Microplastic Contamination of Seafood Intended for Human Consumption: A Systematic Review and Meta-Analysis

Background: Microplastics (MPs) have contaminated all compartments of the marine environment including biota such as seafood; ingestion from such sources is one of the two major uptake routes identified for human exposure. Objectives: The objectives were to conduct a systematic review and meta-analysis of the levels of MP contamination in seafood and to subsequently estimate the annual human uptake. Methods: MEDLINE, EMBASE, and Web of Science were searched from launch (1947, 1974, and 1900, respectively) up to October 2020 for all studies reporting MP content in seafood species. Mean, standard deviations, and ranges of MPs found were collated. Studies were appraised systematically using a bespoke risk of bias (RoB) assessment tool. Results: Fifty studies were included in the systematic review and 19 in the meta-analysis. Evidence was available on four phyla: mollusks, crustaceans, fish, and echinodermata. The majority of studies identified MP contamination in seafood and reported MP content <1 MP/g, with 26% of studies rated as having a high RoB, mainly due to analysis or reporting weaknesses. Mollusks collected off the coasts of Asia were the most heavily contaminated, coinciding with reported trends of MP contamination in the sea. According to the statistical summary, MP content was 0–10.5 MPs/g in mollusks, 0.1–8.6 MPs/g in crustaceans, 0–2.9 MPs/g in fish, and 1 MP/g in echinodermata. Maximum annual human MP uptake was estimated to be close to 55,000 MP particles. Statistical, sample, and methodological heterogeneity was high. Discussion: This is the first systematic review, to our knowledge, to assess and quantify MP contamination of seafood and human uptake from its consumption, suggesting that action must be considered in order to reduce human exposure via such consumption. Further high-quality research using standardized methods is needed to cement the scientific evidence on MP contamination and human exposures. https://doi.org/10.1289/EHP7171


Table of Contents
Code for R used in the meta-analysis Table S1. Search strategy for MEDLINE and EMBASE (OVID) with Medical Subject Headings (MeSH) searching. Table S2. Search strategy for Web of Science. Table S3. Formulae for combining groups (Higgins et al., 2011: Table 7.7.a). Table S4. Risk of bias (RoB) assessment tool template. Table S5. Particle extraction procedure details.    Table S9. Composition identification process characteristics for molluscan studies. Table S10. Meta-analysis data pooled by family of molluscs. Table S11. Molluscan studies subgroup analysis results.  S13. Composition identification process characteristics for fish studies. Table S14. GRADE certainty framework assessment for seafood studies. Table S15. Estimation of microplastic (MP) particles yearly uptake from the consumption of seafood across countries.   . Spearman correlation analysis between (A) the amount of MPs/g in mussels and the percentage of the particles analysed for composition, (B) the amount of MPs/g in mussels and the number of the particles analysed for composition. R is the correlation coefficient with the corresponding p value. Figure S4. Influence analysis forest plots of random-effects model for clams using the leave-oneout method, sorted by (A) effect size estimate, expressed as microplastics per g (MPs/g) and 95% confidence interval (CI) and (B) heterogeneity expressed in I 2 . The pooled effect is recalculated each time leaving out one study. In both figures results are ordered from low to high. Figure S5. Forest plot for random-effects model results for clams excluding two high RoB studies (Baechler et al., 2020;Wang et al., 2019). The x axis represents the standardized mean difference (SMD) expressed in microplastics per gram (MPs/g). TE is the MP content reported by each study and seTE is the calculated standard error (SE). The vertical line is the line of null effect where MP content is 0. The grey boxes represent the pooled effect estimate and the whiskers the CI 95%. The size of the boxes is proportional to the study weight. The diamond is the combined point estimate and CI 95%, and the dotted line is the overall pooled effect. The black box represents the 95% prediction interval. Figure S6. Influence analysis forest plots of random-effects model for mussels using the leaveone-out method, sorted by (A) effect size estimate, expressed as microplastics per g (MPs/g) and 95% confidence interval (CI) and (B) heterogeneity expressed in I 2 . The pooled effect is recalculated each time leaving out one study. In both figures results are ordered from low to high. Li J. et al (2018) a, samples collected from environment, Li J. et al (2018) b, samples collected from market. Figure S7. Influence analysis Baujat Plot of random-effects model for mussels. The horizontal axis illustrates statistical heterogeneity as measured by Cochran's Q statistic. The vertical axis illustrates the influence on the pooled result. Li J. et al (2018) a, samples collected from environment, Li J. et al (2018) b, samples collected from market. Figure S8. Influence analysis forest plots of random-effects model for oysters using the leaveone-out method, sorted by (A) effect size estimate, expressed as microplastics per g (MPs/g) and 95% confidence interval (CI) and (B) heterogeneity expressed in I 2 . The pooled effect is recalculated each time leaving out one study. In both figures results are ordered from low to high. Figure S9. Forest plot for random-effects model results for all molluscan families excluding the five high RoB studies (Baechler et al., 2020;Hermabessiere et al., 2019;Wang et al., 2019;Webb et al., 2019;Zhao et al., 2018) . The x axis represents the standardized mean difference (SMD) expressed in microplastics per gram (MPs/g). TE is the MP content reported by each study and seTE is the calculated standard error (SE). The vertical line is the line of null effect where MP content is 0. The grey boxes represent the pooled effect estimate and the whiskers the confidence interval (CI) 95%. The size of the boxes is proportional to the study weight. The diamond is the combined point estimate and CI 95%, and the dotted line is the overall pooled effect. The black box represents the 95% prediction interval. Li J. et al (2018) a, samples collected from environment; Li J. et al (2018) b, samples collected from market.    . Predicted global yearly maximum microplastic (MP) particles uptake through crustacean consumption. The data were calculated using the FAO (2020) consumption data for crustacean per country and the maximum MPs/g content of crustacean derived from the statistical summary results herein. The numerical data is shown in Table S15. MP data have been classified in ten categories using quantile classification for illustration purposes. The hatched areas illustrate countries for which data on mollusc consumption were not available. Figure S14. Predicted global yearly maximum microplastic (MP) particles uptake through fish consumption. The data were calculated using the FAO (2020) consumption data for fish per country and the maximum MPs/g content of fish derived from the statistical summary results herein. The numerical data is shown in Table S15. MP data have been classified in ten categories using quantile classification for illustration purposes. The hatched areas illustrate countries for which data on fish consumption were not available.

