An Umbrella Review of Meta-Analyses Evaluating Associations between Human Health and Exposure to Major Classes of Plastic-Associated Chemicals

Abstract

Background: Epidemiological research investigating the impact of exposure to plastics, and plastic-associated chemicals, on human health is critical, especially given exponentially increasing plastic production. In parallel with increasing production, academic research has also increased exponentially both in terms of the primary literature and ensuing systematic reviews with meta-analysis. However, there are few overviews that capture a broad range of chemical classes to present a state of play regarding impacts on human health. Methods: We undertook an umbrella review to review the systematic reviews with meta-analyses. Given the complex composition of plastic and the large number of identified plastic-associated chemicals, it was not possible to capture all chemicals that may be present in, and migrate from, plastic materials. We therefore focussed on a defined set of key exposures related to plastics. These were microplastics, due to their ubiquity and potential for human exposure, and the polymers that form the matrix of consumer plastics. We also included plasticisers and flame retardants as the two classes of functional additive with the highest concentration ranges in plastic. In addition, we included bisphenols and per- and polyfluoroalkyl substances (PFAS) as two other major plastic-associated chemicals with significant known exposure through food contact materials. Epistemonikos and PubMed were searched for systematic reviews with meta-analyses, meta-analyses, and pooled analyses evaluating the association of plastic polymers, particles (microplastics) or any of the selected groups of high-volume plastic-associated chemicals above, measured directly in human biospecimens, with human health outcomes. Results: Fifty-two systematic reviews were included, with data contributing 759 meta-analyses. Most meta-analyses (78%) were from reviews of moderate methodological quality. Across all the publications retrieved, only a limited number of plastic-associated chemicals within each of the groups searched had been evaluated in relevant meta-analyses, and there were no meta-analyses evaluating polymers, nor microplastics. Synthesised estimates of the effects of plastic-associated chemical exposure were identified for the following health outcome categories in humans: birth, child and adult reproductive, endocrine, child neurodevelopment, nutritional, circulatory, respiratory, skin-related and cancers. Bisphenol A (BPA) is associated with decreased anoclitoral distance in infants, type 2 diabetes (T2D) in adults, insulin resistance in children and adults, polycystic ovary syndrome, obesity and hypertension in children and adults and cardiovascular disease (CVD); other bisphenols have not been evaluated. Phthalates, the only plasticisers identified, are associated with spontaneous pregnancy loss, decreased anogenital distance in boys, insulin resistance in children and adults, with additional associations between certain phthalates and decreased birth weight, T2D in adults, precocious puberty in girls, reduced sperm quality, endometriosis, adverse cognitive development and intelligence quotient (IQ) loss, adverse fine motor and psychomotor development and elevated blood pressure in children and asthma in children and adults. Polychlorinated biphenyls (PCBs), polybrominated diphenyl ethers (PBDEs) but not other flame retardants, and some PFAS were identified and are all associated with decreased birth weight. In general populations, PCBs are associated with T2D in adults and endometriosis, bronchitis in infants, CVD, non-Hodgkin's lymphoma (NHL) and breast cancer. In PCB-poisoned populations, exposure is associated with overall mortality, mortality from hepatic disease (men), CVD (men and women) and several cancers. PBDEs are adversely associated with children's cognitive development and IQ loss. PBDEs and certain PFAS are associated with changes in thyroid function. PFAS exposure is associated with increased body mass index (BMI) and overweight in children, attention deficit hyperactive disorder (ADHD) in girls and allergic rhinitis. Potential protective associations were found, namely abnormal pubertal timing in boys being less common with higher phthalate exposure, increased high-density lipoprotein (HDL) with exposure to mono(2-ethyl-5-oxohexyl) phthalate (MEOHP) and reduced incidence of chronic lymphocytic lymphoma (a subtype of NHL) with PCB exposure. Conclusions: Exposure to plastic-associated chemicals is associated with adverse outcomes across a wide range of human health domains, and every plastic-associated chemical group is associated with at least one adverse health outcome. Large gaps remain for many plastic-associated chemicals. Recommendations: For research, we recommend that efforts are harmonised globally to pool resources and extend beyond the chemicals included in this umbrella review. Priorities for primary research, with ensuing systematic reviews, could include micro- and nanoplastics as well as emerging plastic-associated chemicals of concern such as bisphenol analogues and replacement plasticisers and flame retardants. With respect to chemical regulation, we propose that safety for plastic-associated chemicals in humans cannot be assumed at market entry. We therefore recommend that improved independent, systematic hazard testing for all plastic-associated chemicals is undertaken before market release of products. In addition because of the limitations of laboratory-based testing for predicting harm from plastic in humans, independent and systematic post-market bio-monitoring and epidemiological studies are essential to detect potential unforeseen harms.

