Improvements in both glycaemic and inflammatory profile in people with type 1 diabetes using automated insulin delivery systems

Abstract

Aims

Automated insulin delivery (AID) systems have been developed to achieve optimal glycaemic targets to prevent or slow the progression of type 1 diabetes (T1DM) and its complications. The aim of this study is to analyse the glycemic and inflammatory profile in people with T1DM after one year of using an AID system.

Materials and methods

A longitudinal study was performed, including 33 patients who started their treatment with an AID system (Medtronic 780G – 19%, Tandem Control-IQ – 27% and Roche-Diabeloop – 54%). A biochemical analysis was performed, and a blood sample was collected prior to pump implantation and at 3, 6, and 12 months.

Results

A significant increase in time in range was observed from the first month, with significant differences always relative to baseline (p < 0.001). The coefficient of variation was significantly reduced from implantation and this reduction was maintained up to one year (p = 0.002). Similarly, significant reductions in HbA1c (p < 0.001) and blood glucose levels (p = 0.001) were observed. The expression of IL6, IL1β, TNFα and VEGF was analysed in the peripheral mononuclear cells of the same cohort, observing a significant decrease in IL1β (p = 0.047), as well as a trend of decrease in the gene expression levels of the other molecules. In addition, correlations between these genes and certain biochemical parameters were observed.

Conclusions

After one year of AID system use, we observed a significant reduction in IL1β expression, independent of baseline glycaemic control. This reduction correlated with total cholesterol, HDL, and LDL levels. Moreover, individuals with poorer initial glycaemic control showed a greater decrease in IL6 levels. These findings suggest that AID use may contribute to modulating specific inflammatory markers in people with T1DM, with differential effects depending on initial metabolic status.

Resumen

Objetivos

Los sistemas de asa cerrada híbridos (AID) se han desarrollado con el objetivo de alcanzar un control glucémico óptimo, previniendo o retrasando la progresión de la diabetes mellitus tipo 1 (DM1) y sus complicaciones. El objetivo de este estudio fue analizar el perfil glucémico e inflamatorio en personas con DM1 tras un año de uso de un sistema AID.

Material y métodos

Se realizó un estudio longitudinal en 33 pacientes que iniciaron tratamiento con un sistema AID (Medtronic 780G-19%, Tandem Control-IQ-27% y Roche Diabeloop-54%). Se realizaron análisis bioquímicos y se tomaron muestras de sangre antes de la implantación del sistema y a los tres, seis y 12 meses.

Resultados

Se observó un aumento significativo del tiempo en rango desde el primer mes, con diferencias significativas respecto al valor basal (p < 0,001). El coeficiente de variación se redujo significativamente desde la implantación y esta reducción se mantuvo hasta el año (p = 0,002). Asimismo, se observaron descensos significativos en los niveles de HbA1c (p < 0,001) y de glucosa en sangre (p = 0,001). Se analizó la expresión génica de IL-6, IL-1β, TNF-α y VEGF en células mononucleares de sangre periférica (PBMCs) de la misma cohorte, observándose una disminución significativa de IL-1β (p = 0,047), así como una tendencia a la reducción en la expresión de las demás moléculas. Además, se identificaron correlaciones entre la expresión de estos genes y ciertos parámetros bioquímicos.

Conclusiones

Tras un año de uso de un sistema AID, se observó una reducción significativa en la expresión de IL-1β, independientemente del control glucémico basal. Esta disminución se correlacionó con los niveles de colesterol total, HDL y LDL. Además, los individuos con un peor control glucémico inicial mostraron una mayor reducción en los niveles de IL-6. Estos hallazgos sugieren que el uso de sistemas AID podría contribuir a la modulación de marcadores inflamatorios específicos en personas con DM1, con efectos diferenciales según el estado metabólico inicial.

