Innovative approaches to early detection of cardiovascular disease and diabetes risk: focus on glucose and cholesterol measuring

Abstract

Cardiovascular diseases and diabetes mellitus are major noncommunicable diseases contributing significantly to disease burden on individuals and societies. Both are largely preventable through early intervention and effective risk factor management. Prevention requires timely data about population-level risk factors and identification of high-risk individuals. The Joint Action JACARDI reviewed self-measuring methods for blood glucose and cholesterol. Results indicate that dried blood spot and capillary blood stored in microtubes/capillary tubes provide close agreement with venous samples for glycated hemoglobin, and dried blood spot samples for triglycerides, but sample collection and handling require standardization. Meanwhile, they are not suitable for diagnostic purposes.

Introduction

Cardiovascular diseases (CVDs) and diabetes mellitus (DM) are two major noncommunicable diseases (NCDs) [1]. Globally, CVDs are the number 1 cause of death, and approximately 18 million people died due to CVDs in 2019 (32% of all deaths) [2]. DM and kidney diseases due to DM caused over 2 million deaths globally in 2021, and high blood glucose is a significant cause of death [3]. Together, CVDs and DM place a significant burden on individuals, healthcare system, and societies.

Both CVDs and DM are largely preventable through lifestyle changes, and preventing diseases is deemed more cost-effective than treating already existing diseases. Timely data about population-level risk factors are needed for evidence-informed policy actions and targeted prevention programmes. At the individual level, early detection of high-risk individuals is important to ensure effective interventions. To complement resource-intensive and costly screening in clinical settings, there is a call for alternative methods that enable self-measuring of risk factors.

The market of the self-measuring tools is still in the early phase of development, and universal standards and protocols are lacking. In the framework of EU Joint Action JACARDI [4], we reviewed the current status of existing devices for self-measurement of cholesterol and glucose in non-clinical setting.

Methods

To review alternatives to often expensive and resource-demanding population-level screening interventions in clinical settings, we searched for new methods for self-measuring of blood glucose and cholesterol in non-clinical settings. A structured literature search for methods supporting self-measurement of cholesterol and blood glucose was conducted in PubMed in November 2024. To complement the limited search results from PubMed, we conducted manual search, using gray literature, and Internet searches. At the end, only six publications provided adequate information.

Results

New, emerging technologies using dried blood spot (DBS) or capillary samples, and noninvasive and minimal noninvasive methods are nowadays available as an alternative for traditional, expensive, and resource-intensive clinical screening initiatives. We identified six studies (Table 1) reviewing and/or validating these new methods.

Table 1.

Reviewed publications and their key findings about measurement of glucose and cholesterol with novel methods.

