Diurnal variations to proinflammatory markers in individuals with diabetes compared to healthy controls: protocol for a systematic review and meta-analysis

Abstract

Introduction

Diabetes mellitus is a highly prevalent metabolic disorder associated with chronic, low-grade inflammation. Of recent interest is the association between diabetes and circadian rhythm disruption. The aim of this review is to evaluate and synthesise clinical evidence for whether diabetes affects homeostatic diurnal patterns to proinflammatory markers in the human body. This could inform the optimal timing of immune-targeted therapies over the course of the day.

Methods and analysis

This systematic review will include primary clinical research studies reporting on diurnal variations, defined as an afternoon/evening (PM) minus a morning (AM) value, within a timeframe of 12±4 hours, for predefined proinflammatory markers, in individuals with diabetes (type 1 or type 2) compared with healthy controls. A search of online databases (Cochrane CENTRAL, Ovid MEDLINE and Ovid Embase) will be performed. Grey literature searches will be performed in clinical trial registries. Two review authors will independently screen retrieved citation records at the title/abstract and full-text levels. Study quality will be assessed using an appropriate National Institute of Health quality assessment tool. A meta-analysis will be performed if more than one study reports equivalent data for any outcome. Statistical heterogeneity will be assessed using the χ² test. Where a meta-analysis is not possible or unlikely to be meaningful, a narrative synthesis of the findings will be provided.

Ethics and dissemination

Ethics approval is not required for this systematic review as no original data will be collected. The results will be disseminated through peer-reviewed publication and conference presentations.

PROSPERO registration number

CRD420251115780.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• We will conduct a comprehensive systematic review and meta-analysis that aims to capture all relevant primary clinical research evidence related to how diurnal variations in proinflammatory markers differ between individuals with diabetes and healthy controls.

• The review will be conducted using rigorous methodology, in accordance with the Cochrane Handbook, and with results reported as stated in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement.

• As it is not feasible to investigate all potential proinflammatory markers, a finite number have been selected as outcome measures for this review.

• Only studies published in English will be included in the review.

• The review only considers studies involving individuals with type 1 and type 2 diabetes mellitus.

Background

Diabetes mellitus encompasses a group of metabolic disorders characterised by hyperglycaemia, due to altered production and/or response to insulin.¹ Diabetes has profound effects on millions of individuals worldwide. The International Diabetes Federation estimated that, as of 2021, approximately 537 million adults live with diabetes.² In individuals with diabetes, chronic hyperglycaemia is associated with systemic inflammation, dyslipidaemia and oxidative stress.³ This can affect multiple organs throughout the body, resulting in serious health complications such as nephropathy, cardiovascular disease, neuropathy and diabetic retinopathy.⁴

Various aetiological and pathophysiological processes underlie the development of diabetes.¹ Type 1 diabetes mellitus (T1DM) is characterised by autoimmune-mediated pancreatic islet cell damage and impaired insulin secretion, with genetic factors playing a key role in T1DM development.⁵ Type 2 diabetes mellitus (T2DM) can instead be attributed to a range of genetic and environmental factors, such as high caloric intakes, obesity, sedentary lifestyles and age, resulting in pancreatic beta cell dysfunction and insulin resistance.^{1 4 6} Another factor thought to contribute to the development of diabetes is environmental disruption, such as circadian rhythm misalignment.⁷ Circadian rhythms are internal biological clocks that operate on a 24-hour basis. They are controlled by the circadian system, consisting of the suprachiasmatic nuclei of the hypothalamus, and numerous secondary clocks in the brain and peripheral organs.⁸ The rhythmic variations of the circadian system play a critical role in the maintenance of metabolic and inflammatory homeostasis, orchestrating physiological processes relevant to diabetes, including insulin secretion, glucose uptake and lipid metabolism.^{3 8 9}

The temporal oscillations of the suprachiasmatic nuclei depend on several genes known as ‘clock genes’. These include CLOCK, BMAL1, PER1/2/3 and CRY1/2 genes, and encode transcription factors that regulate the rhythmic expression of genes associated with metabolism, immunity and cell-cycle regulation.^{3 10} Disruption to these tightly regulated rhythms, for example, as seen in night-shift workers, has been consistently associated with an increased risk of obesity, T2DM, chronic inflammation and cardiovascular disease.37 11,13 Evidence also exists for a relationship between circadian clock gene disruption and the development and progression of T2DM. Yu and colleagues found decreased mRNA levels of core clock genes, such as CLOCK, BMAL1, PER1 and CRY1/2, in the peripheral blood cells of individuals with T2DM.¹⁴ Participants with T2DM also showed elevated levels of proinflammatory markers, including interleukin-6 (IL-6), tumour necrosis factor-α (TNF-α) and C reactive protein (CRP), compared with healthy controls.¹⁴ Notably, reduced circadian clock gene expression was closely associated with elevated plasma levels of IL-6 and TNF-α.¹⁴