Introduction

Since the discovery of insulin in 1921, type 1 diabetes mellitus (T1DM) treatment has been continuously evolving, not only in the pharmacological, but also in the technological field, helping to transform and improve the clinical management of the disease.^1,2 In this line, although HbA1c has traditionally been used to measure glycaemic control over time, this value is only an approximation of glucose control, and other factors such as time in range (TIR), glycaemic variability (GV) or the number of severe hypo- and hyper-glycaemic events have acquired great importance in the management of people with T1DM.^3,4 In recent years, automated insulin delivery (AID) systems have emerged to improve clinical management of T1DM. AID systems integrate three different components: a real-time continuous glucose monitor (CGM), a control algorithm, and an insulin pump. The integration of these three technologies has allowed better control of glycaemic status in T1DM, both in adults and children, increasing substantially the TIR of patients and reducing HbA1c levels, as well as reducing the number and time in hypo- and hyper-glycaemic episodes.⁵ However, despite the important role of glycaemic control on T1DM development and progression, diabetes is a multifactorial and very complex autoimmune disease, and other factors such as the inflammatory and vascular status of patients should be analysed for an early prediction of micro- and macrovascular complications.⁶

Among the biochemical and pathophysiological changes induced by fluctuations in glucose levels are oxidative stress and low-grade chronic inflammation.⁷

Regarding the inflammatory response, intermittent exposure to elevated glucose levels has been shown to promote the secretion of interleukin 6 (IL6), tumour necrosis factor alpha (TNFα), and various inflammatory cytokines. However, it is not only hyperglycaemic episodes that trigger these inflammatory processes; intermittent episodes of hypoglycaemia are also involved. It has been demonstrated that hypoglycaemia stimulates the circulation of certain leukocyte subgroups in the blood and induces pro-inflammatory alterations, both in healthy individuals and in patients with type 1 diabetes.⁸

The aim of this study was to analyse changes in the glycaemic, vascular, and inflammatory profile of T1DM subjects, after one year of the implementation of an AID system.

Material and methods

Study design and population

A one-year prospective longitudinal study was conducted. Thirty-three adult volunteers between 37 and 56 years with T1DM, enrolled in the Asturias Automatic Insulin Devices Registry⁹ and who started treatment with an AID system at the Endocrinology and Nutrition Service of the Central University Hospital of Asturias between November 2019 and August 2022 were included in the study. Written informed consent was obtained from all participants and the study was conducted in accordance with the principles of the Declaration of Helsinki for human research. The protocol was approved by the Ethical Committee of the Central University Hospital of Asturias (Project no. 2023.463, Oviedo, Asturias, Spain). Three different AID systems were included in the study: Minimed™ 780G system (MM780G; MiniMed Medtronic, Northridge, California), integrated with the Guardian Sensor 4 (Medtronic, Northridge, California), and the Accu-Chek® Insight Diabeloop™ system (Roche, Basel, Switzerland) and the Tandem t:slim X2 Control IQ™ system (Tandem Control-IQ; Tandem Inc., San Diego, California), both linked with the Dexcom G6 (Dexcom Inc., San Diego, CA) system. The choice of the AID system was made without randomization, depending on the device availability and physician's discretion. In no case did participation in this study benefit the waiting list for the implementation of the AID system. Previous use of a continuous glucose infusion system was not taken into account considered when selecting participants and was considered as another study variable. All patients are trained through a structured diabetes education program, including a session prior to the day of system implementation and training on the day of implementation. Among the initial participants, three discontinued the use of the AID system and four did not attend subsequent follow-up visits; therefore, they were excluded from the study, and a total of 26 patients completed the follow-up period. It is important to note that this was a pilot study designed to assess the potential impact of AID systems, which coincided with their initial implementation in the public healthcare system of Asturias, Spain. Due to the progressive roll-out and prioritization based on clinical criteria and existing waiting lists, recruitment was constrained, preventing randomization or equal distribution across device types. Consequently, the limited sample size was a result of real-world clinical conditions.

Study variables

All study participants passed a complete medical anamnesis, and underwent routine anthropometry, body composition (total and segmental; Tanita, 780MA, Tokyo, Japan; Table 1) and biochemical analysis at baseline and at three, six and twelve months after AID implantation (Table 1).

Table 1.

Clinical and demographic description and biochemical and glycemic variations.

Gender (males/females)	6/20
Age (years)	43.00 [37.50–56.75]
Diagnosis age (years)	19.50 [11.25–32.75]
T1D duration (years)	24.00 [18.00–34.50]
Previous pump (% yes)	69

