Relationship between key continuous glucose monitoring-derived metrics and specific cognitive domains in patients with type 2 diabetes mellitus

doi:10.21203/rs.3.rs-2357399/v1

Download PDF

Research Article

Relationship between key continuous glucose monitoring-derived metrics and specific cognitive domains in patients with type 2 diabetes mellitus

https://doi.org/10.21203/rs.3.rs-2357399/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 20 May, 2023

Read the published version in BMC Neurology →

You are reading this latest preprint version

Objective

Continuous glucose monitoring (CGM)-derived time in range (TIR) is closely associated with micro- and macro-vascular complications in type 2 diabetes mellitus (T2DM). This study was performed to investigate the relationship between key CGM metrics and specific cognitive domains in patients with T2DM.

Methods

A total of 96 outpatients with T2DM were recruited in this study. A battery of neuropsychological tests was performed to evaluate cognitive function, including memory, executive functioning, visuospatial ability, attention, and language. Participants wore a blinded flash glucose monitor (FGM) for 72 h. The key FGM metrics were calculated, including TIR, time below range (TBR), and time above range (TAR). Furthermore, the glycemia risk index (GRI) was also calculated by the GRI formula. Binary logistic regression was used to assess risk factors for TBR, and we further analyzed the associations between neuropsychological test results and TBR/TAR/TIR/GRI with multiple linear regressions.

Results

A total of 45.8% of patients with T2DM had hypoglycemia (TBR < 3.9) measured by FGM. The Spearman analysis results revealed that a higher TBR < 3.9 was correlated with worse performance on trail making test A (TMTA), clock drawing test (CDT) and cued recall scores (P < 0.05). The logistic regression analysis results revealed that TMTA (OR = 1.010, P = 0.036) and CDT (OR = 0.429, P = 0.016) scores were independent factors influencing the occurrence of TBR < 3.9. Multiple linear regressions revealed that TBR < 3.9 (β = -0.214, P = 0.033), TAR > 13.9 (β = -0.216, P = 0.030) and TAR 10.1–13.9 (β = 0.206, P = 0.042) were significantly correlated with cued recall scores after adjusting for confounding factors. TIR and GRI had no correlation with neuropsychological test results (P > 0.05).

Conclusion

A higher TBR < 3.9 and TAR > 13.9 were associated with worse cognitive functions (memory, visuospatial ability, and executive functioning). A higher TAR of 10.1–13.9 was associated with better memory performance. For patients with T2DM, glycemic targets can be relaxed to 10.1–13.9 mmol/L, which may slow the decline in cognitive function.

time below range

time in range

time above range

type 2 diabetes mellitus

cognitive impairment

Type 2 diabetes mellitus (T2DM) is the most prevalent chronic disease worldwide. People with T2DM are at risk of life-threatening acute and chronic complications, especially T2DM-caused central nervous system damage, which can cause functional and structural changes in the brain that lead to cognitive impairment and behavioral deficits [1]. Chronic hyperglycemia, recurrent hypoglycemic episodes and glycemic excursions have been implicated as potential causative factors of cognitive impairment [2–4], and for this reason, continuous glucose monitoring (CGM) has evolved as an essential part of diabetes management for diabetic patients with cognitive impairment. With the development of CGM technology, flash glucose monitoring (FGM) systems have increased in popularity in recent years because of the highly detailed information and enhanced accuracy that they provide. The key FGM-derived metrics enable quantitative evaluation of the quality of short-term glycemic control, including time in range (TIR), time below range (TBR) and time above range (TAR) [5]. However, as a single metric, TIR does not indicate whether the out-of-range readings are too low or too high; therefore, several researchers have proposed a new evaluation index of the blood glucose-glycemia risk index (GRI). The GRI is based on weighted combinations of TBR (the hypoglycemia component) and TAR (the hyperglycemia component) [6]. Key FGM-derived metrics and GRI provide more direct, comprehensive and complete information on blood glucose.

Human cognitive function consists of seven areas: learning, memory, visuospatial ability, executive functioning, language, complex attention, perceptual-motor function, and social cognition [7]. Cognitive impairment refers to the impairment of one or more of these cognitive functions. Previous studies have shown that TIR, TBR and severe hypoglycemia are associated with cognitive impairment [8]. However, no studies have expounded upon the relationships between TBR/TAR/TIR/GRI and different cognitive domains in patients with T2DM. In this study, we conducted a cross-sectional analysis of the relationships between TBR/TAR/TIR/GRI and specific cognitive domains in patients with T2DM to prevent the occurrence and development of cognitive impairment.

2.1. Participants

A total of 96 outpatients with T2DM were recruited from the First Hospital of Hebei Medical University. This study recruited outpatients who met the following inclusion criteria: (1) participants were between 40 and 75 years old; and (2) patients with T2DM met the 1999 WHO diagnostic criteria for diabetes mellitus and had been on a stable glucose-lowering regimen over the past 3 months. The exclusion criteria included (1) diabetic ketoacidosis, hyperglycemic-hyperosmolar state or severe hypoglycemia and recurrent episodes of hypoglycemia within the past 3 months, thyropathy, parathyropathy or other endocrinopathies; (2) major medical illness, such as severe anemia, serious heart disease, liver and kidney dysfunction or other systemic diseases; (3) acute infection; (4) malignant tumors and autoimmune diseases; (5) severe neurological or psychiatric diseases, alcoholism or abuse of psychotropic medicines; and (6) large infarcts, infection, tumor, or multifocal gray matter and/or white matter lesions observed on magnetic resonance imaging (MRI).

