This meta-analysis was conducted following the Meta-analysis of Observational Studies in Epidemiology (MOOSE) reporting guideline (12) and the Preferred Reporting Items for a Systematic Review and Meta-analysis (PRISMA) guideline ( Tables S1, S2 ) (13). The protocol for this meta-analysis was registered with PROSPERO (CRD42022341823).

We searched the PubMed, Web of Science, Cochrane Library, and Embase databases for relevant articles published up to August 16th, 2022. Our search strategy consisted of MeSH terms and entry terms with no restrictions on the language of the articles. We also scanned the reference lists of eligible articles for additional eligible articles that were not retrieved during the literature search. The details of the search strategy are presented in Table S3 .

The inclusion criteria were as follows (1): Patients: adult AITD patients (2); Control: adult in good health (3); Outcomes: platelet count (PLT), MPV, and platelet distribution width (PDW) (4); Study type: case control studies, cohort studies, and cross sectional studies (5); Studies included at least 20 patients to obtain good reliability.

The exclusion criteria were as follows (1): Animal experiments, review, conference abstracts, case reports and meta-analyses (2); Combined with severe primary systemic diseases such as cancer, acquired immune deficiency syndrome (3); Pregnant or lactating women, hematological diseases, and any other diseases could interfere with the platelet parameters (4); Subjects who had undergone thyroid surgery or iodine-131 therapy (5); Full-text or sufficient data could not be extracted.

Studies from the database were managed using EndNote X9 software to remove duplicate articles. The included articles based on inclusion and exclusion criteria were screened by T-SL and X-YY, who worked independently. The disagreement was discussed with another author (K-YZ) and subsequently resolved via consensus. Extracted research information includes (1): Background information: first author, publication year, country, study design, sample size, age, sex (2); Platelet parameters: PLT, MPV, and PDW. Values were presented as mean ± standard deviation (SD). When relevant data were missing, the corresponding author was contacted to obtain the information. If the same study contained multiple groups of available data, it would be divided into multiple independent individuals for data extraction. The process was performed independently by T-SL and X-YY. Any disagreements were resolved by discussion and consensus.

According to the recommendations of the Agency for Healthcare Research and Quality (AHRQ) for the quality evaluation criteria of observational studies, the Newcastle-Ottawa Scale (NOS) was used to evaluate cohort studies and case control studies (14–16), and the AHRQ methodology checklist was used to evaluate cross sectional studies (17). The NOS evaluates the quality of research from three aspects: the selection of the study groups; the comparability of the groups; and the ascertainment of either the exposure or outcome of interest for case control or cohort studies, respectively ( Methods 1, 2 ). The quality of the study was assessed as follows: low quality = 0-3 stars; moderate quality = 4-6 stars; and high quality = 7-9 stars (18). The AHRQ methodology checklist includes 11 items: the definition of information source; inclusion and exclusion criteria; time period and continuity for identifying patients; blinding of personnel; assessments for quality assurance; confounding and missing data; and response rates and completeness of patients ( Methods 3 ). If the answer is “unclear” or “no”, the item score is “0”. If the answer is “yes”, the score of the item is “1”. The quality of the study was assessed as follows: low quality = 0-3; moderate quality = 4-7; and high quality = 8-11 (17, 19). The quality of eligible articles was independently assessed by two investigators (T-SL and X-YY). Any disagreement was resolved by the third investigator (K-YZ).

The standardized mean difference (SMD) and 95% confidence interval (95% CI) of PLT, MPV, and PDW between AITD patients and healthy controls in each study were calculated and estimated. For studies using median, range and/or interquartile range, we used the methods proposed by Wan et al. (20) and Luo et al. (21) to estimate the sample mean and SD. Positive SMDs indicated higher platelet parameter values in AITD patients. If the changes in the outcomes were not reported. We used the conversion formulas recommended in the Cochrane Handbook Version 6.4 to calculate the changes in outcomes (22).

The I² statistic was used to quantitatively evaluate the heterogeneity of studies. Studies with I² of 0-25%, 25-75%, and > 75% were considered to have low, moderate, and high heterogeneity, respectively. A random effect model was applied regardless of heterogeneity. Meta-regression analyses were performed to determine the source of heterogeneity. According to the disease type, thyroid function, quality assessment, region, and study type, we conducted subgroup analyses. The groups were as follows: GD, HT, and AITD; Hyperthyroid/Before treatment, Hypothyroid/Before treatment, Euthyroid/After treatment, and Unknown; High quality (NOS = 7-9 stars or AHRQ score = 8-11), Moderate quality (NOS = 4–6 stars or AHRQ score = 4-7), and Low quality (NOS = 0-3 stars or AHRQ score = 0-3); Turkey, China, and Others; Case control study and Cross sectional study.

Besides, sensitivity analysis was performed to evaluate the robustness of the results by removing one single study in each turn. If 10 or more studies were included, we would use the funnel plots and Egger’s test to investigate publication bias. If there was publication bias, the trim and fill method was conducted to rectify the funnel plot asymmetry.