	Basal	3 months	6 months	12 months
Waist (cm)	84.50 [77.25–92.5]	84.25 [72.25–90.75]	84.00 [73.00–89.50]	85.00 [70.00–92.00]
BMI (kg/m^2)	26.55 [23.90–29.23]	27.70 [23.90–29.65]	27.10 [23.60–29.70]	26.70 [23.00–30.80]
Fat mass (%)	33.45 [24.03–37.40]	34.10 [25.70–37.00]	34.10 [25.00–37.50]	29.20 [25.00–38.90]
Glucose (mg/dl)	159.5 [105.0–186.5]	123.0 [104.5–150.5]	113.5 [77.3–148.8]	113.0 [82.8–126.5]	***
HbA1c (%)	7.0 [6.5–7.5]	6.5 [6.2–6.9]	6.7 [6.3–7.0]	6.7 [6.3–7.2]	***
HbA1c (mmol/mol)	53 [48–58]	48 [44–52]	50 [45–53]	50 [45–55]	***
HbA1c (%)≤6.5	11/26 [42.3]	12/23 [52.2]	10/25 [40]	12/26 [46.2]
chTOT (mg/dl)	179.0 [165.3–190.8]	171.0 [161.0–184.0]	168.5 [153.8–191.3]	172.0 [154.5–191.3]
HDLch (mg/dl)	69.0 [60.3–81.0]	70.0 [63.0–78.5]	68.0 [61.3–78.8]	72.5 [63.5–81.3]
LDLch (mg/dl)	87.5 [66.5–98.3]	67.0 [56.0–78.5]	88.0 [72.0–101.0]	89.5 [76.3–98.5]	**
TG (mg/dl)	82.5 [61.0–92.0]	84.0 [76.0–98.0]	63.5 [55.8–70.0]	64.5 [54.4–69.8]	***

	Basal	2 weeks	1 month	3 months	6 months	12 months
% Info	97 [93–99]	96 [95–97]	97.9 [93.25–99]	97.5 [93.5–99]	97 [93–99]	98 [94.25–99]
Glucose (mg/dl)	152.0 [141–174]	146.0 [136–156]	143.0 [137.3–153]	146.5 [140.3–155.8]	152.5 [137.3–157]	150 [142.5–159.8]	*
GMI (%)	6.9 [6.7–7.6]	6.8 [6.6–7]	6.75 [6.6–7]	6.8 [6.7–7]	6.95 [6.6–7.1]	6.9 [6.7–.7.1]	*
CV (%)	36.9 [34.1–40.9]	31.0 [29.0–35.0]	31.0 [28.9–35.0]	30.7 [27.3–33.7]	31.5 [28.4–35.8]	29.8 [28.3–33.8]	***
TIR (%)	64.2 [54.0–75.0]	76.0 [69.0–84.0]	77.5 [72.0–81.0]	76.0 [68.8–80.8]	75.0 [68.3–81.0]	75.0 [68.5–77.8]	***

Data is shown as median [IQ25–IQ75] HbA1c ≤6.5% is shown as the percentage of patients with ≤6.5 levels.

BMI: body mass index; chTOT: total cholesterol; TG: triglycerides; CV: coefficient of variation; GMI: glucose management indicator; TIR: time in range.

Friedman differences among groups; ***p < 0.001; **p < 0.01; *p < 0.05.

All biochemical parameters (blood glucose, total cholesterol, LDL, HDL, triglycerides (TG) and HbA1c) from the four visits as well as data from the participant's CGM system at different time points (previous to AID implementation, two weeks and one, three, six and twelve months after implementation), were collected in the REDCap® Database. Additionally, data from the systems were downloaded from the available web-based software: TIR (percentage of time per day in which the patient has 70–180 mg/dl of glucose) which was analysed according to international consensus on TIR,³ and the coefficient of variation (CV) of glucose.¹⁰

Blood collection and sample preparation

Overnight fasting peripheral blood samples were collected from all subjects at every visit in EDTA-containing Vacutainer tubes (BD Biosciences, New Jersey, USA). Blood samples were immediately centrifuged at 2000 revolutions per minute (rpm) for 15 min at 4 °C. The top layer containing the plasma was divided into aliquots and stored at −80 °C until further analysis. Additionally, peripheral blood mononuclear cells (PBMC) were isolated from whole blood using the commercial reagent Lymphoprep™ (STEMCELL Technologies, Vancouver, Canada) according to the manufacturer's instructions.

RNA isolation and quantification

For RNA extraction from PBMC samples, cells were firstly crushed using RLT buffer by mechanical lysis with a needle (25G) and a syringe: RNA was then isolated using the commercial kit RNeasy® Mini Kit (Qiagen, Hilden, Germany) according to manufacturer's instructions. Isolated total RNA was quantified using Nanodrop 1000 (Applied Biosystems, ThermoFisher, Waltham, MA, USA).