Inflammation is considered a key contributor to the pathophysiology of diabetes. Many proinflammatory markers have been associated with diabetes development.¹⁵ IL-6 is considered a major proinflammatory cytokine, while CRP is an acute-phase protein driven by increased IL-6 secretion.¹⁶ Elevated plasma levels of IL-6 and CRP have been found to predict the development of T2DM.¹⁶ Interleukin-1β (IL-1β) is another proinflammatory cytokine that has been associated with pancreatic beta cell damage, while TNF-α and interleukin-8 (IL-8) have been implicated in the pancreatic beta cell inflammation and/or progression of T1DM.^{15 17 18} Matrix metalloproteinases (MMPs) are enzymes involved in the degradation and remodelling of the extracellular matrix, as well as the regulation of proinflammatory cytokines.¹⁹ Vascular endothelial growth factor (VEGF), a pro-angiogenic and proinflammatory factor, also has an established role in the development of diabetic complications such as proliferative retinopathy. In addition, white blood cell (WBC) counts provide a non-specific indicator of inflammatory status. Elevated WBC counts are also considered a risk factor for diabetes.²⁰

As a chronic inflammatory condition associated with circadian rhythm disruption, there has been interest in understanding whether diabetes is associated with disruptions to the diurnal patterns of proinflammatory markers and immune cell features. To our knowledge, there is no published systematic review addressing the extent to which homeostatic diurnal variations in immune-related features, both in the blood and at tissue-specific locations, are dysregulated in individuals with diabetes. Elucidating how temporal fluctuations in immune function are altered in diabetes could have potential implications for diagnostic tests and targeted therapies.

Methods and analysis

This systematic review and meta-analysis will be conducted and reported following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.²¹ The Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA-P) checklist was used when writing this report.²¹ The planned start date of the systematic review is May 2026. It is expected to conclude in September 2026. The systematic review protocol has been registered in the International Prospective Register for Systematic Reviews (PROSPERO), registration number CRD420251115780. Protocol amendments will be uploaded through PROSPERO.

Eligibility criteria

This review will include all studies published from the date of database inception until the search date. Studies will be deemed eligible for inclusion in accordance with the following criteria:

Types of studies

We will include primary clinical research studies that sought to investigate diurnal variations. This includes, but is not limited to, interventional studies (randomised controlled trials (RCTs) and non-randomised trials), cohort studies, case–control studies and cross-sectional studies. Studies will not be limited by publication status, with studies on clinical trial registries eligible for inclusion. Studies involving the overlay of an intervention, such as RCTs, may be included; however, only baseline (pre-intervention) data will be extracted. Conference abstracts, narrative reviews, systematic reviews, case reports and editorials will be excluded. Preclinical studies and studies published in languages other than English will also be excluded.

Types of participants

The review will include studies that have evaluated individuals with diabetes (type 1 or type 2), with no restriction on participant age or gender. All studies where diabetes has been diagnosed using objective criteria, as defined by the study authors, such as glycated haemoglobin (HbA1c) ≥6.5% or fasting blood glucose ≥7.0 mmol/L, will be included. Studies involving the evaluation of individuals with gestational diabetes or pre-diabetes (including impaired glucose tolerance) will be excluded.

Types of comparators

The comparator participant group will be healthy controls. We will consider ‘healthy controls’ to pertain to those individuals without diabetes, as reported by the study authors. These individuals may have other systemic conditions, which will not exclude them from being considered in this systematic review. Where available, data relating to HbA1c values and fasting blood glucose levels will be extracted. Studies not reporting on HbA1c or fasting blood glucose levels for healthy control participants will still be included.

Types of outcome measures

We will exclude studies based on the outcome measures reported. We will only include studies reporting on the prespecified outcome measures, regardless of the reporting format.

Search methods for the identification of studies

Systematic searches will be conducted in Cochrane CENTRAL, Ovid MEDLINE and Ovid Embase, from database inception to the search date. The planned search date is May 2026. Search terms, including structured keywords and controlled vocabulary, are based on the key elements of the review question: diabetes, diurnal, circadian and inflammation. These terms will not be limited by language. Full search strategies are available in the online supplemental material. The search period will extend from database inception to the search date. Grey literature capture will involve searches of ClinicalTrials.gov and the WHO International Clinical Trials Registry. Reference lists of included studies will also be searched.

Data records and management

Study selection

Citation records from the database searches will be compiled in Covidence systematic review software.²² Duplicate study entries will be removed. Information from publications reporting on the same clinical study population will be merged, with the primary reference identified as the publication describing the largest sample size.