All statistical analyses were done with Stata Version 15.0 (StataCorp) from September 6 to 10, 2022. The results were considered significant when a 2-tailed P < 0.05.

Through database searches, 759 records were found. After screening, 19 articles (5, 9, 11, 23–38) were finally selected for the meta-analysis. The specific screening process is shown in Figure 1 , articles excluded after full text review are listed in Methods 4.

A total of 19 articles were included, 2 of which were from the same cross sectional study (26, 27). The characteristics of the 19 included articles are shown in Table S4 . Among them, part of the data from 4 articles (23, 30, 36, 37) were presented as median (min-max) or median (25–75th percentile), which were transformed according to the methods mentioned in Data Synthesis and Data Analysis. A comparison table of the original data and the corresponding converted data is shown in Table S5 . Based on the NOS and AHRQ methodology checklist, 17 articles (5, 11, 23–36, 38) were classified as high quality, and 2 articles (9, 37) as moderate quality. Detailed assessment scores are shown in Table S4 .

18 articles (26 available data) reported the difference in PLT between the AITD group and the control group. The results showed that PLT values were significantly increased in AITD patients when compared with healthy people (SMD: 0.164, 95% CI: 0.044 to 0.285, p = 0.008, Figure 2 ). For MPV, we pooled the effects of 9 articles (14 available data) and found that MPV values were significantly increased in AITD patients when compared with healthy people (SMD: 0.256, 95% CI: 0.013 to 0.500, p = 0.039, Figure 3 ). In the analysis of PDW (3 articles, 4 available data), no significant difference was found between the AITD group and the control group (SMD: 0.060, 95% CI, -0.164 to 0.284, p = 0.601, Figure 4 ).

Meta-regression analyses were performed according to covariates including disease type, thyroid function, quality assessment, study type, and region. The results showed that these covariates were not possible sources of heterogeneity ( Table S6 ). For PDW, since the number of included studies was less than 10, no meta-regression analysis was conducted.

Furthermore, we conducted subgroup analyses according to the disease type, thyroid function, quality assessment, region, and study type. As shown in Figure 2 , HT group had significantly higher PLT than healthy people (SMD: 0.277, 95% CI: 0.156 to 0.399), while the difference in GD group was not statistically significant (SMD: 0.007, 95% CI: -0.198 to 0.211). For thyroid function, PLT in thyroid function unknown group (SMD: 0.140, 95% CI: 0.039 to 0.242) and hypothyroid/before treatment group (SMD: 0.344, 95% CI: 0.097 to 0.592) was significantly higher than that of the control group, while there was no significant difference in hyperthyroid/before treatment group (SMD: -0.005, 95% CI: -0.292 to 0.282) and euthyroid/after treatment group (SMD: 0.314, 95% CI: -0.000 to 0.629). For studies of different quality scores, regions, and study types, the PLT in high quality group (SMD: 0.175, 95% CI: 0.047 to 0.303), China (SMD: 0.200, 95% CI: 0.134 to 0.267) and Turkey group (SMD: 0.253, 95% CI: 0.128 to 0.377), and cross sectional group (SMD: 0.254, 95% CI: 0.062 to 0.446) was significantly higher than that in the control group, while the difference in moderate quality group (SMD: 0.032, 95% CI: -0.212 to 0.276), other countries group (SMD: -0.176, 95% CI: -0.514 to 0.162), and case control group (SMD: 0.067, 95% CI: -0.058 to 0.192) was not significant ( Figures S1A–E ).

For MPV, we conducted subgroup analyses according to disease type, thyroid function, and study type. As shown in Figure 3 , MPV in GD group (SMD: 0.261, 95% CI: 0.121 to 0.402), hyperthyroid/before treatment group (SMD: 0.533, 95% CI: 0.117 to 0.950), euthyroid/after treatment group (SMD: 0.704, 95% CI: 0.237 to 1.171), and case control group (SMD: 0.543, 95% CI: 0.065 to 1.022) was significantly higher than that of the control group, while there was no significant difference in HT group (SMD: 0.290, 95% CI: -0.157 to 0.736), AITD group (SMD: -0.104, 95% CI: -0.444 to 0.236), thyroid function unknown group (SMD: -0.005, 95% CI: -0.257 to 0.248), hypothyroid/before treatment group (SMD: 0.182, 95% CI: -0.672 to 1.035) and cross sectional group (SMD: 0.048, 95% CI: -0.135 to 0.231, Figures S2A–E ). Due to insufficient data, we did not conduct subgroup analysis on MPV based on quality assessment and region. There were too few studies on PDW, and no subgroup analysis was conducted.

Sensitivity analysis showed that no single study significantly influenced the difference in PLT between AITD patients and healthy people. However, the stability of the pooled MPV was not good. After more than half of the studies were eliminated separately, the pooled results of the remaining studies were not statistically significant (95% CI exceeded 0, Table S7 ). In addition, funnel plots were roughly symmetrical ( Figure S3 ), and Egger’s test further showed that there was no publication bias among included studies in PLT and MPV ( Table S8 , Figure S4 ). For PDW, we did not conduct publication bias analysis because fewer than 10 studies were included.