A total amount of 0.8 μg of isolated RNA was reverse transcribed into cDNA using High-Capacity cDNA Reverse Transcription kit (Applied Biosystems, ThermoFisher, Waltham, MA, USA). Finally, gene expression analysis was conducted by RT-PCR using TaqMan® Gene Expression Assays (Applied Biosystems, ThermoFisher, Waltham, MA, USA; Supplemental Table S1) and the Applied Biosystems Prism 7900HT Sequence Detection System (Applied Biosystems) according to the manufacturer's instructions. Gene expression data are expressed as target gene RNA expression relative to the corresponding housekeeping mean gene expression (ΔCT = mean CT gene − mean CT value of the housekeeping RNA). The relative expression of each RNA was reported as 2^−ΔCT.

Statistical analyses

Descriptive analysis of continuous variables was performed by calculating their median and interquartile range, while categorical variables were expressed as percentages. Data were analysed using the Friedman test to assess differences, with Conover's post hoc comparison between time points. Correlations were performed using Spearman's test. A p-value less than 0.05 was considered statistically significant. Statistical analysis was done using the JASP version 0.18.1.0 statistical software all the images were developed using R environment RStudio 2024.12.1.

Results

Effect of hybrid closed-loop insulin delivery system implementation on glycaemic control and lipid control

A total of 26 patients enrolled in the study completed the 1-year evaluation, of which 14 used the Diabeloop® system, 7 used the Control-IQ® system and 5 used the MM780G® system. As shown in demographic Table 1, the age of participants ranged between 38 and 57 years old, with a median age at diagnosis of 20 years and a median diabetes duration of 24 [18.00–34.50] years. Sixty nine percent of the participants had previously used an insulin pump (in no case an AID system).

Implementation of the system led to the achievement of the international consensus targets (CV <36% and TIR >70%) and a reduction from 159.5 mg/dl at baseline to 113.0 mg/dl after 1 year of AID use (Table 2). Specifically, there is a decrease in CV from 36.9% at baseline to 29.8% at 12 months and TIR increased from 64.2% at baseline to 75.0% at 12 months. Regarding the other two significant parameters, time above range (TAR; time with glucose values more than 180 mg/dl) and time below range (TBR; time with glucose values less than 70 mg/dl), remarkable changes are observed (Table 1). This improvement is not only reflected in the glycaemic parameters but also translates into better lipid parameters for the patient, such as blood glucose, cholesterol, and glycated haemoglobin levels.

Table 2.

Mean of percentage of time spent in different glycemic ranges.

	TBR2	TBR1	TIR	TAR1	TAR2
Baseline	0.70	3.56	63.27	2.45	8.43
3 months	0.44	2.00	75.08	1.75	4.96
6 months	0.50	2.15	74.08	1.75	5.77
12 months	0.54	1.73	74.69	1.80	4.65
HbA1c ≤6.5
Baseline	0.38	3.88	74.38	18.13	3.25
12 months	0.65	1.65	71.06	20.12	5.71
HbA1c >6.5
Baseline	0.81	3.13	57.50	27.56	10.81
12 months	0.33	1.89	81.56	14.00	2.67

TBR2: percentage of time with glucose levels <54 mg/dl; TBR1: percentage of time with glucose levels 70–54 mg/dl; TIR: percentage of time with glucose levels 70–180 mg/dl; TAR1: percentage of time with glucose levels 180–250 mg/dl; TAR2: percentage of time with glucose levels >250 mg/dl.

The cohort was stratified into two groups based on HbA1c levels at the beginning of the study: first, those participants who started the study with HbA1c lower than or equal 6.5% and second those with HbA1c levels higher than 6.5%. As depicted in Table 2, the first group showed better results in TIR initially, and these levels remained stable throughout the year. On the contrary, the group with higher initial levels of HbA1c exhibits initially a worst control based on lower TIR, developing a significant improvement by the end of the year, even better than in the other group. Regarding TAR, the first group experienced two points increase, whereas patients with higher HbA1c at the beginning of the study achieved a 50% reduction.

