y, Li X, Fu Y, et al. Whole-genome sequencing unveils the outbreak of Mycoplasma pneumoniae in mainland China. Lancet Microbe. 2024;5(9):100870. doi: 10.1016/S2666-5247(24)00086-7 [DOI] [PubMed] [Google Scholar]
  • 11.Kenri T, Yamazaki T, Ohya H, et al. Genotyping of Mycoplasma pneumoniae strains isolated in Japan during 2019 and 2020: spread of p1 gene type 2c and 2j variant strains. Front Microbiol. 2023;14:1202357. doi: 10.3389/fmicb.2023.1202357 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Tian XJ, Dong YQ, Dong XP, et al. P1 gene of Mycoplasma pneumoniae in clinical isolates collected in Beijing in 2010 and relationship between genotyping and macrolide resistance. Chin Med J (Engl). 2013;126(20):3944–3948. doi: 10.3760/cma.j.issn.0366-6999.20131643 [DOI] [PubMed] [Google Scholar]
  • 13.Diaz MH, Benitez AJ, Cross KE, et al. Molecular detection and characterization of Mycoplasma pneumoniae among patients hospitalized with community-acquired pneumonia in the United States. Open Forum Infect Dis. 2015;2(3):ofv106. doi: 10.1093/ofid/ofv106 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Jiang Y, Kang H, Dou H, et al. Comparative genomic sequencing to characterize Mycoplasma pneumoniae genome, typing, and drug resistance. Microbiol Spectr. 2024;12(8):e0361523. doi: 10.1128/spectrum.03615-23 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Sun Y, Li P, Jin R, et al. Characterizing the epidemiology of Mycoplasma pneumoniae infections in China in 2022-2024: a nationwide cross-sectional study of over 1.6 million cases. Emerg Microbes Infect. 2025;14(1):2482703. doi: 10.1080/22221751.2025.2482703 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Rowlands RS, Sauteur M, Beeton PM, et al. Mycoplasma pneumoniae: not a typical respiratory pathogen. J Med Microbiol. 2024;73(10):001910. doi: 10.1099/jmm.0.001910 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Lee KL, Lee CM, Yang TL, et al. Severe Mycoplasma pneumoniae pneumonia requiring intensive care in children, 2010-2019. J Formos Med Assoc. 2021;120(1 Pt 1):281–291. doi: 10.1016/j.jfma.2020.08.018 [DOI] [PubMed] [Google Scholar]
  • 18.Tang Y, Fu X, Li X, et al. Multi-omics analysis of human plasma reveals reprogramming of tryptophan metabolism associated with inflammation in Mycoplasma pneumoniae pneumonia in children. J Infect. 2025;91(1):106525. doi: 10.1016/j.jinf.2025.106525 [DOI] [PubMed] [Google Scholar]
  • 19.Lee JK, Choi YY, Sohn YJ, et al. Persistent high macrolide resistance rate and increase of macrolide-resistant ST14 strains among Mycoplasma pneumoniae in South Korea, 2019-2020. J Microbiol Immunol Infect. 2021;55(5):910–916. doi: 10.1016/j.jmii.2021.07.011 [DOI] [PubMed] [Google Scholar]
  • 20.Wang Y, Xu B, Wu X, et al. Increased macrolide resistance rate of M3562 Mycoplasma pneumoniae correlated with macrolide usage and genotype shifting. Front Cell Infect Microbiol. 2021;11:675466. doi: 10.3389/fcimb.2021.675466 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Fan L, Li D, Zhang L, et al. Pediatric clinical features of Mycoplasma pneumoniae infection are associated with bacterial P1 genotype. Exp Ther Med. 2017;14(3):1892–1898. doi: 10.3892/etm.2017.4721 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Berlot JR, Krivec U, Mrvič T, et al. Mycoplasma pneumoniae P1 genotype indicates severity of lower respiratory tract infections in children. J Clin Microbiol. 2021;59(8):e00220-21. doi: 10.1128/JCM.00220-21 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. Br Med J. 2021;372:n71. doi: 10.1136/bmj.n71 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Walker J, Holm Hansen C, Martin P, et al. Prevalence of depression in adults with cancer: a systematic review. Ann Oncol. 2013;24(4):895–900. doi: 10.1093/annonc/mds575 [DOI] [PubMed] [Google Scholar]
  • 25.Waites KB, Bade DJ, Bébéar C, et al. Methods for antimicrobial susceptibility testing for human mycoplasmas; approved guideline. Wayne (PA: ): Clinical and Laboratory Standards Institute; 2011. [PubMed] [Google Scholar]
  • 26.The Subspecialty Group of Respiratory tSoP, Chinese Medical Association, China National Clinical Research Center of Respiratory Diseases, Editorial Board, Chinese Journal of Pediatrics. 2023 . Evidence-based guideline for the diagnosis and treatment of Mycoplasma pneumoniae pneumonia in children. Chin J Pediatr. 2024;62(12):1137–1144. [DOI] [PubMed] [Google Scholar]
  • 27.Munn Z, Moola S, Lisy K, et al. Methodological guidance for systematic reviews of observational epidemiological studies reporting prevalence and cumulative incidence data. Int J Evid Based Healthc. 2015;13(3):147–153. doi: 10.1097/XEB.0000000000000054 [DOI] [PubMed] [Google Scholar]
