Person‐Reported and Continuous Glucose Monitoring (CGM) Recorded Hypoglycaemia in People With Insulin‐Treated Type 2 Diabetes and Mild to Moderate Chronic Kidney Disease; Post Hoc Analysis From the Hypo‐METRICS Study

ABSTRACT

Background and Aims

People with diabetes and chronic kidney disease (CKD) may be at increased risk of iatrogenic hypoglycaemia. We compared rates of sensor‐detected (SDH) and person‐reported hypoglycaemia (PRH) in people with insulin‐treated Type 2 diabetes (T2D) with and without CKD in the Hypo‐METRICS study.

Materials and Method

Insulin‐treated T2D patients with > 1 hypoglycaemic episode within the last 3 months wore a blinded continuous glucose monitor (CGM) and recorded PRH on the bespoke Hypo‐METRICS app for 10 weeks. SDH was defined as per consensus guidelines. The participants were grouped based on estimated glomerular filtration rate (eGFR), that is CKD 1 (eGFR ≥ 90 mL/min; n = 117), CKD 2 (eGFR 60–89 mL/min; n = 140), and CKD 3 (eGFR 30–59 mL/min; n = 56). We used descriptive statistics, Kruskal‐Wallis rank sum tests, and a negative binomial models to control for confounders.

Results

In total, 313 participants with T2DM were included. Median age increased with declining eGFR (p < 0.001). Median (IQR) events per week for SDH < 3.9 mmol/L were 1.72 (0.66–3.43) in CKD 1, 2.12 (0.85–4.55) in CKD 2, and 2.22 (1.09–4.57) in CKD 3 (p = 0.12). For SDH < 3.0 mmol/L, these rates were 0.20 (0.00–0.50), 0.21 (0.10–0.61), and 0.20 (0.00–0.71), respectively (p = 0.88). PRH events had medians of 1.05 (0.40–2.12), 1.31 (0.61–2.42), and 1.28 (0.68–2.25) for CKD 1, CKD 2, and CKD 3 (p = 0.24).

Conclusion

In people with insulin‐treated T2DM and CKD stages 2 or 3, the rates of SDH and PRH do not significantly differ from those in people without CKD.

Trial Registration: ClinicalTRials.gov: NCT04304963

1. Introduction

In individuals with Type 2 diabetes (T2DM) and chronic kidney disease (CKD; eGFR < 90 mL/min/1.73 m²), elevated HbA_1c levels are strongly and independently associated with increased risks of mortality, hospitalization, myocardial infarction, stroke, and progression to end‐stage renal disease [1, 2]. However, intensification of glucose‐lowering therapy in the population with moderate to severe CKD has been linked to a higher incidence of both self‐treated and severe hypoglycaemia [3, 4, 5, 6]. Moderate to severe CKD itself predisposes to hypoglycaemia through impaired renal gluconeogenesis and reduced clearance of exogenous insulin and sulphonylureas [6, 7, 8]. In addition, advanced age, longer diabetes duration, and altered counterregulatory responses to hypoglycaemia are common in people with moderate to severe CKD and further increase the risk of hypoglycaemia [5, 6, 8]. The risk of hypoglycaemia in the CKD population with mild to moderate CKD is, however, less studied.

The Hypoglycaemia–Measurement, Thresholds, and Impacts (Hypo‐METRICS) study aimed to evaluate the clinical, psychological, and health economic consequences of both symptomatic and asymptomatic hypoglycaemia, and to establish an evidence‐based definition of sensor‐detected hypoglycaemia among people with insulin‐treated diabetes, more than half of whom with Type 2 diabetes. The aim of this post hoc analysis of the Hypo‐METRICS study was to explore the rates of sensor‐detected hypoglycaemia and person‐reported hypoglycaemia in people with insulin‐treated T2DM and mild to moderate CKD.

2. Method

The Hypo‐METRICS study, conducted as part of the European Union's Innovative Medicines Initiative program, Hypoglycaemia—Redefining Solutions for Better Lives (Hypo‐RESOLVE) [9, 10], was a 10‐week international, multicentre, observational, prospective study. The trial was approved by the Scientific Ethics Committee of the Capital Region of Denmark (H‐20044611), conformed to the Declaration of Helsinki, and written informed consent was obtained from all participants.