Keywords: Evidence synthesis; Human health; Plastic-associated chemicals; Plastics; Umbrella review.

Figure 1

PRISMA flow diagram [35] presenting process of study identification, selection and final inclusion in the review project and the outcomes reported in this manuscript.

Figure 2

Harvest plot of exposure to plastic-associated chemicals and birth outcomes. Plastic-associated chemicals included are bisphenol A (BPA) (pink); phthalate monoester metabolites (blue), encompassing monomethyl phthalate (MMP), monoethyl phthalate (MEP), mono-n-butyl phthalate (MnBP), monoisobutyl phthalate (MiBP), monobenzyl phthalate (MBzP), mono(2-ethylhexyl) phthalate (MEHP), mono(2-ethyl-5-hydroxyhexyl) phthalate (MEHHP), mono(2-ethyl-5-oxohexyl) phthalate (MEOHP), mono(2-ethyl-5-carboxypentyl) phthalate (MECPP), and molar sum of the di(2-ethylhexyl) phthalate metabolites (∑DEHP); flame retardants (green) encompassing polychlorinated biphenyl (PCB), polybrominated diphenyl ethers (PBDEs), 2,2’,4,4’-tetrabromodiphenyl ether (BDE-47), 2,2’,4,4’,5-pentabromodiphenyl ether (BDE-99), 2,2’,4,4’,6-pentabromodiphenyl ether (BDE-100), 2,2’,4,4’,5,5’-hexabromodiphenyl ether (BDE-153); and per- and polyfluoroalkyl substances (PFAS) (orange), encompassing perfluorooctanoic acid (PFOA) and perfluorooctane sulfonate (PFOS). Outcomes are either dichotomous (†) or measured on a continuous scale (‡). Outcome measures include ‡birth weight, ‡birth length, ‡head circumference, ‡ponderal index, ‡gestational age, †secondary sex ratio and †spontaneous pregnancy loss. Each bar represents an individual effect estimate from the corresponding review, which is indicated by the number below each bar. The height of the bar represents the quality score of the review assessed using the AMSTAR tool. Low quality reflects a score of 1–4, moderate quality a score of 5–8 and high quality a score of 9–11. Dark filled bars represent the main analyses of each review; light filled bars represent sub-group analyses. Bars have been assigned as an increase or decrease (columns) in the measure where the change is statistically significant. Remaining bars appearing under ‘no change’ indicate direction of effect as an increase (>), no clear trend (–) (the estimate of relative risk was 1 or regression coefficient or mean difference was 0), or decrease (<) in the measure or risk estimate.

Figure 3

Harvest plot of exposure to plastic-associated chemicals and child reproductive outcome measures. Plastic-associated chemicals included are bisphenol A (BPA) (pink); and phthalate diesters diethylhexyl phthalate (DEHP) and di-n-butyl phthalate (DnBP) and monoester metabolites (blue), including monomethyl phthalate (MMP), monoethyl phthalate (MEP), mono-n-butyl phthalate (MnBP), monobenzyl phthalate (MBzP), mono(2-ethylhexyl) phthalate (MEHP), mono(2-ethyl-5-hydroxyhexyl) phthalate (MEHHP), mono(2-ethyl-5-oxohexyl) phthalate (MEOHP), mono(2-ethyl-5-carboxypentyl) phthalate (MECPP), and mono (3-carboxypropyl) phthalate (MCPP). Outcomes are either dichotomous (†) or measured on a continuous scale (‡). Outcomes measured include †precocious puberty, ‡anogenital distance measured by anoclitoral and anofourchette distance in girls and anoscrotal and anopenile distance in boys, †abnormal timing/age of puberty/early puberty measured by pubarche, menarche, thelarche and testicular volume. Each bar represents an individual effect estimate from the corresponding review, which is indicated by the number below each bar. The height of the bar represents the quality score of the review assessed using the AMSTAR tool. Moderate quality reflects a score of 5–8. Dark filled bars represent the main analyses of each review. Bars have been assigned as an increase or decrease (columns) in the measure where the change is statistically significant. Remaining bars appearing under ‘no change’ indicate direction of effect as an increase (>), or decrease (<) in the measure or risk estimate.