2.2. Clinical and Biochemical Measurements

Patient information on age, sex, diabetes duration, education, and medication prescriptions, such as glucose-lowering drugs, was collected by doctors. Each patient underwent a physical examination that included measurements of height, weight, and blood pressure. Blood samples were drawn by an experienced nurse after a 10-h overnight fast. Total cholesterol, triglycerides, high-density lipoprotein (HDL), and low-density lipoprotein (LDL) were determined by applying standard enzymatic methods using a biochemical analyzer (US Beckman AU5800 automatic chemistry analyzer, Beckman Coulter, America). Fasting plasma glucose (FPG) levels were assayed by using the glucose hexokinase method. HbA1c was measured by using high-performance liquid chromatography with a Hemoglobin A1c analyzer (Afinion 2 HbA1c, Abbott Diagnostics Technologies AS, Norway).

2.3. Neuropsychological assessment

Several neuropsychological tests, including the Mini-Mental State Examination (MMSE), Montreal Cognitive Assessment (MoCA), Digit Span Test forward (DST forward) and backward (DST backward), Trail Making Test A and B (TMTA and TMTB), Boston Naming Test (BNT), self-assessment of depression, Clock Drawing Test (CDT), Verbal Fluency Test (VFT) and Auditory Verbal Learning Test (AVLT), were performed to evaluate cognitive function, such as memory, executive functioning, visuospatial ability, attention and language. The AVLT included immediate memory total scores, scores on 20-min delayed recall, cued recall and long delayed recognition [9–11].

2.4. FGM parameters

Each patient wore a blinded FGM (FreeStyle Libre; Abbott Diabetes Care, Witney, UK) for 72 h. The glucose data were automatically recorded and saved every 15 min, and finger blood correction was exempted. The sensor of the FGM system was inserted on day 0 and removed after 72 h. After 72 h of monitoring, the data were downloaded to a computer and analyzed. Key FGM-derived metrics were calculated. These metrics included the following: (1) TIR: % of readings and time 3.9–10.0 mmol/L; (2) TBR: % of readings and time < 3.9 mmol/L (TBR < 3.9), further TBR were classified as level 1 TBR and level 2 TBR, Level 1 TBR: % of readings and time 3.0-3.8 mmol/L (TBR 3.0-3.8), Level 2 TBR: % of readings and time < 3.0 mmol/L (TBR < 3.0); (3) TAR: % of readings and time > 10.0 mmol/L; further TAR were classified as level 1 TAR and level 2 TAR, Level 1 TAR: % of readings and time 10.1–13.9 mmol/L (TAR 10.1–13.9), Level 2 TAR: % of readings and time > 13.9 mmol/L (TAR > 13.9); (4) Time ≥ 3.9 mmol/L: % of readings and time ≥ 3.9 mmol/L (T ≥ 3.9) [5]. GRI was calculated through the GRI formula.

2.5. Statistical analysis:

Statistical analyses were carried out with the IBM SPSS 25.0 software package (IBM, America). Continuous variables with a normal distribution are presented as the mean ± SD, and those not conforming to a normal distribution are expressed as the interquartile range [M (QL, QU)]. For continuous variables with normal or skewed distributions, Student’s t test, one-way ANOVA or the Mann‒Whitney U test were used for comparisons between groups. ᵪ²-test was applied for categorical variables. Spearman analysis was applied to analyze the relationships between glucose metrics and neuropsychological test results. Binary logistic regression was applied to analyze the independent associations between TBR and neuropsychological test results. Multilinear regression analysis was applied to analyze the correlations between neuropsychological test results and FPG, FGM metrics and GRI.

3.1. Characteristics of the study participants

The clinical characteristics of the 96 patients are shown in Table 1. The mean age of the 96 patients was 61.40 ± 7.84 years, the duration of diabetes was 13.05 ± 7.81 years, and HbA1c was 7.85% (7.20%, 8.70%). A total of 45.8% of patients with T2DM had hypoglycemia (TBR < 3.9) measured by FGM. According to TBR < 3.9, the participants were divided into groups based on absence or presence of hypoglycemia (< 3.9 mmol/L). There were no significant differences in sex, age, duration of disease, education, blood pressure, BMI or blood lipids between these two groups (P > 0.05). Compared with T2DM patients with T ≥ 3.9, T2DM patients with TBR < 3.9 showed increased TIR and glucose coefficient of variation (CV) values (P < 0.01) and lower HbA1c, FPG, TAR 10.1–13.9, TAR > 13.9, and mean glucose (MG) values (all P < 0.01). T2DM patients with TBR < 3.9 exhibited significantly worse performance on TMTA and CDT (all P < 0.05) (Table 2). TBR was further classified as level 1 TBR (TBR 3.0-3.8) and level 2 TBR (TBR < 3.0). Compared with T2DM patients with T ≥ 3.9, T2DM patients with TBR 3.0-3.8 exhibited significantly worse performance on the TMTA and CDT (all P < 0.05), as shown in Supplementary Table 1.