Changes on the inflammatory profile of people with T1DM after one year with a hybrid closed-loop insulin delivery system

Cytokine variation levels between time points of the study are shown in Supplemental Table S2. The mRNA expression levels of four cytokines, interleukin 1β (IL1β), interleukin 6 (IL6), tumour necrosis factor (TNFα) and vascular endothelial growth factor (VEGF), were analysed by RT-qPCR. As shown in Fig. 1A, a statistically significant decrease in IL1β mRNA expression levels occurs (p = 0.047), while IL6 mRNA levels increase (Fig. 1B). Moreover, a reduction in VEGF mRNA expression levels is observed at the initial point (Fig. 1C); however, these levels remain constant throughout the rest of the time points studied. Finally, regarding TNFα, changes are observed after six months, resulting in a reduction in mRNA expression levels that persists until the endpoint, where there is a slight increase (Fig. 1D).

Fig. 1.

Changes in the inflammatory profile. (A–D) Box plots and density diagrams of inflammatory cytokines logmRNAs expression measured by RT-PCR. (A) IL1β gene expression. (B) IL6 gene expression. (C) VEGF gene expression. (D) TNFα gene expression. Significance: p value <0.05 (*). (E–G) Correlation between changes in the lipidic and the inflammatory profile after 6 months of the implementation of an AID system. (E) IL1β expression versus total cholesterol versus IL1β. (F) IL1β versus HDL cholesterol. (G) IL1β versus LDL cholesterol.

To further investigate the effect of AID implementation on the metabolic and inflammatory profile of people with T1DM, we analysed the correlation between changes in gene expression and changes in biochemical parameters. Our findings highlight an inverse correlation between lipid changes after 6 months of use of the AID system with changes in IL1β mRNA expression levels, specifically with total cholesterol (Fig. 1E) and LDLch (Fig. 1F), but not with HDLch (Fig. 1G).

Influence of initial glycaemic control on inflammatory profile of patients with T1DM after one year with a hybrid closed-loop insulin delivery system

To investigate whether initial glycaemic control could influence the inflammatory profile, we divided our cohort according to their initial HbA1c percentage, classifying patients according to whether they had a baseline HbA1c of less than or equal to 6.5% or greater than 6.5% and subsequently analysed the change in inflammatory profile of patients after one year from the implantation of the AID system (Fig. 2A–D). Only statistically significant changes in IL6 mRNA expression were observed between groups. Initially, patients with better glycaemic control showed increased IL6 mRNA expression, which was after the first 3 months, but from six months ahead, IL6 mRNA expression remained similar in both groups of patients (Fig. 2B).

Fig. 2.

Comparison of mRNA cytokine expression by groups of HbA1c baseline levels. (A) IL1β gene expression. (B) IL6 gene expression. (C) VEGF gene expression. (D) TNFα gene expression. Significance: p value <0.05 (*).

Discussion

The improvement in glycaemic profile is essential for a good management of diabetes and the prevention of associated comorbidities.^3,11 For this reason, the pharmaceutical industry has focused its efforts on developing a device that allows the patient to achieve the established glycaemic targets according to the international consensus on TIR and CV,¹² resulting in AID systems whose benefits in controlling glycaemic levels in people with T1DM have been previously reported.^13,14 In our cohort, we have observed that this improvement is achieved from the first weeks of use and maintained for at least one year.

Considering the limitations of BMI as a sole indicator of obesity, we also evaluated body composition to obtain a more precise assessment of fat distribution. Although no statistically significant differences were found in any of the body composition parameters studied, the observed reduction in body fat percentage may still hold clinical relevance, as upper body adiposity, particularly in the trunk and arm regions, has been associated with increased cardiovascular and metabolic risk in individuals with diebetes.¹⁵

When analysing glycemic control is the optimization of plasma HbA1c levels, which is associated with reduced risks of both macrovascular and microvascular complications in both patients with type 1 and type 2 diabetes.¹⁶ In our cohort, as previously reported in other populations,¹⁴ HbA1c was reduced from 7.0% to 6.5% within the first three months and remained below 7.0% throughout the intervention period. However, although important, HbA1c and glucose measurements are not enough, and for a more complete understanding of glycaemic control, TIR and CV should also be analysed. In this line, there was a significant increase in TIR of approximately 13% and while it is true that patients initially begin with a CV level very close to the ideal (36%), a reduction in CV was also observed in the whole cohort of participants, giving remarkable results after the implementation of an AID system.

We then divided our population into two groups, regarding on their initial HbA1c percentage (HbA1c ≤6.5% prior to AID system implementation, and HbA1c >6.5% prior to AID implementation) and observed that, although both groups improved their TIR after one year of AID usage, the group with higher initial levels of HbA1c had a greater increase in TIR, from a baseline of 57.50–81.56% after 12 months.