  • 28.Wells GA, Shea B, O'Connell D, et al. The Newcastle – Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses [cited 2025 January 16]. Available from: https://www.ohri.ca/programs/clinical_epidemiology/oxford.asp
  • 29.WHO . WHO regional offices [cited 2025 January 16]. Available from: https://www.who.int/about/who-we-are/regional-offices.
  • 30.Yamazaki T, Kenri T.. Epidemiology of Mycoplasma pneumoniae infections in Japan and therapeutic strategies for macrolide-resistant M. pneumoniae . Front Microbiol. 2016;7:693. doi: 10.3389/fmicb.2016.00693 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Kogoj R, Praprotnik M, Mrvič T, et al. Genetic diversity and macrolide resistance of Mycoplasma pneumoniae isolates from two consecutive epidemics in Slovenia. Eur J Clin Microbiol Infect Dis. 2018;37(1):99–107. doi: 10.1007/s10096-017-3106-5 [DOI] [PubMed] [Google Scholar]
  • 32.Kenri T, Okazaki N, Yamazaki T, et al. Genotyping analysis of Mycoplasma pneumoniae clinical strains in Japan between 1995 and 2005: type shift phenomenon of M. pneumoniae clinical strains. J Med Microbiol. 2008;57(Pt 4):469–475. doi: 10.1099/jmm.0.47634-0 [DOI] [PubMed] [Google Scholar]
  • 33.Kim EK, Youn YS, Rhim JW, et al. Epidemiological comparison of three Mycoplasma pneumoniae pneumonia epidemics in a single hospital over 10 years. Korean J Pediatr. 2015;58(5):172–177. doi: 10.3345/kjp.2015.58.5.172 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Lee JK, Lee JH, Lee H, et al. Clonal expansion of macrolide-resistant sequence type 3 Mycoplasma pneumoniae, South Korea. Emerg Infect Dis. 2018;24(8):1465–1471. doi: 10.3201/eid2408.180081 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Kim K, Jung S, Kim M, et al. Global trends in the proportion of macrolide-resistant Mycoplasma pneumoniae infections: a systematic review and meta-analysis. JAMA Netw Open. 2022;5(7):e2220949. doi: 10.1001/jamanetworkopen.2022.20949 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Kenri T, Suzuki M, Sekizuka T, et al. Periodic genotype shifts in clinically prevalent Mycoplasma pneumoniae strains in Japan. Front Cell Infect Microbiol. 2020;10:385. doi: 10.3389/fcimb.2020.00385 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Zhao F, Liu JR, Shi WX, et al. Antimicrobial susceptibility and genotyping of Mycoplasma pneumoniae isolates in Beijing, China, from 2014 to 2016. Antimicrob Resist Infect Control. 2019:8:18. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Chen Y, Jia X, Gao Y, et al. Increased macrolide resistance rate of Mycoplasma pneumoniae correlated with epidemic in Beijing, China in 2023. Front Microbiol. 2024;15:1449511. doi: 10.3389/fmicb.2024.1449511 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Smit PW, Eggink D, Paltansing S, et al. Mycoplasma pneumoniae MLST detected in the upsurge of pneumonia during the 2023 to 2024 winter season in the Netherlands. Sci Rep. 2025;15(1):6985. doi: 10.1038/s41598-025-88990-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Nilsson AC, Björkman P, Welinder-Olsson C, et al. Clinical severity of Mycoplasma pneumoniae (MP) infection is associated with bacterial load in oropharyngeal secretions but not with MP genotype. BMC Infect Dis. 2010;10:39. doi: 10.1186/1471-2334-10-39 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Li L, Ma J, Guo P, et al. Molecular beacon based real-time PCR p1 gene genotyping, macrolide resistance mutation detection and clinical characteristics analysis of Mycoplasma pneumoniae infections in children. BMC Infect Dis. 2022;22(1):724. doi: 10.1186/s12879-022-07715-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Zhang L, Chen Z, Shen Z, et al. Genotyping of Mycoplasma pneumoniae in nasopharyngeal aspirate from children with pneumonia in Zhejiang Province. Article in Chinese. Chin J Pediatr. 2011;49(10):750–754. [PubMed] [Google Scholar]
  • 43.Hung HM, Chuang CH, Chen YY, et al. Clonal spread of macrolide-resistant Mycoplasma pneumoniae sequence type-3 and type-17 with recombination on non-P1 adhesin among children in Taiwan. Clin Microbiol Infect. 2021;27(8):1169.e1–1169.e6. doi: 10.1016/j.cmi.2020.09.035 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Material
TEMI_A_2614742_SM4587.docx (2.7MB, docx)

Data Availability Statement

This meta-analysis did not require the collection of new data, but rather the analysis of previously published data. Extracted and aggregated data from published studies are available on reasonable request to the corresponding author.

Articles from Emerging Microbes & Infections are provided here courtesy of Taylor & Francis

RESOURCES