Code for R used in the meta-analysis
#fit the random effects model using the DerSimonian-Laird estimator for tau 2 : microfiber* 10 microfibre* 11 micro-fiber* 12 micro-fibre* 13 particle* 14 particle size/ 15 pellet* 16 fragment* 17 film* 18 filament* 19 rubber/ 20 5 and 6 21 5 and 7 22 5 and 8 23 5 and 9 24 5 and 10 25 5 and 11 26 5 and 12 27 5 and 13 28 5 and 14 29 5 and 15 30 5 and 16 31 5 and 17 32 5 and 18 33 5 and 19 34 1 or 2 or 3 or 4 or 20 or 21 or 22 or 23 or 24 or 25 or 26 or 27 or 28 or 29 or 30 or 31 or 32 or 33 35 food quality/ or food dye/ or food ingredient/ or canned food/ or food packaging/ or food contamination/ or food industry/ or food insecurity/ or cooked food/ or food safety/ or food analysis/ or food chain/ or fast food/ or dried food/ or sea food/ or food handling/ or food security/ or food/ 36 water table/ or drinking water/ or water quality/ or tap water/ or water pollutant/ or water contamination/ or water pollution/ 37 sea food/ 38 fish/  Table S3. Formulae for combining groups (Higgins et al., 2011: Table 7.7.a).

RoB tool additional explanation
For the majority of the items in the tool the rating 'high' and 'low' is based on a yes/no answer or a numerical value. The rating 'unclear' is assigned when the study does not report sufficient information to make a judgment or when the associated risk is unknown. In order to achieve maximum transparency all items are discussed herein in detail.

Study design
1. Is the design appropriate for the questions of the study? Yes/No Explanation: The design must be observational (non-analytic) which is the appropriate design for an environmental study in this field.

Has the interpretation of the results been based on the outcomes of the analysis?
Explanation: The conclusions should be based on the reported results from the experimental analysis and statistical analysis.

Quality of reporting
Methodology 20. Have the methods used in the study been reported in detail?
Explanation: All methods throughout the study protocol should be reported in detail in the main paper or in supporting material so that the study can duplicated and verified. The answer to this question also draws from the previous answers given in the tool in the domains of sampling and analysis.

Has the study recognized limitations?
Explanation: Authors should report how their results relate to the wider picture of their field and whether have identified important limitations. Understanding and reporting limitations relates to both internal and external validity of the study.   The table shows the rating (high, unclear, low) for the four domains and the overall rating for each study expressed in number of studies and percentage of studies. Table S7. RoB rating for all seafood studies.
The table shows the rating for the four domains and the overall rating for each study. Red (-) indicates high RoB, green (+) indicates low RoB and yellow (?) indicates unclear RoB (Unclear RoB is given to a study when substantial information to make an informed assessment have not been reported).      For the meta-analysis, the evidence will not be downgraded as the results of the studies that are rated as of high RoB will not be included in the summary of evidence.
For the statistical summary, in the molluscs the evidence will not be downgraded as only 29% of the studies have been rated as of high RoB. For the crustacean the evidence will be downgraded as 80% of the studies have been rated as of high RoB. For the fish, the evidence will not be downgraded as only 25% of the studies have been rated as of high RoB. For echinodermata, the evidence will not be downgraded.

Meta-analysis
Heterogeneity across studies was high.

Statistical summary/ narrative analysis
The range of MP content was high for all phyla.
Downgrade the meta-analysis evidence due to high heterogeneity as measured by I 2 , Chi 2 , and corresponding p value.