Figure 4

Harvest plot of exposure to plastic-associated chemicals and adult reproductive outcome measures. Plastic-associated chemicals included are bisphenol A (BPA) (pink); phthalate monoester metabolites (blue), including monomethyl phthalate (MMP), monoethyl phthalate (MEP), mono-n-butyl phthalate (MnBP), monobenzyl phthalate (MBzP), mono(2-ethylhexyl) phthalate (MEHP), mono(2-ethyl-5-hydroxyhexyl) phthalate (MEHHP), mono(2-ethyl-5-oxohexyl) phthalate (MEOHP), mono(2-ethyl-5-carboxypentyl) phthalate (MECPP), molar sum of the di(2-ethylhexyl) phthalate metabolites (∑DEHP), and mono (3-carboxypropyl) phthalate (MCPP); and flame retardants (green) encompassing polychlorinated biphenyl (PCB). Outcomes are either dichotomous (†) or measured on a continuous scale (‡). Outcomes measured include †endometriosis, †sperm concentration, ‡†sperm motility, †sperm morphology, †sperm volume, ‡sperm motion measured via straight line velocity, curvilinear velocity, linearity, and ‡sperm DNA damage measured via comet assay (comet extent), comet assay (percentage [%] DNA in tail) and comet assay (tail distributed moment). Each bar represents an individual effect estimate from the corresponding review, which is indicated by the number below each bar. The height of the bar represents the quality score of the review assessed using the AMSTAR tool. Low quality reflects a score of 1–4 and moderate quality a score of 5–8. Dark filled bars represent the main analyses of each review; light filled bars represent sub-group analyses. Bars have been assigned as an increase or decrease (columns) in the measure where the change is statistically significant. Remaining bars appearing under ‘no change’ indicate direction of effect as an increase (>), no clear trend (–) (the estimate of relative risk was 1 or regression coefficient or mean difference was 0), or decrease (<) in the measure or risk estimate.

Figure 5

Harvest plot of exposure to plastic-associated chemicals and endocrine outcome measures. Plastic-associated chemicals included are bisphenol A (BPA) (pink); phthalate monoester metabolites (blue), encompassing monomethyl phthalate (MMP), monoethyl phthalate (MEP), monoisobutyl phthalate (MiBP), monobenzyl phthalate (MBzP), mono(2-ethylhexyl) phthalate (MEHP), mono(2-ethyl-5-hydroxyhexyl) phthalate (MEHHP), mono(2-ethyl-5-oxohexyl) phthalate (MEOHP), mono(2-ethyl-5-carboxypentyl) phthalate (MECPP), mono(3-carboxypropyl) phthalate (MCPP) and molar sum of the di(2-ethylhexyl) phthalate metabolites (∑DEHP); flame retardants (green) encompassing polychlorinated biphenyl (PCB), 2,3’,4,4’,5-pentachlorobiphenyl (PCB 118) group (gp) II, 2,2’,3,4,4’,5’-hexachlorobiphenyl (PCB 138) (gp II), 2,2’,4,4’,5,5’-hexachlorobiphenyl (PCB 153) (gp III), 2,2’,3,4,4’,5,5’-heptachlorobiphenyl PCB 180) (gp III), polybrominated diphenyl ethers (PBDEs); and per- and polyfluoroalkyl substances (PFAS) (orange), encompassing perfluorohexane sulfonate (PFHxS), perfluorooctanoic acid (PFOA), and perfluorooctane sulfonate (PFOS). Outcomes are either dichotomous (†) or measured on a continuous scale (‡). Outcomes measured include thyroid function measured by levels of ‡free thyroxine (fT4), ‡thyroxine (TT4), ‡thyroid-stimulating hormone (TSH), and ‡triiodothyronine (T3), †type 2 diabetes (T2D), ‡insulin resistance (HOMA-IR), ‡fasting insulin, ‡2-hour (hr) insulin, ‡fasting glucose and ‡2-hour glucose. Each bar represents an individual effect estimate from the corresponding review, which is indicated by the number below each bar. The height of the bar represents the quality score of the review assessed using the AMSTAR tool. Low quality reflects a score of 1–4, moderate quality a score of 5–8 and high quality a score of 9–11. Dark filled bars represent the main analyses of each review; light filled bars represent sub-group analyses. Bars have been assigned as an increase or decrease (columns) in the measure where the change is statistically significant. Remaining bars appearing under ‘no change’ indicate direction of effect as an increase (>), no clear trend (–) (the estimate of relative risk was 1 or regression coefficient or mean difference was 0), or decrease (<) in the measure or risk estimate.