2.1. Participants

For this study, people with insulin‐treated diabetes (both Type 1 and insulin‐treated Type 2) on stable glucose lowering therapy comprising at least one injection of insulin per day, who had experienced at least one hypoglycaemic episode within the last 3 months, were recruited. Main exclusion criteria were the use of hybrid closed loop systems or an eGFR below 30 mL/min. In this post hoc analysis, we included data from all the participants with insulin‐treated T2DM and stratified them according to eGFR as per Kidney Disease: Improving Global Outcomes (KDIGO) guidelines [11].

2.2. Study Design

The study methods have previously been described in full [2]. In brief, the 10‐week study consisted of five visits: a baseline visit for the recording of demographics, blood sampling for measurement of serum creatinine, and obtaining other clinical data. Blood samples were analysed at the local laboratory at the site. At this visit, the participants were provided with the Abbott Freestyle Libre 2 sensor and blinded reader. A bespoke Hypo‐METRICS app for the recording of daily functioning and person‐reported hypoglycaemia was installed on their smartphone [12]. Hereafter, the participants had virtual visits every 2 weeks (Weeks 2, 4, and 6) with study personnel to solve any study‐related issues and to record changes in medical treatment and events of severe hypoglycaemia.

At the last visit after 10 weeks, we collected study devices and downloaded data from the blinded sensor.

During the entire study duration, the participants continued their usual manner of recording glucose values for self‐management, either with self‐monitoring of blood glucose monitoring (BGM) or CGM. Study personnel did not make changes to the treatment during the study.

For this analysis we used the CKD‐EPI 2009 for the eGFR equation.

2.3. Hypoglycaemia Definitions

Person‐reported hypoglycaemia (PRH) was defined as symptomatic episodes that resolved on carbohydrate ingestion or an incidental measure of glucose < 4.0 mmol/L and were recorded by the participants in the Hypo‐METRICS app every day during the 10 weeks, as described previously [10].

Sensor‐detected hypoglycaemia (SDH) was defined as hypoglycaemic events with a sensor‐glucose < 3.9 mmol/L or < 3.0 mmol/L lasting for at least 15 min as per consensus guidelines [4]. The sensor recorded data in 5‐min intervals that were interpolated to 1‐min intervals at download.

Matched hypoglycaemia was defined as an SDH where a PRH occurred within an hour of the SDH (to account for factors like delays in recording PRH, sensor lag, and recall errors regarding the timing of events). Data were excluded if the CGM use in a day was less than 70% and if the app had not been used at least once during the day.

2.4. CKD Groups

To maintain statistical power in the analyses, the participants were divided into three CKD groups:

• −CKD 1: eGFR≥90 mL/min/1.73 m².
• −CKD 2 (mildly decreased): eGFR 60–89 mL/min/1.73 m².
• −CKD 3 (mildly to moderately decreased): eGFR 30–59 mL/min/1.73 m².

The trial was registered on ClinicalTrials.gov (NCT04304963). All participants provided written informed consent.

2.5. Statistical Analysis

To test whether there was a statistically significant difference in hypoglycaemic outcome variables—SDH < 3.9, SDH < 3.0, PRH, and time below range (TBR)—and the CKD groups, we used a Kruskal Wallis rank sum test. In a multivariable regression, we included variables with a p‐value ≤ 0.1 in unadjusted univariate analysis. We created a fully adjusted negative binomial model for each for the outcomes SDH < 3.9, SDH < 3.0, and PRH. A post hoc sensitivity power calculation was performed to estimate the minimum detectable standardized difference between CKD groups, using β = 0.8 (80% power) and a two‐sided α = 0.05. All data analyses and graphs were produced using the statistical software R (version 4.5.2).

3. Results

The analysis included 313 participants with insulin‐treated T2DM, 117 grouped as CKD 1 with a mean (SD) eGFR of 102 mL/min (14.4), 140 participants with CKD 2 and a mean eGFR of 78.6 mL/min (8.3), and 56 participants with CKD 3 and a mean eGFR of 46.7 mL/min (9.9).

The participants with CKD were significantly older than the non‐CKD participants, ranging from a median age of 56 years (IQR 49–63) in CKD 1 to 70 years (IQR 62–74) in the CKD 3 group (Table 1). Diabetes duration did not significantly vary across groups. There were more women in the CKD 1 group compared to the CKD 3 group: 25% versus 43% respectively. HBA_1c did not differ between groups (Table 1). Hypoglycaemia awareness (Gold, Clarke and Hillerød‐method) [13] did not differ between CKD groups.

