Association of carotid intima-media thickness and dyslipidaemia in patients with type 2 diabetes: a protocol for systematic review and meta-analysis

Protocol and registration

This protocol for systematic literature review and meta-analysis is reported according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses for Protocols guideline 201522 (online supplemental appendix 1). The protocol is registered with the International Prospective Register of Systematic Reviews (PROSPERO), registration number CRD42023451731.

Ethics approval

The present study will use data pooled from studies published in peer-reviewed journals and will, therefore, not require ethical approval from an institutional ethics committee.

Patient and public involvement

None.

Objectives and outcomes questions

The objective of this is to evaluate the association of carotid intima thickness and dyslipidaemia in patients with diabetes. The study aims to answer the research questions based on Population, Exposure, Compares, Outcomes and Study design criteria:

What is the association between CIMT and risk of developing dyslipidaemia in patients with diabetes?

PICO eligibility criteria

• Population: Type 2 diabetic adults aged 18 years and above.

• Exposure: T2DM.

• Comparator: The review will include studies that have patients with non-diabetes to serve controls and aged 18 years and above.

• Outcomes: Dyslipidaemia and impaired CIMT.

• Study design: The proposed systematic review and meta-analysis will include cross-sectional studies, prospective case controls, prospective and retrospective cohorts.

Study selection

The studies for inclusion will be selected based on eligibility criteria and this will be performed by two (KM and RGGM) investigators independently to minimise the risk of selection bias. The initial screening process will include screening by title, abstracts, keywords and the overall aim of the study guided by PECO guidelines and the eligibility criteria. Studies deemed appropriate for the review and meta-analysis based on screening the title and abstract will be downloaded by the investigators independently for the full-text screening process. In case of disagreements regarding screening, data extraction and quality assessment, a final decision will be made by contacting independent investigators (WP and EM) to evaluate the study, domain in question.

Inclusion criteria

1. Studies investigating the association of CIMT and dyslipidaemia in type 2 patients with diabetes.

2. Used the ultrasound method to measure CIMT.

3. Measured lipids profile (total cholesterol, triglycerides, low high-density lipoprotein, low-density lipoprotein).

4. Included participants diagnosed with T2DM and non-diabetic controls.

5. Published in English. In cases where a study reported CIMT and lipids profiles in T2DM for different DM duration periods, the longest period will be considered.

Exclusion criteria

1. Studies without measurement of CIMT or lipid profiles.

2. Studies in patients with other conditions other than T2DM.

3. Studies not published in English.

4. Systematic reviews and meta-analysis or any reviews.

5. Studies in models of diabetes.

Search strategy

Google Scholar, PubMed, Science Direct and Web of Sciences databases will be searched up to 30 June 2023, following a standardised strategy developed following standardised medical subject headings (MeSH) terms and search strategy. Reference lists of the retrieved studies will also be searched to identify additional eligible studies. The following medical subject heading (MeSH) will be used to identify relevant studies from databases: “carotid intima-media thickness” ([MeSH] OR “CIMT” [MeSH] OR “carotid atherosclerosis” [MeSH]) AND “dyslipidemia” ([MeSH] OR “high total cholesterol”, “high triglycerides”, “low high-density lipoprotein cholesterol”, and “elevated low-density lipoprotein cholesterol”) AND “type 2 diabetes mellitus” ([MeSH] OR “hyperglycemia”). A piloted search strategy is attached in supplementary file (refer to online supplemental appendix B).

Data management

Statistical analysis

Meta-analysis will be performed in case where two or more studies are available reporting on same outcome measures, otherwise narrative approach will be taken. Data extracted from each study will be analysed using IBM SPSS statistical software (V.29) and metaHun, online meta-analysis software. Sample size, mean and corresponding SD will be used to estimate the effect size across all outcomes (CIMT, lipid profiles) in T2DM compared with controls. In cases where the selected reported data in the form of median and ranges, this will be converted to mean and SD following guidelines by Wan et al.19 The effect size will be reported as Cohen’s d/standardised mean difference and 95% CIs. Forest and funnels plots will be used to present the overall effect estimates. The magnitude of effect size will be categorised as small, medium and large when equivalent to 0.2, 0.5 and 0.8, respectively. Probability values of <5% will be regarded as statistically significant. The p˃0.05 for Egger’s test will be classified as potential publication bias.

Subgroup analysis

The heterogeneity in quantitative analysis is defined as variations in study outcomes between studies. We will use the I² statistic test to assess statistical heterogeneity.20 The I² values of <50% and >75% will be classified as minimal and substantial statistical heterogeneity, respectively. In the presence of heterogeneity, random-effect model meta-analysis will be used across the studies. However, in cases of no evidence of heterogeneity (I²=0%), a fixed-effect model meta-analysis will be conducted. Should there be evidence of heterogeneity (I² ˃50%), subgroup analysis will be performed based on predetermined covariates of interest such as age, gender, population size and country where the study was performed.

Sensitivity analysis

Sensitivity analysis will be explored by one study exclusion method to confirm the stability of the effect size.24 Any changes observed in the reanalysed effect size will be compared with the original effect size to identify the magnitude of the change.

Publication bias

Publication bias plays a role in research, in this proposed study publication bias will be assessed if there are more than ten studies analysing the same parameters. Therefore, in this study, we will assess publication graphically by visualising funnel plots and statistically with the Egger’s regression test.25 Any asymmetry noted through funnel plots will imply the presence of bias while a symmetrically shaped funnel plot will indicate the absence of publication bias. Additionally, Egger’s test p≤0.05 will confirm the presence of publication bias while p>0.05 will exclude publication bias among the studies analysed.

Grading of evidence

The grading of recommendations assessment, development and evaluation will be used to grade the quality of analysed evidence using GRADEPro.26 All domains including risk of bias, inconsistency, indirectness, imprecision and publication bias will be evaluated independently by RGGM and KM. Quality will be judged as low, very low, moderate of high based on the upgrading or downgrading of each domain.