This meta-analysis adhered to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines. The study protocol was registered in the PROSPERO database (Registration No.: CRD420251003956).

A comprehensive literature search was conducted in PubMed, Web of Science, and Embase from their inception until September 30, 2024. Considering that in some regions “Shrimp” and “Prawn” are used interchangeably while in others they refer to distinct species, both terms were included. The search combined allergy-related terms (“Allergy”, “Hypersensitivity”, “Anaphylaxis”) with prevalence-related terms (“Prevalence”, “Incidence”, “Epidemiology”, “Risk factors”) using Boolean operators. The search was limited to studies on humans published in English.

In addition to the articles retrieved from the above databases, we also included studies mentioned in the reference lists of relevant articles. These studies were identified from known review articles and reference lists. To ensure the quality and scientific validity of the included studies, we primarily relied on research published in peer-reviewed journals. Therefore, we decided not to include gray literature in this study.

Two independent reviewers (Jiaqi Chen and Qiang Zhang) screened the retrieved studies according to predefined criteria, with disagreements resolved through discussion or consultation with a third reviewer (Yongli Ying). The inclusion criteria were as follows: 1. Study Design: Cross-sectional, cohort, or case-control studies. 2. Population: Individuals of all ages with reported SA prevalence. 3. Outcomes: Prevalence data based on one or more of the following: Self-reported symptoms, Physician diagnosis, Skin prick test (SPT), Serum specific IgE (sIgE), Food challenge tests. The exclusion criteria were: 1. Review articles, case reports, conference abstracts, and editorials. 2. Studies lacking relevant prevalence data. 3. Studies using duplicate datasets (only the most comprehensive data was included). 4. Studies that do not meet the definitions or criteria for skin prick test (SPT) or serum specific IgE (sIgE) diagnosis.

In this study, we adopted a comprehensive approach to identifying SA without distinguishing between IgE-mediated and non-IgE-mediated reactions or differentiating between food allergies and food intolerances. SA was defined based on the following criteria: 1. self-reported symptomatic SA: Individuals experiencing allergic symptoms after shrimp consumption. Referring to an individual experiencing allergic symptoms after shrimp consumption. 2. Self-reported physician-diagnosed SA: SA diagnosed by a physician as reported by the individual. 3. SPT- or sIgE-based SA: Elevated specific IgE levels (≥0.35 kU/L) or a positive skin prick test (SPT) result (wheal ≥ 3 mm larger than the control). 4. SPT- or sIgE-based symptomatic SA: Positive SPT or sIgE results accompanied by allergic symptoms. and 5. Food challenge-confirmed SA: SA established through open, single-blind, or double-blind placebo-controlled food challenge tests.

Two independent reviewers (Jiaqi Chen and Xuying Zhang) extracted data using a pre-designed data extraction form, including study characteristics (first author, publication year, country, study design, sample size, and age group) and outcome measures (prevalence rates and diagnostic methods).

Quality assessment was performed using different appraisal tools based on study design. For cross-sectional studies, the Joanna Briggs Institute (JBI) Critical Appraisal Checklist was used, with studies scoring ≥6 considered high quality, 4–5 as moderate quality, and ≤3 as low quality. For cohort studies, the Newcastle-Ottawa Scale (NOS) was applied, with studies scoring ≥7 classified as high quality, 4–6 as moderate quality, and ≤3 as low quality. Disagreements between reviewers were resolved through discussion.

Meta-analysis was conducted using Comprehensive Meta-Analysis (CMA) software. Heterogeneity was assessed using Cochran's Q test and the I² statistic, with I² > 50% indicating substantial heterogeneity. Random-effects models were applied to pool the prevalence of SA in the presence of high heterogeneity, while fixed-effects models were used for studies with low heterogeneity (I² < 50% and P > 0.05). Publication bias was evaluated using Egger's regression test (8) and Funnel plot. A P-value of <0.05 was considered statistically significant for all tests.

For prevalence estimates, we did not differentiate between lifetime and point prevalence. If a study reported both, we used lifetime prevalence for analysis. We prioritized extracting raw data from studies whenever possible; if raw data were unavailable, weighted data were used instead. To estimate the true prevalence of SA, we adjusted for participant dropout in some studies based on the proportion of individuals who completed the study. If a study reported both shrimp and prawn allergy prevalence, we included the data with the larger sample size.

To analyze age-based differences, we categorized study participants as children (<18 years old) or adults (≥18 years old). For studies that included both age groups, classification was based on the predominant age distribution within the study.