Table 1

Characteristics of study participants
Variable	Total	T ≥ 3.9	TBR < 3.9	P value
Sample	96	52	44	-
Male (n, %)	57 (59.38%)	27 (51.92%)	30 (68.18%)	0.106
Education junior college degree or above (n, %)	41 (42.71%)	20 (38.46%)	21 (47.73%)	0.087
Age (years)	61.40 ± 7.84	61.06 ± 8.57	61.80 ± 6.96	0.648
Diabetes duration (years)	13.05 ± 7.81	13.00 ± 7.69	13.11 ± 8.03	0.944
SBP (mmHg)	133.89 ± 13.88	133.88 ± 14.37	133.89 ± 13.44	1.000
DBP (mmHg)	80.68 ± 9.07	81.15 ± 8.61	80.11 ± 9.66	0.578
BMI (kg/m2)	26.11 ± 3.91	26.64 ± 4.15	25.48 ± 3.56	0.150
Total cholesterol (mmol/L)	4.84 ± 1.09	4.90 ± 1.12	4.76 ± 1.06	0.510
Triglycerides (mmol/L)	1.39 (0.98, 2.19)	1.68 (0.93, 2.46)	1.31 (1.01, 1.85)	0.184
HDL (mmol/L)	1.18 (1.03, 1.39)	1.19 (1.04, 1.40)	1.16 (1.01, 1.32)	0.476
LDL (mmol/L)	3.00 ± 0.79	3.03 ± 0.76	2.97 ± 0.83	0.753
HbA1c (%)	7.85 (7.20, 8.70)	8.15 (7.43, 9.08)	7.45 (6.70, 8.18)	< 0.001
FPG (mmol/L)	8.50 (7.13, 10.47)	8.83 (7.69, 11.58)	7.91 (6.21, 9.44)	0.005
TIR (%)	61.66 ± 27.76	52.34 ± 31.25	72.68 ± 17.73	< 0.001
TAR 10.1–13.9 (%)	20.50 (11.26, 33.27)	29.30 (16.75, 41.75)	16.08 (6.00, 21.05)	< 0.001
TAR > 13.9 (%)	2.91 (0.00, 18.00)	8.60 (0.09, 28.86)	1.05 (0.00, 6.81)	0.002
SD (mmol/L)	2.60 ± 0.84	2.59 ± 0.73	2.61 ± 0.96	0.883
GRI	34.32 (20.00, 64.60)	39.75 (18.18, 83.30)	29.38(20.98, 47.03)	0.158
MG (mmol/L)	9.27 ± 2.76	10.50 ± 2.88	7.81 ± 1.71	< 0.001
CV (%)	26.50 (23.65, 33.00)	24.77 (21.61, 27.02)	32.15 (26.85, 37.50)	< 0.001
MAGE (mmol/L)	5.66 (4.21, 7.00)	5.80 (4.36, 6.81)	5.24 (3.98, 7.66)	0.903
Use antidiabetes agents
Insulin (n, %)	47 (48.96%)	27 (51.92%)	20 (45.45%)	0.528
Data are presented as means (standard deviations) or medians (interquartile ranges) for continuous variables or numbers (percentages) for categorical variables. A two-tailed value of p < 0.05 was considered as statistically significant (bold).
SBP, systolic blood pressure; DBP, diastolic blood pressure; BMI, body mass index; HDL, high-density lipoprotein; LDL, low-density lipoprotein; HbA1c, hemoglobin A1c; FPG, fasting plasma glucose; TIR, time in range; TAR, time above range; TBR, time below range; SD, standard deviation; GRI, glycemia risk index; MG, mean glucose; CV, coefficient of variation; MAGE, mean amplitude of glycemic excursions; T ≥ 3.9, time ≥ 3.9.

Table 2

Neuropsychological test information in all patients
Variable	T ≥ 3.9	TBR < 3.9	P value
Sample	52	44
MMSE	28.00 (27.00, 29.00)	28.00 (26.00, 29.00)	0.418
MoCA	23.37 ± 2.69	22.70 ± 3.35	0.286
Immediate memory total scores	23.81 ± 5.07	22.34 ± 5.16	0.165
20-min delayed recall	8.33 ± 2.20	7.64 ± 3.09	0.206
Cued recall	10.50 (9.00, 11.75)	10.00 (8.00, 11.00)	0.118
Long delayed recognition	12.00 (11.00, 13.75)	12.00 (10.00, 14.00)	0.965
DST forward	7.00 (6.00, 8.00)	6.00 (6.00, 8.00)	0.114
DST backward	4.00 (4.00, 5.00)	4.00 (3.00, 5.00)	0.503
TMTA	38.43 (32.43, 60.25)	62.79 (38.16, 150.00)	0.010
TMTB	88.00 (57.93, 300.00)	300.00 (72.37, 300.00)	0.053
BNT	23.50 (22.00, 27.00)	24.00 (21.00, 26.00)	0.799
CDT	3.00 (2.00, 3.00)	2.00 (1.25, 3.00)	0.028
VFT	18.00 (15.00, 20.00)	17.00 (14.00, 21.00)	0.507
Depression scale	51.88 (47.50, 57.50)	52.50 (47.50, 56.25)	0.997
Negative emotion	86.50 (76.00, 95.00)	87.50 (76.75, 92.75)	0.953
Data are presented as means (standard deviations) or medians (interquartile ranges) for continuous variables. A two-tailed value of p < 0.05 was considered as statistically significant (bold).
MMSE, Mini-Mental State Examination; MoCA, Montreal Cognitive Assessment; DST, Digit Span Test; TMT, Trail Making Test; BNT, Boston Naming Test; CDT, Clock Drawing Test; VFT, Verbal Fluency Test; T ≥ 3.9, time ≥ 3.9; TBR, time below range.