Two review authors will independently perform title and abstract screening, and resolve any discrepancies by consensus or by consulting a third review author (if required). Studies deemed relevant, or potentially relevant, for inclusion will undergo independent full-text screening by two review authors; conflicts will be discussed to reach consensus, consulting a third review author (if required). For studies excluded at full-text screening, the primary reason for exclusion will be recorded. A PRISMA flow diagram will be used to summarise the study selection process.

Information and data extraction

Data extraction will be undertaken independently by two review authors in Covidence, using a standardised data extraction form. Extracted data will be summarised in tabular format. Any discrepancies in data extraction will be discussed for consensus among the review team. We will attempt to contact the study authors of relevant studies by email if further information or clarification is required. If we fail to receive a response from the corresponding author within four weeks, or if the authors are unable to provide us with the requested information, we will use the information that is available. Extracted information and data from each eligible study will comprise:

1. Article details: year of publication, journal of publication and publication status.

2. Study details: study design, research question type, study setting, country study conducted, dates study conducted, study duration, corresponding author details (name, institution, email and address) and whether study authors were contacted for further information.

3. Population details: participant eligibility criteria, participant numbers per group, baseline characteristics of the study populations (including age, ethnicity, gender or sex (as reported), HbA1c values and fasting blood glucose levels), and for the diabetes group: type of diabetes (type 1 or type 2) and diabetes diagnostic criteria/definition.

4. Outcomes of the study:

   1. Methodological information (for each quantified parameter): unit of analysis (per person, per eye, average of both eyes, if applicable), unit of measure, method of analysis and time points measured.

   2. Quantitative data for diurnal variation: ‘change score’ (latter (afternoon/evening (PM)) timepoint value minus baseline (morning (AM)) timepoint value) ± SD, where reported, or the mean±SD at baseline and subsequent timepoints (change score to be derived), or other relevant measures of central tendency and variability. Other relevant measures of central tendency and variability will be converted to mean±SD for pooled analyses.

5. Other details: sources of funding statement (present or absent), actual source of funding, conflicts of interest statement (present or absent) and nature of any conflict of interest.

Quantitative data will not be extracted when the measured timepoints are unclear or outside the prespecified diurnal window of interest (ie, 12±4 hours between timepoints). Where multiple timepoint values fall within a time range of 12±4 hours, the timepoints closest to 12 hours will be used for data extraction.

Study quality assessment

Risk of bias assessments will be performed independently by two review authors for each of the included studies, using an appropriate risk of bias tool. The National Institute of Health (NIH) quality assessment tools will be used for all study types.²³ The ‘quality assessment of controlled intervention studies’ tool will be used for interventional studies, while non-interventional studies will be evaluated using the ‘quality assessment tool for observational cohort and cross-sectional studies’. Any discrepancies in assessments between the two review authors will be discussed to reach consensus, consulting a third review author if required.

Effect measures

Outcome measures

Outcome measures comprise the difference in diurnal variation (defined as the PM–AM value, measured 12±4 hours apart) in key proinflammatory markers (measured in any units) between healthy control participants (ie, without diabetes) and individuals with diabetes (where T1DM and T2DM populations are considered as separate comparisons).

The primary outcomes of this review are the difference in diurnal variation in IL-6 levels (quantified using any concentration unit), measured from blood, between (1) individuals with T1DM and healthy controls, and (2) individuals with T2DM and healthy controls.

Secondary outcomes are the difference in diurnal variation between (1) individuals with T1DM and healthy controls, and (2) individuals with T2DM and healthy controls, for each of:

• IL-6 levels, measured from biofluids other than blood (eg, tears, aqueous, cerebrospinal fluid, saliva and urine).

• IL-1β measured from any biofluid.

• IL-8 measured from any biofluid.

• TNF-α measured from any biofluid.

• MMP-9 measured from any biofluid.

• VEGF measured from any biofluid.

• CRP measured from any biofluid.

• Immune cell density, measured from blood or any bodily tissue (eg, cornea, retina).

We will consider the diurnal variation of proinflammatory markers in different biofluids or bodily tissues as separate outcome measures, in view of the anticipated heterogeneity.

Unit of analysis issues

In general, the unit of analysis will be per person. In studies involving the analysis of ocular fluids or tissues, the unit of analysis will be per eye. If the results of both eyes are evaluated and the average values of both eyes are reported, the unit of analysis will be per person. In studies reporting on both eyes separately, we will extract the correlation coefficient, where available, to adjust for within-person correlation. We will contact study investigators to obtain further information if necessary.