Regarding the narrow relationship between diabetes and vascular comorbidities, to ensure good management of T1DM, not only glycaemic control but also the lipid profile must be considered. In fact, exposure to high LDLch levels has been associated with an increased risk of retinopathy and nephropathy, independent of glucose control,¹⁷ while Hadjadj et al.¹⁸ report that serum TG play a role in the development and progression of renal and retinal microvascular disease and the development of nephropathy in T1DM patients.

In our study group, a significant reduction in LDLch levels was initially observed, but it returned to baseline levels after 12 months. In contrast, a reduction in TG was achieved from the sixth month and maintained throughout the year. Therefore, it could be beneficial to monitor these parameters to prevent the onset of comorbidities.

It has been postulated that changes in glucose levels, translated into higher GV, as well as changes in the lipid profile, are associated with worse diabetes prognosis, by promoting an activation of inflammatory response, which accelerates the progression of diabetic complications having a higher incidence of cardiovascular disease, such as cardiomyopathy, nephropathy, retinopathy.^19,20

It has been shown that people with diabetes have a higher population of M1-type macrophages, which are responsible for secreting large amounts of pro-inflammatory cytokines such as TNFα and IL1β, leading to insulin resistance, compared to healthy individuals, where the balance between M1 and M2 macrophages is maintained. The increase in the M1/M2 macrophage ratio leads to the production of a large number of pro-inflammatory cytokines, which exacerbate the progression of diabetic complications.²¹

Therefore, we decided to investigate how the implementation of an AID system, with its associated improvement in glycaemic and metabolic control, could also have an impact on the inflammatory profile of T1DM patients. Nevertheless, to the best of our knowledge, this is the first study to investigate the association between AID implementation and inflammatory profile in T1DM patients. Specifically, the differential expression of four different inflammation-related genes (IL1β, IL6, TNFα and VEGF) was evaluated in PBMC from T1DM volunteers.

After the implementation of the AID system, the expression of IL1β mRNA in the PBMC of the T1DM patients showed a significant reduction. The IL1β has potent proinflammatory activity, inducing the release of a wide variety of proinflammatory mediators. It is associated with common diseases such as type 2 diabetes and cardiovascular disease.²² In fact, we also observed that changes in IL1β were negatively associated with several lipid parameters, total and LDLch. The close relationship between both parameters has been extensively studied in the field of cardiovascular pathology and atherosclerosis,²³ but it has never been associated with better glycaemic control in people with T1DM, and the implementation of an AID system may have synergistic effects on the lipid and inflammatory profile of these patients. It should be noted that, when patients were stratified by glycemic control at the beginning of the study, increased expression of IL1β mRNA was observed in patients with better glucose control. These patients had similar reduced levels at the end of the study, although no statistical differences between groups were observed. This can be explained by the fact that modest reductions in body fat can lower IL1β expression and improve insulin sensitivity independently of major changes in HbA1c.²⁴

Regarding the expression of IL6, its behaviour was more ambiguous, with a slight increase in its expression after 3 months of use of the AID system and a slight decrease at the end of the study. These results are not surprising given the apparent pro- and anti-inflammatory duality associated with this interleukin, as well as its low expression levels under normal conditions.²⁵ Interestingly, changes in this interleukin are also associated with the patient's initial glycaemic control measured by its HbA1c percentage. However, the relationship between this cytokine and glycaemic control has mainly been studied in cohorts of non-diabetic²⁶ or people with type 2 diabetes,²⁷ with studies in T1DM patients being scarce and inconclusive.²⁸ Nevertheless, the relationship between the expression of IL6 and the long-term glycaemic control of patients seems to be of key importance, which is why it seems necessary to continue to investigate this aspect.

In addition to the aforementioned interleukins, there are two cytokines whose relationship with the development of vascular pathologies has been widely demonstrated: TNFα and VEGF.²⁹ While TNFα is a pleiotropic proinflammatory cytokine, responsible for a wide range of inflammatory responses and is also one of the most important biomarkers involved in the pathogenesis of T1DM,³⁰ VEGF is a potent angiogenic cytokine released by hypoxic cells, activated platelets, leukocytes, and cancer cells.³¹

Even though the role of TNFα in T1DM is still unclear, patients with long standing T1DM show higher levels of TNF-α, which may be a key factor involved in β-cell destruction; in fact, previous studies show an infiltration of TNFα producing cells in pancreatic islets.³² Besides this, although a slight fluctuation was observed during the year of AID treatment, no significant differences were observed in this cytokine due to the improvement in glycaemic control.