3.2. Association between the key FGM-derived metrics and cognitive function

3.2.1. Associations between TBR < 3.9 and neuropsychological test results

The correlation analysis of TBR < 3.9 with neuropsychological test results was performed with Spearman’s analysis. TBR < 3.9 was positively correlated with TMTA scores (P < 0.01), while TBR < 3.9 was negatively correlated with cued recall and CDT scores (P < 0.05) (Fig. 1). Binary logistic regression analysis was carried out to further analyze the independent association between TBR and neuropsychological test results. The binary dependent variables were TBR < 3.9 (T2DM patients with TBR < 3.9 were coded as 1, and T2DM patients with T ≥ 3.9 were coded as 0), and the independent variables were age, duration of disease, education, BMI, and neuropsychological test results (scores on MMSE, MoCA, CDT, TMTA and cued recall). The results revealed that TMTA (OR = 1.010, P = 0.036) and CDT (OR = 0.429, P = 0.016) scores were independent influencing factors for the occurrence of TBR < 3.9 (Fig. 2). After adjusting for age, sex, diabetes duration, education and BMI, significant associations existed between CDT scores (OR = 0.379, P = 0.008) and TBR < 3.9. There was still a significant correlation after further respective adjustment of scores on MMSE, MoCA and TMTA (P < 0.05) (Table 3).

Table 3

Association between CDT scores and TBR < 3.9 after controlling for confounding factors
	TBR < 3.9
	OR	95% CI	P value
Model 1
CDT	0.379	(0.185–0.778)	0.008
Model 2
CDT	0.385	(0.188–0.790)	0.009
MMSE	0.894	(0.666–1.199)	0.454
Model 3
CDT	0.393	(0.189–0.815)	0.012
MoCA	0.954	(0.793–1.147)	0.616
Model 4
CDT	0.385	(0.186-0.800)	0.011
TMTA	1.010	(0.999–1.021)	0.071
MMSE, Mini-Mental State Examination; MoCA, Montreal Cognitive Assessment; TMTA, Trail Making Test A; CDT, Clock Drawing Test.
Model 1 was adjusted for age, sex, duration of disease, education and BMI
Model 2 includes all variables in model 1 plus MMSE
Model 3 includes all variables in model 1 plus MoCA
Model 4 includes all variables in model 1 plus TMTA

3.2.2. Associations between FPG and neuropsychological test results

FPG was negatively correlated with scores on DST forward, BNT, delayed recall and cued recall (P < 0.01 or P < 0.05) (Fig. 3).

3.2.3. Associations between cued recall scores and key FGM-derived metrics

Multiple linear regression was used to investigate the associations of TBR < 3.9, TIR, TAR 10.1–13.9, TAR > 13.9, FPG and cued recall scores, and the results revealed that TBR < 3.9 (β = -0.224, P = 0.021), TAR 10.1–13.9 (β = 0.263, P = 0.009) and FPG (β = -0.459, P = 0.000) were independent influencing factors for cued recall scores (Table 4). After adjusting for age, sex, diabetes duration, education and BMI, FPG (β = -0.307, P = 0.003), TBR < 3.9 (β = -0.214, P = 0.033), TAR > 13.9 (β = -0.216, P = 0.030) and TAR 10.1–13.9 (β = 0.206, P = 0.042) were significantly correlated with cued recall scores, and there was still a significant correlation after the adjustment of the depression scale (P < 0.05). After adjusting for FPG, the correlation of TAR > 13.9 (β = -0.031, P = 0.818) disappeared, but TBR < 3.9 (β = -0.327, P = 0.001) and TAR 10.1–13.9 (β = 0.343, P = 0.001) were even more significantly correlated with cued recall scores (Table 5).

Table 4

Results of multiple linear regression analysis of cued recall scores
Variable	Partial regression coefficient B	Standard error	β	t	P value	95% CI
Constant	6.551	1.838	-	3.564	0.001	2.899–10.202
BMI	0.145	0.058	0.224	2.518	0.014	0.031–0.260
Education	0.756	0.237	0.278	3.192	0.002	0.286–1.227
FPG	-0.399	0.083	-0.459	-4.790	0.000	-0.564- -0.234
TBR < 3.9 (%)	-0.101	0.043	-0.224	-2.343	0.021	-0.187- -0.015
TAR 10.1–13.9 (%)	0.044	0.016	0.263	2.669	0.009	0.011–0.076
BMI, body mass index; FPG, fasting plasma glucose; TAR, time above range; TBR, time below range; CI confidence interval
Multiple linear regression, dependent variable - cued recall scores; independent variable – age, duration of disease, education, BMI, TIR, FPG, TBR < 3.9, TAR 10.1–13.9, TAR > 13.9.

Table 5

Association between cued recall scores and FPG/TBR/TAR/TIR/GRI after controlling for confounding factors
	β	Adjustment R2	P value
Model 1
FPG	-0.307	0.149	0.003
Model 2
TBR < 3.9 (%)	-0.214	0.107	0.033
TAR > 13.9 (%)	-0.216	0.109	0.030
TAR 10.1–13.9 (%)	0.206	0.103	0.042
TIR (%)	0.086	0.068	0.399
GRI	-0.154	0.085	0.129
Model 3
TBR < 3.9 (%)	-0.327	0.240	0.001
TAR > 13.9 (%)	-0.031	0.140	0.818
TAR 10.1–13.9 (%)	0.343	0.247	0.001
Model 4
TBR < 3.9 (%)	-0.212	0.099	0.036
TAR > 13.9 (%)	-0.218	0.102	0.030
TAR 10.1–13.9 (%)	0.203	0.095	0.046
FPG, fasting plasma glucose; TIR, time in range; TAR, time above range; TBR, time below range;
GRI, glycemia risk index;
Model 1 was adjusted for age, sex, duration of disease, education and BMI
Model 2 was adjusted for age, sex, duration of disease, education and BMI
Model 3 includes all variables in model2 plus FPG
Model 4 includes all variables in model2 plus depression scale

3.2.4. Associations between TIR/GRI and neuropsychological test results

Multiple linear regression was used to investigate the associations between TIR/GRI and neuropsychological test results, and the results revealed that TIR and GRI had no correlation with neuropsychological test results (P > 0.05) (Supplementary Table 2).