Figure 6

Harvest plot of exposure to plastic-associated chemicals and children’s neurodevelopmental outcome measures. Plastic-associated chemicals included are phthalates (blue) where exposure was determined based on monoester metabolites monoethyl phthalate (MEP), mono-n-butyl phthalate (MnBP), monoisobutyl phthalate (MiBP), monobenzyl phthalate (MBzP), molar sum of all di(2-ethylhexyl) phthalate metabolites measured (∑DEHP), and best single measure of metabolite(s) of di(2-ethylhexyl) phthalate (DEHP m.); flame retardants (green) including polybrominated diphenyl ethers (PBDEs) where exposure was determined based on a prevalent congener 2,2’,4,4’-tetrabromodiphenyl ether (BDE-47); and per- and polyfluoroalkyl substances (PFAS) (orange) including perfluorooctanoic acid (PFOA) and perfluorooctane sulfonate (PFOS). Outcome measures are either dichotomous (†) or measured on a continuous scale (‡). Outcomes include ‡Cognitive Development and Intelligence Quotient (IQ) (measured on the Mental Development Index (MDI) of the Bayley Scales of Infant Development, 2nd ed. (BSID-II), Cognitive Development subscale of the Bayley Scales of Infant and Toddler Development, 3rd ed. (Bayley-III), General Cognitive Scale (GCS) of the McCarthy Scales of Children’s Abilities (MSCA), and Full Scale IQ (FSIQ) of the Wechsler Preschool & Primary Scale of Intelligence (WPPSI) or Wechsler Intelligence Scale for Children (WISC)); ‡Fine Motor/Psychomotor Development (measured on the Psychomotor Development Index (PDI) of BSID-II, and Fine Motor subscale of Bayley-III) and †Attention Deficit Hyperactive Disorder (ADHD) (measured with the Attention Problems Syndrome Scale of the Child Behaviour Checklist (CBCL), the Hyperactivity/Inattention subscale of the Strengths and Difficulties Questionnaire (SDQ) and the ADHD Diagnostic and Statistical Manual of Mental Disorder 4th ed (DSM-IV)). Each bar represents an individual effect estimate from the corresponding review, which is indicated by the number below each bar. The height of the bar represents the quality score of the review assessed using the AMSTAR tool. Low quality reflects a score of 1–4, moderate (mod) quality a score of 5–8 and high quality a score of 9–11. Dark filled bars represent the primary analyses of each review; unfilled bars represent sub-group analyses. Bars have been assigned as an increase or decrease (columns) in the measure where the change is statistically significant. Remaining bars appearing under ‘no change’ indicate direction of effect as an increase (>), no clear trend (–) (the estimate of relative risk was 1 or regression coefficient or mean difference was 0), or decrease (<) in the measure or risk estimate.