On the other hand, VEGF levels decreased throughout the treatment period, although without reaching statistical significance, possibly due to the small sample size of our study. VEGF is responsible for increasing vascular permeability, provoking protein kinase C activation, and releasing certain proteins that are elevated in plasma during diabetic endothelial dysfunction.²⁹ Schlingemann et al., suggest that in diabetes, poor glycaemic control alters platelets to release more VEGFα upon activation, as demonstrated by an independent correlation between VEGFα and HbA1c.³³

The study used a longitudinal design, allowing changes in the expression of inflammatory genes to be followed over time. The findings have direct clinical implications, offering potential biomarkers for monitoring and treating T1DM. The study has some limitations that should be acknowledged. First, as it is a pilot study, the sample size may not have been large enough to detect all possible variations in inflammatory gene expression. Additionally, not all potential confounding factors could be controlled for, such as the presence of other comorbidities or the type of AID sensor and system used. Lastly, the individuals included in this study were long-standing diabetic patients, most of whom had previously been treated with a pump and had reasonably good baseline HbA1c levels. Thus, a new cohort of patients in earlier stages of the disease and/or with worse baseline glycemic control should be included.

In this study, in addition to demonstrating the benefits in terms of glycaemic control associated with the introduction of an AID, independent of the patient's initial level of control, we have evaluated for the first time the effect that improving glycaemic control in people with T1DM has on the overall inflammatory state of the individual.

Our findings show a significant reduction in IL1β expression after one year of AID use, regardless of baseline glycaemic control. Interestingly, this reduction was associated with lipid profile parameters, including total cholesterol, HDL, and LDL levels. Moreover, individuals with poorer initial glycaemic control exhibited a greater decrease in IL6 levels, suggesting a more pronounced anti-inflammatory response in this subgroup.

These results highlight the potential of AID systems not only to improve metabolic outcomes but also to modulate systemic inflammation, which could have important implications for the long-term cardiovascular and metabolic risk in people with T1DM. Further studies are warranted to elucidate the mechanisms linking glycaemic regulation, lipid metabolism, and inflammatory pathways in the context of advanced diabetes technologies.

Our results, although still preliminary, open the door to novel studies that will allow us to demonstrate that good control of glycaemic parameters in patients with T1DM translates into changes in the inflammatory profile and, therefore, in a reduction in the incidence of diabetes complications.

CRediT authorship contribution statement

All the authors contributed to the study concept and design. Material preparation, data collection and analysis were performed by A.V.G., E.V.-F., P.P., J.A.-B., A.C.-I., M.G.-V. and C.L. The first draft of the manuscript was written by A.V.G., E.V.-F., and C.L. A.V.G., E.V.-F., J.A.-B., M.G.-V., T.G.-V., E.M.T., E.D., P.P., and C.L. commented on previous versions of the manuscript. All authors have read and agreed to the published version of the manuscript.

Ethical approval

The protocol was approved by the Ethical Committee of the Central University Hospital of Asturias (Project no. 2023.463, Oviedo, Asturias, Spain).

Prior presentation

Parts of this study have been submitted for presentation at Annual ESE Young Endocrinologists & Scientists (EYES) meeting 8–11 September 2023; 64 Congreso Sociedad Española de Endocrinología y Nutrición (SEEN) 18–20 October 2023 and XXXV Congreso de la Sociedad Española de Diabetes 10–13 April 2024.

Funding

This project is funded by the Government of the Principality of Asturias through the Agency for Science, Business Competitiveness and Innovation of the Principality of Asturias and co-financed by the European Union, through the Grants for Research Groups of Organizations of the Principality of Asturias for the year 2024, with file number IDE/2024/000705. Carmen Lambert is recipient of a Sara Borrel grant from the Instituto de Salud Carlos III (CD23/00037). Elsa Villa is recipient of an FPU grant from the Ministry of Education of Spain. TG-V was supported by a Río Hortega research contract (CM24/00080) from the Instituto de Salud Carlos III.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Acknowledgments

We would like to thank all the participants in the study for their collaboration, as well as the entire nursing team for their collaboration. We also thank Fundación Caja Rural and Sociedad Asturiana de Diabetes, Endocrinología, Nutrición y Obesidad (SADENO) for their continuous support.