All of the participants were stratified according to tertiles of TIR (TIR < 57%; 57% ≤ TIR < 79%; TIR ≥ 79%). At the same incidence of TBR < 3.9, compared with T2DM patients with 57% ≤ TIR < 79%, T2DM patients with TIR ≥ 79% exhibited significantly better performance on the BNT (P = 0.020), and there was no difference in other neuropsychological test results (P > 0.05) (Supplementary Table 3).

The present study demonstrated that a higher TBR < 3.9 was correlated with worse performance on TMTA, CDT and cued recall scores. TMTA and CDT scores were independent factors influencing the occurrence of TBR < 3.9. A higher TAR of 10.1–13.9 was correlated with better performance on cued recall scores. A higher TAR > 13.9 was correlated with lower performance on cued recall scores.

Previous studies have shown that severe hypoglycemia and recurrent episodes of hypoglycemia were related to cognitive impairment in T2DM, and severe hypoglycemia was closely related to Alzheimer’s disease [12, 13]. It has been demonstrated that TBR is negatively correlated with MoCA scores [8]. In contrast, no correlation was found in our study between TBR < 3.9 and MoCA scores, and we revealed correlations between TBR < 3.9 and specific cognitive domains in T2DM. Differences in study populations might account for the variations in findings. We recruited outpatients with T2DM who had been on a stable glucose-lowering regimen over 3 months, while patients who experienced severe hypoglycemia and recurrent episodes of hypoglycemia within the past 3 months were excluded. One study demonstrated that severe hypoglycemia was associated with worse cognition (global cognition, language, executive functioning and episodic memory) in older adults with type 1 diabetes mellitus [14]. In line with this study, our study revealed that a higher TBR < 3.9 was correlated with worse performance on TMTA, CDT and cued recall scores. Recent studies confirmed that compared with other neuropsychological tests, cued recall and CDT were thought to be more sensitive in identifying mild cognitive impairment and preferably valuable in predicting conversion from mild cognitive impairment to Alzheimer’s disease [10, 15]. Therefore, TBR < 3.9 was associated with cognitive impairment (memory dysfunction, deficits in visuospatial ability, and impaired executive functioning). In addition, FGM provided an improved opportunity to capture nocturnal asymptomatic hypoglycemia (NAH) events, and we found that 81.8% of patients with T2DM had NAH. NAH was associated with neurological damage, and this was consistent with previous studies [16]. The occurrence of NAH in patients might be related to disorders of rapid eye movement (REM) sleep phases and the suppression of counterregulatory hormone responses to hypoglycemia during REM sleep [17]. Furthermore, we compared the TMTA and CDT between T2DM patients with TBR 3.0-3.8 and T2DM patients with T ≥ 3.9. T2DM patients with TBR 3.0-3.8 exhibited significantly worse performance on the TMTA and CDT. This result indicates that nonsevere asymptomatic hypoglycemia was associated with cognitive impairment (deficits in visuospatial ability and impaired executive functioning). We did not observe any association between nonsevere asymptomatic hypoglycemia and cognitive performance in the global cognition, language and attention domains. Previous studies have shown a bidirectional association between severe hypoglycemia and cognitive impairment. Severe hypoglycemia can lead to cognitive function decline, and cognitive impairment may result in a greater risk of severe hypoglycemia [18]. Our findings were consistent with those of numerous studies on T2DM, and we observed that TBR < 3.9 was associated with disruption of memory, executive functioning and visuospatial ability. In addition, we found that deficits executive functioning and visuospatial ability could lead to the occurrence of TBR < 3.9, and the impairment of visuospatial ability was a more important factor for the occurrence of TBR < 3.9. The mechanism underlying the association between TBR < 3.9 and cognition is still unclear. Hypoglycemic episodes can cause neuronal death in the hippocampus and cerebral cortex and increase the consumption of alternate respiratory substrates such as ketone bodies, glycogen and monocarboxylate in the brain. These factors can cause mitochondrial dysfunction and brain function damage [19, 20].