Figure 7

Harvest plot of exposure to plastic-associated chemicals and nutritional outcome measures. Plastic-associated chemicals included are bisphenol A (BPA) (pink) and phthalate monoester metabolites (blue), including monomethyl phthalate (MMP), monoethyl phthalate (MEP), mono-n-butyl phthalate (MnBP), monoisobutyl phthalate (MiBP), monobenzyl phthalate (MBzP), mono(2-ethylhexyl) phthalate (MEHP), mono(2-ethyl-5-hydroxyhexyl) phthalate (MEHHP), mono(2-ethyl-5-oxohexyl) phthalate (MEOHP), mono(2-ethyl-5-carboxypentyl) phthalate (MECPP), mono-n-octyl phthalate (MnOP) and mono (3-carboxypropyl) phthalate (MCPP). Outcome measures are either dichotomous (†) or measured on a continuous scale (‡). Outcomes measured include †obesity including abdominal obesity and generalized obesity, †overweight including generalized overweight, ‡Body Mass Index (BMI) and ‡BMI z score, †elevated waist circumference and ‡waist circumference. Each bar represents an individual effect estimate from the corresponding review, which is indicated by the number below each bar. The height of the bar represents the quality score of the review assessed using the AMSTAR tool. Moderate quality reflects a score of 5–8. Dark filled bars represent the primary analyses of each review; light filled bars represent sub-group analyses. Bars have been assigned as an increase or decrease (columns) in the measure where the change is statistically significant. Remaining bars appearing under ‘no change’ indicate direction of effect as an increase (>), no change (–) (the estimate of relative risk was 1 or regression coefficient or mean difference was 0), or decrease (<) in the measure or risk estimate.

Figure 8

Harvest plot of exposure to plastic-associated chemicals and circulatory outcome measures. The plastic-associated chemicals included are bisphenol A (BPA) (pink); phthalate monoester metabolites (blue), including monomethyl phthalate (MMP), monoethyl phthalate (MEP), mono-n-butyl phthalate (MnBP), monoisobutyl phthalate (MiBP), monobenzyl phthalate (MBzP), mono(2-ethylhexyl) phthalate (MEHP), mono(2-ethyl-5-hydroxyhexyl) phthalate (MEHHP), mono(2-ethyl-5-oxohexyl) phthalate (MEOHP), mono(3-carboxypropyl) phthalate (MCPP); mono(2-ethyl-5-carboxypentyl) phthalate (MECPP), and the molar sum of the di(2-ethylhexyl) phthalate metabolites (∑DEHP); and flame retardants (green) including polychlorinated biphenyl (PCB). Outcome measures are either dichotomous (†) or measured on a continuous scale (‡). Outcomes measured include serum lipids encompassing concentrations in low-density lipoprotein (LDL), high-density lipoprotein (HDL), total cholesterol (TC), triglycerides (TG) and apolipoprotein B (ApoB); child systolic blood pressure (SBP); child diastolic blood pressure (DBP); cardiovascular disease (CVD; for BPA and phthalates children also included with sampling frame [17]); CVD mortality; cerebrovascular disease mortality; hypertension and hypertension mortality. Each bar represents an individual effect estimate from the corresponding review, which is indicated by the number below each bar. The height of the bar represents the quality score of the review assessed using the AMSTAR tool. Low quality reflects a score of 1–4 and moderate (mod) quality a score of 5–8. Dark filled bars represent the primary analyses of each review; light filled bars represent sub-group analyses. Bars have been assigned as an increase or decrease (columns) in the measure where the change is statistically significant. Remaining bars appearing under ‘no change’ indicate direction of effect as an increase (>), no change (–) (the estimate of relative risk was 1 or regression coefficient or mean difference was 0), or decrease (<) in the measure or risk estimate.

Figure 9

Harvest plot of exposure to plastic-associated chemicals and respiratory outcomes. Plastic-associated chemicals included are phthalate monoester metabolites (blue), encompassing monoethyl phthalate (MEP), mono-n-butyl phthalate (MnBP), monoisobutyl phthalate (MiBP), monobenzyl phthalate (MBzP), mono(2-ethylhexyl) phthalate (MEHP), mono(2-ethyl-5-hydroxyhexyl) phthalate (MEHHP), mono(2-ethyl-5-oxohexyl) phthalate (MEOHP), mono(2-ethyl-5-carboxypentyl) phthalate (MECPP), molar sum of the di(2-ethylhexyl) phthalate metabolites (∑DEHP), mono(carboxyisooctyl) phthalate (MCOP), monocarboxyisononyl phthalate (MCNP), and mono (3-carboxypropyl) phthalate (MCPP); flame retardants (green) including polychlorinated biphenyl (PCB); and per- and polyfluoroalkyl substances (PFAS) (orange) including perfluorohexane sulfonate (PFHxS), perfluorooctanoic acid (PFOA), perfluorooctane sulfonate (PFOS) and perfluorononanoic acid (PFNA). Outcomes are dichotomous (†) and include risk of asthma, bronchitis, wheeze and allergic rhinitis. Each bar represents an individual effect estimate from the corresponding review, which is indicated by the number below each bar. The height of the bar represents the quality score of the review assessed using the AMSTAR tool. Low quality reflects a score of 1–4, moderate (mod) quality a score of 5–8 and high quality a score of 9–11. Dark filled bars represent the primary analyses of each review; light filled bars represent sub-group analyses. Bars have been assigned as an increase or decrease (columns) in the measure where the change is statistically significant. Remaining bars appearing under ‘no change’ indicate direction of effect as an increase (>), no change (–) (the relative estimate was 1), or decrease (<) in the estimate.