Source	Study design	Measured matrix/analyte	Key findings	Main conclusions
1. Affan et al. (2014) [5]	Systematic review and meta-analysis. Pooled data from 16 studies comparing two sampling methods.	Matrix: dried blood spots (DBS) vs. venous samples. Analytes: HbA1c, total cholesterol (TC), triglycerides (TG), HDL, LDL.	HbA1c showed close agreement (DBS = 0.9858venous + 0.3809), except for assays based on affinity chromatography. TG results were directly comparable (DBS = 0.9557 venous + 0.1427). TC required significant adjustment (DBS = 0.6807 venous + 1.151). The association was consistent at diagnostic HbA1c levels (>6.5%).	DBS assays are clearly associated with venous samples for HbA1c and selected lipids. Standardization is required for sample collection, transportation, storage, and analysis before DBS can be considered mainstream.
2. Colley et al. (2022) [7]	Systematic review and narrative synthesis (30 articles). Assessed accuracy, reliability, and patient acceptability of self-collected, posted samples.	Matrix: capillary blood in microtubes/capillary tubes (micro/cap) or as dried blood spots (capDBS) vs. venous whole blood (WB). Analyte: HbA1c.	Micro/cap showed high agreement; all reported limits of agreement (LoA) met IFCC quality targets (±5 mmol/mol). MD range for micro/cap was −0.4 to 1.4 mmol/mol. CapDBS showed comparatively poorer accuracy (LoA width 11.7–16.8 mmol/mol) and failed to meet IFCC targets. Increased analysis delay appeared to reduce correlation and agreement for DBS samples. Self-collection kits were highly acceptable (83%–96% satisfaction).	Microtubes/capillary tubes are promising for remote HbA1c monitoring due to accuracy and reliability. CapDBS samples have a limited role in developing healthcare economies due to poorer accuracy.
3. Nathan et al. (2021) [6]	Validation study comparing two same-day collections (capillary and venous) analyzed by the same HPLC method. Included 122 subjects (type 1 and type 2 diabetes).	Matrix: participant self-collected capillary blood (in capillary tubes/BioRad HCCS) vs. standard venous whole blood. Analyte: HbA1c.	Capillary and venous HbA1c results were highly correlated (R2 = 0.993). 96.7% of sample differences were within ±0.2% (2.2 mmol/mol). The overall regression showed near equality (capillary HbA1c = 0.0671 + 0.9942 × venous HbA1c). Feasibility: 97% of participants surveyed felt the method would be easy to do at home.	Participant-collected and shipped capillary blood samples collected via the tube kit are comparable and highly correlated with conventional venous results, supporting their use for remote clinical data collection.
4. Shang et al. (2021) [8]	Database review/systematic search of glucose monitoring products avoiding invasive blood sampling. Classified products as noninvasive optical (NIO-GM), noninvasive fluid sampling (NIFS-GM), or minimally invasive (MI-GM).	Matrix: interstitial fluid (ISF), blood, sweat, tears, saliva, or a combination. Analyte: glucose.	Identified 65 products; 13 had regulatory clearance. MI-GM products (inserted sensors measuring ISF) are generally more accurate, robust, and successful, with four having FDA non-adjunctive status. NIO-GM is painless but typically less accurate. NIFS-GM (using fluids like sweat or saliva) suffers from a significant time delay (lag time) between blood and fluid measurements.	MI-GM technologies are advancing rapidly and are key components of digital health tools. The major technical challenge remains the lag time between blood and measurement fluid concentrations, necessitating standardization of comparator matrixes (capillary vs. venous blood) for ISF measurements.
5. Kale et al. (2017) [9]	Prospective clinical study involving 50 subjects (Study Group with DM, HTN, or CHD risk factors, vs. Healthy Control Group).	Matrix: whole saliva (collected by spitting) vs. serum (venipuncture). Analytes: TC, HDL, TG, LDL, VLDL, and TC/HDL ratio.	Serum: TC and LDL were significantly higher in the study group (P < .05). Saliva: no statistically significant difference was found in saliva lipid profile values between the study group and control group (P > .05). Correlation (serum vs. saliva): good overall correlation was observed. Study group showed a strong correlation for TC (r = 0.60) and VLDL (r = 0.60), and a moderate correlation for HDL (r = 0.53) and LDL (r = 0.56).	An increase in serum lipid profile values leads to a corresponding increase in saliva lipid profile values. Saliva may be useful for identifying individuals with high serum cholesterol levels, but the insignificant difference between risk groups in saliva suggests that lipids exist in a different state of aggregation in saliva.
6. Singh et al. (2014) [10]	Prospective Correlative Study on 100 healthy individuals (medically compromised patients excluded).	Matrix: unstimulated whole saliva vs. serum (after overnight fasting). Analytes: TC, TGL, HDLC, LDLC, VLDLC.	Correlation coefficients for all five lipid parameters assessed were highly statistically significant (P < .01). Correlation was moderate for TC (r = 0.5384), TGL (r = 0.4822), HDLC (r = 0.5689), and VLDLC (r = 0.4388). Correlation was low/small for LDLC (r = 0.3182). Increasing serum mean values showed a corresponding increase in saliva mean values for all five parameters.	Saliva can be used as a noninvasive diagnostic tool for assessing lipid profile levels. The salivary lipid profile values reflect serum lipid values to some extent.

CHD: coronary heart disease DM: diabetes mellitus FDA: the United States Food and Drug Administration HbA1c: Hemoglobin A1c HDL: high-density lipoprotein HDLC: high-density lipoprotein-cholesterol HPLC: high-performance liquid chromatography HTN: hypertension IFCC: the International Federation of Clinical Chemistry and Laboratory Medicine LDL: low-density lipoprotein LDLC: low-density lipoprotein-cholesterol TGL: triglycerides  VLDLC: very low-density lipoprotein-cholesterol

For both cholesterol and glucose, DBS samples can provide cost-effective and user-friendly alternative to venous blood samples, especially when collection, transport, and processing of venous blood samples have logistical challenges, and overall resources are limited. DBS samples provide highly comparable results for glycated hemoglobin (HbA1c) and triglycerides in comparison to venous samples, but sample collection, handling, and analysis require standardized protocols [5].

Collecting capillary blood from a finger-prick represents a minimally invasive method for blood glucose measurement. For measurement of HbA1c, using capillary blood samples, shipped to the laboratory for analysis by an individual, compares well with the venous samples [6]. Similar results were obtained in a systematic review investigating self-collected capillary blood sample methods in comparison to venous samples. DBS samples, however, showed inferior accuracy to capillary blood samples [7].