To date, many studies have revealed that hyperglycemia is strongly correlated with cognitive impairment [21, 22]. TAR and FPG are the main metrics of hyperglycemia. We found that a higher TAR of 10.1–13.9 was associated with better memory performance. After adjusting for various confounding factors, TAR 10.1–13.9 was one of the main factors affecting memory. Therefore, we speculated that blood glucose levels within the range of 10.1–13.9 mmol/L might slow the decline in memory and impede the occurrence and development of cognitive impairment. Our study first revealed that a higher TAR of 10.1–13.9 was associated with better cognitive function. Therefore, it is recommended that the glycemic control target value should be relaxed for T2DM patients with cognitive impairment. One study demonstrated that a higher TAR > 10.0 mmol/L was associated with a lower performance in executive functioning and working memory [22]. Our results are in line with the study discussed above. We found that a higher TAR > 13.9 was correlated with worse memory performance. Previous studies indicated that elevated FPG could increase the risk of dementia [23]. Our findings revealed that higher FPG levels were associated with worse memory, attention and language ability. After adjusting for various confounding factors, FPG was one of the main factors affecting memory. In summary, higher TAR > 13.9 and FPG levels were associated with worse cognitive function. The mechanism underlying the association between very high-glucose hyperglycemia (TAR > 13.9) and cognition is still unknown. Acute high-glucose hyperglycemia may lead to cellular hypoxia and intracellular hyperosmosis in cerebral neurons, which results in neuronal damage. Chronic high-glucose hyperglycemia can cause the accumulation of advanced glycation end products, reactive oxygen species, and proinflammatory cytokines that induce neuronal damage [24, 25].

TIR is closely associated with micro- and macro-vascular complications and mortality in T2DM [26–29]. Some studies have shown that TIR is closely related to cognitive dysfunction [8, 22]. Contrary to a previous study, we did not observe a relationship between TIR and global cognition or specific cognitive domains. The different findings may be attributable to various reasons, including the age and ethnicity of participants and the duration of FGM. Each patient wore a blinded FGM in our study. The blinded FGM could reduce the influence of emotion, diet and exercise on blood glucose. Another reason for this inconsistent outcome might be the high incidence of TBR < 3.9 (45.8%). At the same incidence of TBR < 3.9, patients with higher TIR have better language ability. Therefore, when adjusting the glucose-lowering regimen for T2DM, the DATAA (Download Data, Assess Safety, Time in Range, Areas to Improve, Action Plan) model should be adopted in interpreting the FGM data in elderly patients and patients with cognitive impairment [30], and the first priority is to reduce TBR to target levels and then address TIR. Additionally, individualized glucose-lowering treatment regimens can improve patients’ language ability and prevent the occurrence and development of cognitive impairment.

GRI is a composite metric for evaluating glycemic control by utilizing the percentages of hyperglycemia and hypoglycemia based on CGM data. The lower the GRI is, the better the quality of glycemic control [6]. We did not observe an association between the GRI and specific cognitive domains, which may be related to the limitations of the GRI. It cannot independently reflect the composition of hyperglycemia and hypoglycemia groups. Moreover, GRI cannot predict the occurrence cognitive impairment in T2DM.

This study had several limitations. The cross-sectional design did not allow us to explore the relationship between the TBR/TAR/TIR/GRI level and the development of cognitive function. The present study also had a small sample size, and we need to expand the sample volume and conduct follow-up observation studies on patients in the future.

Through the study on TBR/TAR/TIR/GRI and specific cognitive domains, we found that a higher TBR < 3.9 was associated with worse cognition (memory dysfunction, deficits in visuospatial ability, and impaired executive functioning). At the same incidence of TBR < 3.9, higher TIR exhibited better performance on language ability. In addition, we found that higher TAR 10.1–13.9 was associated with better memory performance, and higher TAR > 13.9 was correlated with worse memory performance. Therefore, when setting glycemic targets for T2DM patients with cognitive impairment, it is important to be relaxed, with a strong focus on reducing TBR < 3.9 and preventing high-glucose hyperglycemia (TAR > 13.9).

Continuous glucose monitoring: CGM; Type 2 diabetes mellitus: T2DM; Flash glucose monitor: FGM; Time in range: TIR; Time below range: TBR; Time above range: TAR; The glycemia risk index: GRI; High-density lipoprotein: HDL; Low-density lipoprotein: LDL; Fasting plasma glucose: FPG; Mini-Mental State Examination: MMSE; Montreal Cognitive Assessment: MoCA; Digit Span Test forward: DST forward; Digit Span Test backward: DST backward; Trail Making Test A: TMTA; Trail Making Test B: TMTB; Boston Naming Test: BNT; Clock Drawing Test: CDT; Verbal Fluency Test : VFT; Auditory Verbal Learning Test: AVLT; Nocturnal asymptomatic hypoglycemia: NAH; Rapid eye movement: REM; Systolic blood pressure: SBP; Diastolic blood pressure: SBP; Body mass index: BMI; Hemoglobin A1c: HbA1c; Standard deviation: SD; Mean glucose: MG; Coefficient of variation, CV; Mean amplitude of glycemic excursions; MAGE; Time ≥ 3.9: T ≥ 3.9.

Ethics approval and consent to participate

The study protocol was approved by the Research Ethics Committee of the First Hospital of Hebei Medical University (approval number: 20210502) and have therefore been performed in accordance with the principles of the Declaration of Helsinki. Written informed consent was obtained from each participant. No potentially identifiable human images or data is presented in this study.

Consent for publication

Not applicable

Availability of data and materials

All data generated or analysed during this study are included in this published article [and its supplementary information files].

Competing interests

The authors declare that they have no conflicts of interest.

Funding

This work was supported by The Science and Technology project of the People's Livelihood in Hebei Province (20377707D), Special Funding for Local Science and Technology Development Guided by the Central Government (206Z7701G) and the Key Projects of Hebei Administration of Traditional Chinese Medicine (Z2022015).