TABLE 1.

Baseline characteristics.

Variable	CKD 1 (eGFR ≥ 90 mL/min) N = 117	CKD 2 (eGFR 60–89 mL/min) N = 140	CKD 3 (eGFR 30–59 mL/min) N = 56
Age (years)*	56 (49–63)	65 (60–72)	70 (62–74)
Diabetes duration	21 (15–26)	23 (18–29)	26 (21–31)
Females (n (%))*	50 (43)	53 (38)	14 (25)
Hypoglycaemia awareness [13] (n(%))
Clarke
Aware	70 (61)	84 (61)	35 (63)
Unclassifiable	14 (12)	18 (13)	8 (14)
Reduced awareness	30 (26)	36 (26)	13 (23)
Gold
Aware	78 (67)	109 (78)	40 (71)
Impaired	39 (33)	31 (22)	16 (29)
Hillerød‐method
Aware	56 (48)	47 (34)	22 (39)
Impaired	39 (33)	64 (46)	21 (38)
Unaware	22 (19)	29 (21)	13 (23)
Country (n (%))
United Kingdom	52 (44)	48 (34)	20 (36)
Austria	14 (12)	30 (21)	9 (16)
Denmark	16 (14)	17 (12)	5 (9)
The Netherlands	34 (29)	42 (30)	22 (39)
France	1 (1)	3 (2)	0 (0)
Glucose monitoring (n (%))
BGM	73 (62)	79 (56)	31 (55)
CGM	44 (38)	61 (44)	25 (45)
Insulin delivery (n (%))
Multiple daily injections (MDI)	65 (56)	82 (59)	32 (57)
Mixed insulin	16 (14)	17 (12)	12 (21)
Insulin pump	2 (2)	5 (4)	2 (4)
Basal once daily	34 (29)	36 (26)	10 (18)
Total daily insulin dose. Median units per day (IQR).
	N = 76	N = 85	N = 36
	54 (26–92)	50 (28–96)	50 (24–84)
Non‐insulin glucose lowering therapy (n (%))
Metformin	72 (62)	91 (65)	29 (52)
Sodium‐Glucose Transport 2 inhibitor (SGLT2i)	29 (25)	36 (26)	9 (16)
Dipeptidyl peptidase 4 inhibitor (DPP‐4)	3 (3)	13 (9)	5 (9)
Sulfonylurea (SU)	11 (9)	21 (15)	7 (13)
Glucagon Like Peptide 1 receptor agonist (GLP‐1)	36 (31)	29 (21)	12 (21)
HbA1c (mmol/mol)	60 (51–67)	57 (51–68)	57 (51–66)

Note: Median and interquartile ranges unless indicated otherwise.

Abbreviations: BGM, blood glucose monitoring; CGM, continuous glucose monitoring; CKD, chronic kidney disease; eGFR, estimated glomerular filtration rate.

^* p < 0.05.

Based on the available group sizes, the study had 80% power at a two‐sided α level of 0.05 to detect pairwise CKD‐group differences in hypoglycaemia ranging from 0.35 to 0.46 standard deviations, with the lowest power in the comparison between CKD 1 and CKD 3.

3.1. Sensor‐Detected Hypoglycaemia (SDH)

The median (IQR) number of SDH (< 3.9 mmol/L) per week was similar across CKD groups, Table 2. Similarly, the number of SDH (< 3.0 mmol/L) per week was comparable across groups, Table 2. Furthermore, there was no statistically significant difference in SDH (< 3.9 or < 3.0 mmol/L) between the CKD groups in the multivariable regression model (Tables S1 and S2).

TABLE 2.

Number of events per week according to CKD group.

Events per week	CKD 1 (n = 117)	CKD 2 (n = 140)	CKD 3 (n = 56)	p
SDH (< 3.9)	1.72 (0.66–3.43)	2.12 (0.85–4.55)	2.22 (1.09–4.57)	0.11
SDH (< 3.0)	0.20 (0.00–0.50)	0.21 (0.10–0.61)	0.20 (0.00–0.71)	0.88
PRH	1.05 (0.40–2.12)	1.31 (0.61–2.42)	1.28 (0.68–2.25)	0.24
Matched (< 3.9)	0.30 (0.10–0.81)	0.40 (0.10–0.91)	0.51 (0.23–0.91)	0.17
Matched (< 3.0)	0.10 (0.00–0.20)	0.10 (0.00–0.20)	0.10 (0.00–0.20)	0.79

Note: p‐value is based on a Kruskal‐Wallis rank sum test.