Figure 10

Harvest plot of prenatal exposure to plastic-associated chemicals and skin-related outcomes in children. Plastic-associated chemicals included are per- and polyfluoroalkyl substances (PFAS) (orange), encompassing perfluorohexane sulfonate (PFHxS), perfluorooctanoic acid (PFOA), perfluorooctane sulfonate (PFOS) and perfluorononanoic acid (PFNA). Outcomes are dichotomous(†) and include atopic dermatitis and eczema. Each bar represents an individual effect estimate from the corresponding review, which is indicated by the number below each bar. The height of the bar represents the quality score of the review assessed using the AMSTAR tool. Moderate (mod) quality reflects a score of 5–8. Dark filled bars represent the primary analyses of each review; light filled bars represent sub-group analyses. Bars have been assigned as an increase or decrease (columns) in the measure where the change is statistically significant. Remaining bars appearing under ‘no change’ indicate direction of effect as an increase (>) or decrease (<) in the measure or risk estimate.

Figure 11

Harvest plot of exposure to plastic-associated chemicals and cancer outcomes. Plastic-associated chemicals included are flame retardants (green), including polychlorinated biphenyl (PCB) further organised by group – gp I (44, 52, 101, 107, 187, 201), gp II (105, 118, 138, 156, 167, 170) and gp III (99, 153, 180, 183, 203) as well as PCB 28. Outcomes are dichotomous (†) and include breast cancer, non-Hodgkin’s lymphoma (NHL), NHL subtypes—chronic lymphocytic leukemia (CLL), diffuse large B-cell lymphoma (DLBCL), follicular lymphoma (FL)—and cancer-specific mortality: all cancer, breast cancer, leukaemia, liver cancer, lung cancer, melanoma, NHL, pancreatic cancer, rectal cancer, stomach cancer and uterine cancer. PCB poisoning refers to populations exposed to PCB-contaminated food and PCB occupational refers to populations occupationally exposed to PCBs. Each bar represents an individual effect estimate from the corresponding review, which is indicated by the number below each bar. The height of the bar represents the quality score of the review assessed using the AMSTAR tool. Low quality reflects a score of 1–4 and moderate (mod) quality a score of 5–8. Dark filled bars represent the primary analyses of each review; light filled bars represent sub-group analyses. Bars have been assigned as an increase or decrease (columns) in the measure where the change is statistically significant. Remaining bars appearing under ‘no change’ indicate direction of effect as an increase (>) or decrease (<) in the measure or risk estimate.

Figure 12

Harvest plot of exposure to plastic-associated chemicals and other outcomes. Plastic-associated chemicals included are flame retardants (green), including polychlorinated biphenyls (PCB) in populations exposed to contaminated food. All outcomes are dichotomous (†). Outcomes measured include mortality attributable to hepatic disease and all-cause mortality. Each bar represents an individual effect estimate from the corresponding review, which is indicated by the number below each bar. The height of the bar represents the quality score of the review assessed using the AMSTAR tool. Low quality reflects a score of 1–4 and moderate quality a score of 5–8. Dark filled bars represent the primary analyses of each review; light filled bars represent sub-group analyses. Bars have been assigned as an increase or decrease (columns) in the measure where the change is statistically significant. Remaining bars appearing under ‘no change’ indicate direction of effect as an increase (>), no change (–) (the relative estimate was 1), or decrease (<) in the measure or risk estimate.