Dealing with missing data

For any studies where outcome data are missing, we will attempt to contact the study authors by email. If we fail to receive a response from the corresponding author within four weeks, or if the authors are unable to provide us with the requested information, we will use the information that is available.

Information and data synthesis

Data synthesis

A meta-analysis will be performed if more than one study reports data on the same prespecified review outcome measure. All predefined outcome measures are continuous. Where studies report the outcome in the same unit, a mean difference will be used. Where studies report data relevant to an outcome measure in different units, a standardised mean difference will be used, unless it is possible to convert all measures from different studies into the same units, in which case a mean difference is preferred. For each outcome measure where data are available in a study, the difference in diurnal variation (‘change score’) will be calculated between the diabetes (T1DM or T2DM) and the control group. The ‘change score’ will always be calculated as the PM value minus the AM value, generating a standard, non-absolute measure.

Inverse-variance weighted meta-analyses will be used to synthesise data across studies, including the reporting of 95% CIs. We plan to use the meta package in the statistical software ‘R’, V.4.5.2, to perform statistical computations for the meta-analysis.²⁴ When data from more than three studies are pooled, a random-effects model will be used. When data from three or fewer studies are pooled, a fixed-effects model will be used. As defined in the prespecified outcome measures, analyses relevant to the type of diabetes (T1DM and T2DM), relative to a comparator (control) group, will be considered separately due to considerations around clinical heterogeneity. In addition, the analysis of proinflammatory markers in biofluids and bodily tissues will be considered in separate analyses, in view of the anticipated heterogeneity.

In measuring heterogeneity, we will also consider the (1) magnitude and direction of the effects of individual studies and (2) strength of evidence for heterogeneity (using a p<0.10 from the χ² test as the criterion for significant heterogeneity).²⁵ In cases of significant heterogeneity, we will not pool the data.

For outcomes where a meta-analysis is not possible due to insufficient data or where it is unlikely to be meaningful due to significant heterogeneity, a narrative synthesis of the findings will be provided. This will include details about the population characteristics, the unit of analysis, the unit of measure and the ‘change score’ for relevant outcome measures.

Subgroup analysis and investigation of heterogeneity

We will consider different biofluids or bodily tissues in separate analyses in view of the anticipated heterogeneity. Subgroup analyses will therefore not be conducted as a part of this review.

Sensitivity analysis

If there are at least three studies included in the meta-analysis, sensitivity analyses that exclude studies deemed to be at high risk of bias (<10 out of 14 as ‘yes’ using the NIH quality assessment tools) will be undertaken to assess the robustness of findings.

Assessment of reporting biases

If at least 10 studies are included in a meta-analysis, we will assess publication bias using funnel plots and Egger’s test. We will explore asymmetries in the funnel plots through consideration of study characteristics, such as sample size.

Summary of findings table and certainty of the evidence

If appropriate, a summary of findings table will be presented where there are at least two studies per outcome. The quality and strength of the cumulative evidence will be assessed using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach.²⁶

Patient and public involvement

Patients and the public were not directly involved in the design of this systematic review and are not anticipated to be involved with its conduct. Opportunities to engage patient consumers in the reporting and dissemination of the findings will be pursued prior to publication of the findings.

Discussion

This protocol describes the planned methodologies and analyses to be undertaken for this systematic review and meta-analysis, with the aim of identifying, appraising and synthesising clinical evidence for how diurnal variations to proinflammatory markers differ between individuals with diabetes and healthy control individuals without diabetes. Strengths of the planned methodologies include that they will be performed with rigour and in accordance with the Cochrane Handbook. The systematic review protocol has been registered a priori to ensure transparency throughout the entirety of the review process. Furthermore, results will be reported in accordance with those defined in the PRISMA statement. This further serves to optimise the transparency, reproducibility, accuracy and completeness of the systematic review.

We would also like to address the selection of a finite number of proinflammatory markers as outcome measures. Although there are numerous inflammatory markers with potential implications in diabetes, our selection aims to ensure the feasibility of the evidence evaluation and analysis. An acknowledged limitation of the review relates to only including studies published in English; this pragmatic decision relates to the time and resources that would be required for the translation of potential non-English papers. The review also only considers studies involving individuals with type 1 and type 2 diabetes, not other forms of the metabolic disease.

The proposed systematic review is expected to advance understanding of how diabetes mellitus affects diurnal variations in proinflammatory markers and immune cell profiles in humans. This knowledge could be useful for guiding the timing of immune-targeted therapies, and thus be of benefit to those living with diabetes and the clinicians providing care to these individuals.

Ethics and dissemination

Ethics approval is not required for this systematic review as no original data will be collected. The results will be disseminated through peer-reviewed publication and conference presentations.