Abbreviations: CKD, chronic kidney disease; Median and IQR, interquartile range; PRH, person‐reported hypoglycaemia; SDH, sensor‐detected hypoglycaemia.

3.2. Person‐Reported Hypoglycaemia (PRH)

The number of PRH events per week was similar across CKD groups, Table 2. Additionally, there was no statistically significant difference in PRH between the CKD groups in the multivariable regression model (Table S3).

3.3. Matched Events

The frequency of matched hypoglycaemic events did not differ significantly among CKD groups, Table 2.

3.4. Time Below Range (TBR)

Time below range was low across all groups, Table 3.

TABLE 3.

Time in ranges according to CKD group.

Time (%)	CKD 1 (n = 117)	CKD 2 (n = 140)	CKD 3 (n = 56)	p
TBR (< 3.9)	1.3 (0.5–2.6)	1.4 (0.6–3.5)	1.6 (0.8–4.1)	0.24
TBR (< 3.0)	0.1 (0–0.3)	0.1 (0–0.3)	0.1 (0–0.3)	0.97
TIR (3.9–10.0)	62.8 (45.8–81.3)	66.6 (53.3–79.8)	70.2 (58.4–79.5)	0.28
TAR (> 10.0)	33.4 (15.6–52.4)	29.8 (15.6–45.2)	25.4 (16.8–38.8)	0.24
TAR (> 13.9)	6.9 (1.6–18.0)	5.7 (1.1–12.0)	3.9 (1.5–7.6)	0.16

Abbreviations: CKD, chronic kidney disease; TAR, time‐above range; TBR, time‐below range; TIR, time in range.

4. Discussion

In this post hoc analysis of the Hypo‐Metrics study, we found no differences in the occurrence of sensor‐detected or person‐reported hypoglycaemia between people with T2DM with or without CKD stage 2–3 (mildly to moderately decreased renal function). This challenges the common perception that moderate CKD confers increased risk of hypoglycaemia in people with T2DM and may have implications for clinical decision‐making regarding glycaemic targets in people with insulin‐treated T2DM and CKD.

Previous studies reported on increased hypoglycaemia risks across various stages of CKD. In a study by Tian et al., conducted in a Singaporean population of 134 participants with T2D, people with CKD (pooled analysis of participants with eGFR < 60 mL/min) had higher odds of Level 1 (OR 1.73, p = 0.011) and Level 2 hypoglycaemic events (OR 2.16, p = 0.002), as well as longer duration of both Level 1 (OR 3.25, p < 0.001) and Level 2 events (OR 5.57, p < 0.001), compared with non‐CKD participants [14]. One possible explanation for the discrepancy with our data is the lower eGFR of the participants in the CKD group of that study. Notably, 27% of the cohort had CKD 4 (eGFR 15–29 mL/min/1.73 m²), whereas our study excluded individuals with an eGFR below 30 mL/min/1.73 m².

In a study by Hong et al., 80 people with T2DM and CKD (pooled analysis of participants with eGFR < 45 mL/min) were monitored with blinded CGM for 14 days. The mean number of hypoglycaemic events per person was 7.5 (< 3.9 mmol/L) and 7.3 (< 3.3 mmol/L), which is higher than in our cohort. The study also showed that the low eGFR was an independent predictor of an increased risk of hypoglycaemia [15]. Compared with our cohort, these participants had lower eGFR (all < 45 mL/min, severely decreased) and included people with an eGFR < 15 mL/min, no control group, and higher hypoglycaemia occurrence per participant.

Our post hoc analysis supports the existing evidence that mild to moderate CKD is associated with only a minimal additional risk of hypoglycaemia.

In the studies mentioned, the authors highlight the importance of sufficient glucose monitoring of glycaemic parameters in individuals with both CKD and diabetes, given the increased risk and complexity of managing glycaemia in this population. We observed a notable discrepancy between the number of SDHs (< 3.0 mmol/L) and those reported by participants as symptomatic (PRHs). This suggests that a substantial proportion of hypoglycaemic events may remain unrecognized by people with diabetes, reinforcing the need for CGM to complement self‐reported symptoms in clinical practice.