Authors' contributions

Huimin Zhou and Shunjiang Xu contributed to conception and design of the study. Lina Wang, Rui Zhang and Lei Jiang acquired and analyzed the data. Shanshan Dong and Chenxu Zhao drafted and revised a significant portion of the manuscript or figures. Zhaoyu Gao conducted the Statistical analysis. Shanshan Dong wrote the paper. All authors have read and approved the final version of manuscript to be published.

Acknowledgements

We thank all the participants for their contribution to this study. We would like to thank AJE (https://china.aje.com/cn/researcher/) for English‑language editing.

ORCID

Shunjiang Xu: https://orcid.org/0000-0002-9441-3900

Shanshan Dong: https://orcid.org/0000-0002-0959-8999

Biessels GJ, Deary IJ, Ryan CM. Cognition and diabetes: a lifespan perspective. Lancet Neurol. 2008;7 2:184 – 90; doi: 10.1016/S1474-4422(08)70021-8.
Crane PK, Walker R, Hubbard RA, Li G, Nathan DM, Zheng H, et al. Glucose levels and risk of dementia. N Engl J Med. 2013;369 6:540–8. doi:10.1056/NEJMoa1215740.
Gao R, Ren L, Zhou Y, Wang L, Xie Y, Zhang M, et al. Recurrent non-severe hypoglycemia aggravates cognitive decline in diabetes and induces mitochondrial dysfunction in cultured astrocytes. Mol Cell Endocrinol. 2021;526:111192. doi:10.1016/j.mce.2021.111192.
Yang J, Zhao Z, Yuan H, Ma X, Li Y, Wang H, et al. The mechanisms of glycemic variability accelerate diabetic central neuropathy and diabetic peripheral neuropathy in diabetic rats. Biochem Biophys Res Commun. 2019;510 1:35–41. doi:10.1016/j.bbrc.2018.12.179.
Battelino T, Danne T, Bergenstal RM, Amiel SA, Beck R, Biester T, et al. Clinical Targets for Continuous Glucose Monitoring Data Interpretation: Recommendations From the International Consensus on Time in Range. Diabetes Care. 2019;42 8:1593–603. doi:10.2337/dci19-0028.
Klonoff DC, Wang J, Rodbard D, Kohn MA, Li C, Liepmann D, et al. A Glycemia Risk Index (GRI) of Hypoglycemia and Hyperglycemia for Continuous Glucose Monitoring Validated by Clinician Ratings. J Diabetes Sci Technol. 2022:19322968221085273; doi:10.1177/19322968221085273.
Sachdev PS, Blacker D, Blazer DG, Ganguli M, Jeste DV, Paulsen JS, et al. Classifying neurocognitive disorders: the DSM-5 approach. Nat Rev Neurol. 2014;10 11:634–42. doi:10.1038/nrneurol.2014.181.
Shang J, Lyu Y, An R, Yuan H. Relationship between mild cognitive impairment and time in the range of blood glucose targets and time below the target range in elderly patients with type 2 diabetes by a cognitive assessment. Chin J Endocrinol Metab. 2020;36 9:789–893.
Nasreddine ZS, Phillips NA, Bedirian V, Charbonneau S, Whitehead V, Collin I, et al. The Montreal Cognitive Assessment, MoCA: a brief screening tool for mild cognitive impairment. J Am Geriatr Soc. 2005;53 4:695–9. doi:10.1111/j.1532-5415.2005.53221.x.
Nesset M, Kersten H, Ulstein ID. Brief Tests such as the Clock Drawing Test or Cognistat Can Be Useful Predictors of Conversion from MCI to Dementia in the Clinical Assessment of Outpatients. Dement Geriatr Cogn Dis Extra. 2014;4(2):263–70. doi:10.1159/000363734.
Wind AW, Schellevis FG, Van Staveren G, Scholten RP, Jonker C, Van Eijk JT. Limitations of the Mini-Mental State Examination in diagnosing dementia in general practice. Int J Geriatr Psychiatry. 1997;12(1):101–8. doi:10.1002/(sici)1099-1166(199701)12:1<101::aid-gps469>3.0.co;2-r.
Han E, Han KD, Lee BW, Kang ES, Cha BS, Ko SH, et al. Severe Hypoglycemia Increases Dementia Risk and Related Mortality: A Nationwide, Population-based Cohort Study. J Clin Endocrinol Metab. 2022;107:5:e1976–86. doi:10.1210/clinem/dgab860.
McCrimmon RJ. Consequences of recurrent hypoglycaemia on brain function in diabetes. Diabetologia. 2021;64 5:971–7. doi:10.1007/s00125-020-05369-0.
Lacy ME, Gilsanz P, Eng C, Beeri MS, Karter AJ, Whitmer RA. Severe Hypoglycemia and Cognitive Function in Older Adults With Type 1 Diabetes: The Study of Longevity in Diabetes (SOLID). Diabetes Care. 2020;43 3:541–8. doi:10.2337/dc19-0906.
Lemos R, Simoes MR, Santiago B, Santana I. The free and cued selective reminding test: Validation for mild cognitive impairment and Alzheimer's disease. J Neuropsychol. 2015;9(2):242–57. doi:10.1111/jnp.12048.
Jauch-Chara K, Hallschmid M, Gais S, Schmid SM, Oltmanns KM, Colmorgen C, et al. Hypoglycemia during sleep impairs consolidation of declarative memory in type 1 diabetic and healthy humans. Diabetes Care. 2007;30:8. doi:10.2337/dc07-0067. 2040–5.
Cryer PE. Mechanisms of hypoglycemia-associated autonomic failure and its component syndromes in diabetes. Diabetes. 2005;54 12:3592–601. doi:10.2337/diabetes.54.12.3592.
Mattishent K, Loke YK. Bi-directional interaction between hypoglycaemia and cognitive impairment in elderly patients treated with glucose-lowering agents: a systematic review and meta-analysis. Diabetes Obes Metab. 2016;18(2):135–41. doi:10.1111/dom.12587.
Rehni AK, Dave KR. Impact of Hypoglycemia on Brain Metabolism During Diabetes. Mol Neurobiol. 2018;55 12:9075–88. doi:10.1007/s12035-018-1044-6.
Zhou Y, Huang L, Zheng W, An J, Zhan Z, Wang L, et al. Recurrent nonsevere hypoglycemia exacerbates imbalance of mitochondrial homeostasis leading to synapse injury and cognitive deficit in diabetes. Am J Physiol Endocrinol Metab. 2018;315(5):E973-E86. doi:10.1152/ajpendo.00133.2018.
Rawlings AM, Sharrett AR, Albert MS, Coresh J, Windham BG, Power MC, et al. The Association of Late-Life Diabetes Status and Hyperglycemia With Incident Mild Cognitive Impairment and Dementia: The ARIC Study. Diabetes Care. 2019;42 7:1248–54. doi:10.2337/dc19-0120.
Sugimoto T, Tokuda H, Miura H, Kawashima S, Ando T, Kuroda Y, et al. Cross-sectional association of metrics derived from continuous glucose monitoring with cognitive performance in older adults with type 2 diabetes. Diabetes Obes Metab. 2022. doi:10.1111/dom.14866.
Wang F, Luo J, Ding D, Zhao Q, Guo Q, Liang X, et al. Elevated Fasting Blood Glucose Level Increases the Risk of Cognitive Decline Among Older Adults with Diabetes Mellitus: The Shanghai Aging Study. J Alzheimers Dis. 2019;67 4:1255–65. doi:10.3233/JAD-180662.
Malone JI, Hanna S, Saporta S, Mervis RF, Park CR, Chong L, et al. Hyperglycemia not hypoglycemia alters neuronal dendrites and impairs spatial memory. Pediatr Diabetes. 2008;9 6:531–9. doi:10.1111/j.1399-5448.2008.00431.x.
Valente T, Gella A, Fernandez-Busquets X, Unzeta M, Durany N. Immunohistochemical analysis of human brain suggests pathological synergism of Alzheimer's disease and diabetes mellitus. Neurobiol Dis. 2010;37 1:67–76. doi:10.1016/j.nbd.2009.09.008.
Lu J, Ma X, Zhou J, Zhang L, Mo Y, Ying L, et al. Association of Time in Range, as Assessed by Continuous Glucose Monitoring, With Diabetic Retinopathy in Type 2 Diabetes. Diabetes Care. 2018;41 11:2370–6. doi:10.2337/dc18-1131.
Li F, Zhang Y, Li H, Lu J, Jiang L, Vigersky RA, et al. TIR generated by continuous glucose monitoring is associated with peripheral nerve function in type 2 diabetes. Diabetes Res Clin Pract. 2020;166:108289. doi:10.1016/j.diabres.2020.108289.
Li J, Li Y, Ma W, Liu Y, Yin X, Xie C, et al. Association of Time in Range levels with Lower Extremity Arterial Disease in patients with type 2 diabetes. Diabetes Metab Syndr. 2020;14 6:2081–5. doi:10.1016/j.dsx.2020.09.028.
Lu J, Wang C, Shen Y, Chen L, Zhang L, Cai J, et al. Time in Range in Relation to All-Cause and Cardiovascular Mortality in Patients With Type 2 Diabetes: A Prospective Cohort Study. Diabetes Care. 2021;44(2):549–55. doi:10.2337/dc20-1862.
Isaacs D, Cox C, Schwab K, Oser TK, Rinker J, Mason MJ, et al. Technology Integration: The Role of the Diabetes Care and Education Specialist in Practice. Diabetes Educ. 2020;46(4):323–34. doi:10.1177/0145721720935123.