Continuous glucose monitoring systems, developed for patients with DM, might be used for early detection of abnormal glucose values if the accuracy is improved. Of the existing numerous bloodless glucose measuring products, only 13 have received regulatory clearance and are currently on the market. These devices are limited for use by patients with DM without intended utilization for screening [8].

Cholesterol can be measured to some extent from saliva. This could turn out to be a feasible noninvasive method for home settings. Some devices and biosensors have been developed using saliva as a sample matrix. Saliva samples have shown fair comparability to venous serum blood samples in lipid profile, however, depending on the lipid fraction in question. The nature of the associations, standardized sample collection, transportation, storage, and analysis methods need to be further investigated and developed before they can be used for screening [9, 10].

Discussion

As CVDs and DM are preventable, timely data on population risk factor levels and early detection of high-risk individuals should be included in the national measuring and screening strategies of the countries facing a substantial burden of NCDs.

To ensure wide population coverage and cost-effectiveness of the measurement and screening procedures, new methods are justified. This is also well-founded by the fact that collecting, handling, processing, and analyzing blood samples in the traditional laboratory settings are time-consuming and expensive. However, self-measuring methods of risk factors for CVDs, such as blood glucose and cholesterol, cannot yet be considered as a mainstream method. Even though there are encouraging agreements with HbA1c and some cholesterol fractions with these minimal-invasive or noninvasive methods in comparison to venous samples, accuracy still lacks consistency. This is mostly because many open questions and restrictions still exist in comparison to the measurements conducted in standardized clinical settings. Accuracy of the methods is one of the key questions, leading to possible false-positive or false-negative results by these novel devices. There is also a risk that individuals conducting the tests are not following the instructions properly, which further leads to inaccuracy of the methods.

Furthermore, currently available methods are not designed for screening purposes, as they are rather intended for self-monitoring of existing or diagnosed diseases such as hypercholesterolemia or DM. Finally, a variety of manufacturers and brands with commercial interests exists but only a small fraction of them have received regulatory approval. When selecting a device, aspects such as clinical validation of the device should be considered.

Based on the current knowledge [7], novel methods and devices are not yet recommended for screening purposes, and they can only be regarded as complementary methods to the measurements conducted in clinical settings. However, further research on these novel methods is needed regarding screening programs, and if the innovative methods and devices develop accuracy and usability, they may prove to be a cost-effective alternative for population-based screening. With ever-increasing reductions in healthcare budgets of EU countries, the cost-effectiveness of the screening methods should be gaining emphasis and importance.

This review highlights the transformative potential of novel, noninvasive, and user-friendly methods for cholesterol and glucose measurements. These findings support two key strategic points for public health: prevention and focusing on the health of the entire population rather than individual patients.

These emerging technologies enable a significant shift in the way we organize population-level screening programs. They have the potential to reduce respondent burden by facilitating self-measurement, allowing individuals to conduct tests in their own homes. This may increase screening uptake by overcoming barriers such as time constraints, travel, and clinic-related anxiety. Home measurement may also minimize the societal costs by reducing the need for extensive clinical infrastructures and professional time.

These results on noninvasive, participant-accepted, and cost-effective measurement methods can support evidence-informed policy decisions both at the national and EU levels by supporting the formulation of screening strategies needed, for example, for the EU Cardiovascular Health Plan.

Funding

This project has received funding from the EU4Health Programme 2021-2027 under Grant Agreement 101126953. Views and opinions expressed are, however, those of the author(s) only and do not necessarily reflect those of the European Union or the European Health and Digital Executive Agency (HaDEA). Neither the European Union nor the granting authority can be held responsible for them.

Disclaimer

NotebookLM was used in the preparation of Table 1.

Key points.

• Self-measurement of cholesterol/glucose is a promising strategy for non-clinical settings; however, according to current knowledge, it is not yet recommended for screening.

• Current evidence supports the effectiveness of self-measurement in disease management; however, future research on non-clinically used devices for screening purposes is called upon.

• There is a need to find alternatives in the research field of screening to traditional clinical measurement, which is often costly and resource-demanding.

• Promoting the use of non-clinical devices and approaches can provide a tool to increase awareness and enhance the prevention of NCDs among the general population.

• Engaging policy-makers, clinical experts, funders, and other relevant stakeholders is crucial in providing support for identifying new, innovative, and cost-effective approaches to early detection of CVDs and DM risk.

Data availability

Manuscript does not use any data, only published documents listed in the references.