Indirectness (downgrading)
The majority of the studies measured the outcome addressed by the question of the review.
Will not downgrade evidence.

Imprecision (downgrading)
The sample size n for the studies varied from 5 to 760 organisms. 95% CI for the studies in the metaanalysis are acceptable for the majority of the studies.
Will not downgrade evidence.

Publication bias (downgrading)
Publication bias was not substantial.
For the meta-analysis the funnel plots did not detect substantial bias. For the statistical summary the country of origin did not point to publication bias. The evidence will not be downgraded.

Large effects (upgrading)
The estimates are not large, consistent and confident enough to allowing an upgrade of the evidence.
Will not upgrade the evidence.

Dose response (upgrading)
Dose response cannot, at this point, be accessed for these studies.
Not applicable.

Opposing plausible residual bias and confounding (upgrading)
All confounders have been considered.
The estimate of effect was controlled for the following possible confounders: contamination of samples post sampling, misrepresentation of particles as MPs. The use of procedural blank samples and the use of a technique to identify the chemical composition of the particles are the built-in fail-safes of the review. The certainty of the evidence will be increased for all studies.
Note: CI, confidence interval; MPs, microplastics; RoB, risk of bias. molluscs, crustacean and fish corresponding to the yearly consumption data (FAO, 2020). The MP content of seafood is based on the statistical summary results (Table 5). Min, minimum microplastic particles uptake per year; max, maximum microplastic particles uptake per year.

Studies included in meta-analysis (n = 19)
Records identified through rerun of database searching (n = 1950); Web of Science (n=845), MEDLINE (n = 909), EMBASE (n = 521) Records after duplicates removed (n = 1374) Titles and abstracts screened (n = 1374) Records excluded (n = 1308) Full-text articles assessed for eligibility (n = 66) Full-text articles excluded, with reasons (n = 50) c Studies included from rerun (n = 16) Figure S1. PRISMA flow chart a The reference lists of the reviews that were discovered by the search were screened, as well as the reference lists of published relevant reports.  Figure S2. Phylogenetic tree for the molluscan phylum. The species that are included in each study are presented in Tables 1 and 2. Figure S4. Spearman correlation analysis between (A) the amount of MPs/g in mussels and the percentage of the particles analysed for composition, (B) the amount of MPs/g in mussels and the number of the particles analysed for composition.
A B Figure S3. Spearman correlation analysis between (A) the amount of MPs/g in mussels and the percentage of the particles analysed for composition, (B) the amount of MPs/g in mussels and the number of the particles analysed for composition. R is the correlation coefficient with the corresponding p value. Overall hetereogeneity contr ibution 0.     L i J . e t a l.
( 2 0 1 5 ) L i J . e t a l.  L i J . e t a l.
( 2 0 1 8 ) a L i J . e t a l.
( 2 0 1 8 ) b N a m e t a l.
(2 0 1 9 ) P h u o n g e t a l.
( 2 0 1 8 a ) V a n C a u w e n b e r g h e a n d L i J . e t a l.
( 2 0 1 8 ) a L i J . e t a l.
( 2 0 1 8 ) b N a m e t a l.
( 2 0 1 9 ) P h u o n g e t a l.
( 2 0 1 8 a ) V a n C a u w e n b e r g h e a n d L i J . e t a l.
( 2 0 1 8 ) a L i J . e t a l.
( 2 0 1 8 ) b N a m e t a l.
( 2 0 1 9 ) P h u o n g e t a l. ( 2 0 1 8 a ) V a n C a u w e n b e r g h e a n d L i J . e t a l.
( 2 0 1 8 ) a L i J . e t a l.
( 2 0 1 8 ) b N a m e t a l.
( 2 0 1 9 ) P h u o n g e t a l.
( 2 0 1 8 a ) V a n C a u w e n b e r g h e a n d W e b b e t a l.
( 2 0 1 9 ) Z h a o e t a l.
( 2 0 1 8 Figure S6. Influence analysis forest plots of random-effects model for mussels using the leave-oneout method, sorted by (A) effect size estimate, expressed as microplastics per g (MPs/g) and 95% confidence interval (CI) and (     Omitting Van Cauwenberghe and Janssen (     consumption. The data were calculated using the FAO (2020) consumption data for crustacean per country and the maximum MPs/g content of crustacean derived from the statistical summary results herein. The numerical data is shown in Table S15. MP data have been classified in ten categories using quantile classification for illustration purposes. The hatched areas illustrate countries for which data on mollusc consumption were not available. Figure S14. Predicted global yearly maximum microplastic (MP) particles uptake through fish consumption. The data were calculated using the FAO (2020) consumption data for fish per country and the maximum MPs/g content of fish derived from the statistical summary results herein. The numerical data is shown in Table S15. MP data have been classified in ten categories using quantile classification for illustration purposes. The hatched areas illustrate countries for which data on fish consumption were not available.