No competing interests reported.

Download PDF

Journal Publication

published 20 May, 2023

Read the published version in BMC Neurology →

Editorial decision: Major revision
27 Mar, 2023
Reviews received at journal
23 Mar, 2023
Reviewers agreed at journal
15 Mar, 2023
Reviewers invited by journal
15 Mar, 2023
Editor assigned by journal
15 Mar, 2023
Editor invited by journal
23 Dec, 2022
Submission checks completed at journal
23 Dec, 2022
First submitted to journal
08 Dec, 2022

You are reading this latest preprint version

Relationship between key continuous glucose monitoring-derived metrics and specific cognitive domains in patients with type 2 diabetes mellitus

Status:

Journal Publication

Version 1

Abstract

Objective

Methods

Results

Conclusion

Figures

1. Introduction

2. Research Design And Methods

2.1. Participants

2.2. Clinical and Biochemical Measurements

2.3. Neuropsychological assessment

2.4. FGM parameters

2.5. Statistical analysis:

3. Results

3.1. Characteristics of the study participants

3.2. Association between the key FGM-derived metrics and cognitive function

3.2.1. Associations between TBR < 3.9 and neuropsychological test results

3.2.2. Associations between FPG and neuropsychological test results

3.2.3. Associations between cued recall scores and key FGM-derived metrics

3.2.4. Associations between TIR/GRI and neuropsychological test results

4. Discussion

5. Conclusion

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1