A total of 823 articles were identified through the search process, of which 40 studies were ultimately included in the systematic review. The detailed selection process is shown in Figure 1. Among these studies, 36 adopted a cross-sectional design, while 4 employed a cohort design. Regarding study populations, 23 studies focused on children, 12 on adults, and 5 included multiple age groups. The quality assessment revealed that 9 studies were of high quality, 17 were of moderate quality, and 14 were of low quality (Table 1).

Among the included studies, 23 assessed SA using self-reported symptoms, with a total sample size of 167,835. The meta-analysis yielded a pooled prevalence of SA of 1.90% (95% CI: 1.44%–2.51%) (Figure 2A). Heterogeneity analysis indicated a high level of heterogeneity among the studies (I² = 97.67%, P < 0.05), suggesting considerable variability in prevalence estimates. Consequently, a random-effects model was applied for the meta-analysis.

Age-stratified analysis showed that the prevalence of self-reported symptomatic SA was 1.29% (95% CI: 0.97%–1.73%) in children, lower than that in adults (3.39%; 95% CI: 2.47%–4.64%). Region-specific analysis revealed prevalence rates of 3.75% (95% CI: 2.48%–5.64%) in Asia, 1.05% (95% CI: 0.62%–1.79%) in Europe, 1.95% (95% CI: 1.42%–2.68%) in America, and 1.06% (95% CI: 0.46%–2.42%) in Africa (Figure 2B).

Egger's regression test showed no evidence of publication bias (P = 0.92)); see Supplementary Figure S1 for the funnel plot.

A total of 9 studies assessed SA based on self-reported physician diagnosis, with a combined sample size of 51,264. The meta-analysis yielded a pooled prevalence of 1.94% (95% CI: 1.10%–3.41%) (Figure 3A). The heterogeneity analysis indicated significant heterogeneity (I² = 99.21%, P < 0.05), necessitating the use of a random-effects model.

Age-stratified analysis revealed that the prevalence in children was 1.85% (95% CI: 1.04%–3.25%), lower than that in adults (3.03%; 95% CI: 1.68%–5.39%). Regional analysis showed prevalence rates of 3.39% (95% CI: 1.88%–6.76%) in Asia and 1.30% (95% CI: 0.62%–2.67%) in America (Figure 3B).

Egger's regression test showed no evidence of publication bias (P = 0.36); see Supplementary Figure S2 for the funnel plot.

Nine studies assessed SA based on SPT- or sIgE-testing, with a total sample size of 38,459. The pooled prevalence was 2.76% (95% CI: 1.91%–3.97%) (Figure 4A). Heterogeneity analysis revealed high variability (I² = 97.01%, P < 0.05), necessitating the use of a random-effects model.

Age-stratified analysis showed a prevalence of 2.61% (95% CI: 1.79%–3.81%) in children, lower than that in adults (4.42%; 95% CI: 2.69%–7.18%). Regional analysis revealed prevalence rates of 1.36% (95% CI: 0.65%–2.85%) in Asia and 5.23% (95% CI: 3.22%–8.40%) in America (Figure 4B).

Egger's regression test showed no evidence of publication bias (P = 0.29); see Supplementary Figure S3 for the funnel plot.

Nine studies assessed SA using SPT- or sIgE-testing combined with symptoms, with a total sample size of 30,779. The pooled prevalence was 0.43% (95% CI: 0.24%–0.78%) (Figure 5A). Heterogeneity analysis showed moderate variability (I² = 62.86%, P < 0.05), necessitating the use of a random-effects model.

Age-stratified analysis revealed that the prevalence in children was 0.37% (95% CI: 0.15%–0.87%), slightly lower than that in adults (0.47%; 95% CI: 0.14%–1.52%). Regionally, the prevalence was 0.68% (95% CI: 0.25%–1.86%) in Asia and 0.18% (95% CI: 0.05%–0.56%) in Europe (Figure 5B).

Egger's regression test showed no evidence of publication bias (P = 0.71); see Supplementary Figure S4 for the funnel plot.

Seven studies assessed SA using food challenge testing, with a total sample size of 4,652. The meta-analysis yielded a pooled prevalence of 0.50% (95% CI: 0.30%–0.85%) (Figure 6A). Heterogeneity analysis indicated moderate variability (I² = 48.49%, P > 0.05), justifying the use of a fixed-effects model.

Age-stratified analysis showed that the prevalence in children was 0.59% (95% CI: 0.31%–1.14%), slightly higher than that in adults (0.45%; 95% CI: 0.19%–1.03%). Regional analysis indicated a prevalence of 0.68% (95% CI: 0.35%–1.33%) in Asia and 0.31% (95% CI: 0.13%–0.72%) in Europe (Figure 6B).

Egger's regression test showed no evidence of publication bias (P = 0.25); see Supplementary Figure S5 for the funnel plot.