A key limitation to our study is the relatively small number of participants with CKD 3, with only 55 individuals having an eGFR < 60 mL/min. Furthermore, our study did not include any participants with an eGFR < 30 mL/min, as people with more advanced CKD were excluded. As albuminuria data were not collected in this study, we were unable to evaluate the potential association between the degree of albuminuria and hypoglycaemia outcomes.

A key strength of our study is the duration of CGM wear for 10 weeks. This prolonged monitoring period provided a comprehensive and robust dataset. Additionally, the study achieved a high completion rate for both CGM use and use of the Hypo‐Metrics app for hypoglycaemia registration. This high level of adherence enhances the reliability of the findings and reflects strong participant engagement, further supporting the validity of the data collected.

5. Conclusion

In people with insulin‐treated T2DM and CKD 2–3 (mild to moderate CKD) who were monitored with blinded CGM for 10 weeks, we found no association between presence or severity of CKD and the frequency of hypoglycaemia, whether measured as sensor‐detected events, person‐reported events, or matched events.

Funding

This work was supported by the Innovative Medicines Initiative 2 Joint Undertaking (JU) under grant agreement 777460. JU receives support from the European Union's Horizon 2020 Research and Innovation Programme and European Federation of Pharmaceutical Industries and Associations and T1D Exchange, Juvenile Diabetes Research Foundation, International Diabetes Federation, and The Leona M. and Harry B. Helmsley Charitable Trust. The industry partners supporting JU include Abbott Diabetes Care, Eli Lilly, Medtronic, Novo Nordisk, and Sanofi‐Aventis. Abbott Diabetes Care provided the continuous glucose monitors used in the study.

Conflicts of Interest

P.D. has received speaker fees for NovoNordisk, Menarini and Grunethal. P.D. has served on advisory boards for Dexcom. R.J.M.—speaker fees for Sanofi, advisory boards for Novonordisk and Lilly. Institution has received research support from NovoNordisk. Eli Lilly, Abbott Diabetes Care, Astra Zeneca, Boeringher Ingelheim, Roche. S.R.H. has provided consultancy to Eli Lilly, Zucara and Vertex for which his organization has received remuneration. M.L.E. has served on advisory boards and/or received lecture fees and/or research support from NovoNordisk, Eli Lilly, AstraZeneca, Medtronic‐Minimed, Dexcom, Ypsomed, Abbott Diabetes Care, Roche, NGM Pharma, Zucara, Pila Pharma, Vertex, VtV therapeutics. J.K.M. is a member in the advisory board of Boehringer Ingelheim, Eli Lilly, Medtronic, NovoNordisk AS, Prediktor A/S, Roche Diabetes Care, Sanofi‐Aventis and received speaker honoraria from Abbott Diabetes Care, Astra Zeneca, Dexcom, Eli Lilly, Medtronic, MSD, NovoNordisk AS, Roche Diabetes Care, Sanofi, Servier and Takeda. E.R. has served as consultant/advisor for Abbott, Air Liquide SI, Astra‐Zeneca, Boehringer‐Ingelheim, Dexcom, Eli‐Lilly, Hillo, Insulet, Medirio, Novo‐Nordisk, Roche, Sanofi‐Aventis, Tandem, and received research support from Dexcom and Tandem. P.C. has received personal fees and/or research support from Abbott Diabetes Care, Insulet, Dexcom, Novo Nordisk, Lilly, Vertex, CML, Medtronic, Roche Diabetes Care and Sanofi Diabetes. U.P.‐B. has served on advisory boards and has received lecture fees from Sanofi and Novo Nordisk. The other authors declare no conflicts of interest.

Supporting information

Table S1: Multivariable negative binomial regression for sensor detected hypoglycaemia (SDH) of < 3.9 mmol/L.

Table S2: Multivariable negative binomial regression for sensor detected hypoglycaemia (SDH) of < 3.0 mmol/L.

Table S3: Multivariable negative binomial regression for person‐reported hypoglycaemia (PRH).

Brøsen J. M. B., Malmsiø D., Divilly P., et al., “Person‐Reported and Continuous Glucose Monitoring (CGM) Recorded Hypoglycaemia in People With Insulin‐Treated Type 2 Diabetes and Mild to Moderate Chronic Kidney Disease; Post Hoc Analysis From the Hypo‐METRICS Study,” Diabetes, Obesity and Metabolism 28, no. 11 (2026): 10664–10669, 10.1111/dom.71247.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.