The characteristics of the total 351 outpatients with T2DM were shown in Table 1. There were 195 (55.6%) men and 156 (44.4%) women with an average age of 54.4 ± 14.4 years. The average systolic blood pressure (SBP) was 125.5 ± 15.9mmHg, and the average diastolic blood pressure (DBP) was 75.9 ± 10.0 mmHg. The mean BMI was 22.9 ± 3.1 kg/m2. The median duration of diabetes was 6.0 years, and the mean HbA1c was 7.1 ± 1.5% (54 ± 16mmol/mol). According to CGM data, none experienced symptomatic hypoglycemia during the three-day CGM. Overall, AH events were captured in 137 outpatients (39.0%), in whom Level 1 and Level 2 AH accounted for 61.3% and 38.7%, respectively. 112 (31.9%) of the AH population occurred during the daytime, and 85 (24.2%) experienced NAH. Hypoglycemia episodes that occurred exclusively at night accounted for 18.2% of the AH population and 29.4% of the NAH population. The number of hypoglycemic events among those with asymptomatic hypoglycemia occurred exclusively in the daytime (DAH*), asymptomatic hypoglycemia occurred exclusively at night (NAH*), and asymptomatic hypoglycemia occurred in both daytime and nighttime were 1.0 (1.0,4.0), 1.0 (0.0,4.0), and 8.0 (3.0,14.5), respectively. The Total TIR was 0.849 (0.667, 0.960). The MAGE, SD, and MBG were 4.2 ± 2.2mmol/L, 1.9 ± 0.9mmol/L, and 8.2 ± 2.1mmol/L, respectively.
Table 1. Clinical characteristics and CGM data of the enrolled outpatients.
Characteristics
|
n=351
|
Sex
|
|
Male
|
195 (55.6)
|
Female
|
156 (44.4)
|
Age, years
|
55.4±14.4
|
SBP, mmHg
|
125.5±15.9
|
DBP, mmHg
|
75.9±10.0
|
BMI, kg/m2
|
22.9±3.1
|
Duration, years
|
6.0 (2.0,11.0)
|
HbA1c, %
|
7.1±1.5
|
HbA1c, mmol/mol
|
54±16
|
C-P, ng/mL
|
1.4 (0.6,2.0)
|
Comorbidity
|
|
|
Hypertension
|
45 (13.4)
|
|
Hyperlipidemia
|
59 (17.5)
|
|
Fatty liver
|
19 (5.6)
|
|
CCVD
|
37 (11.0)
|
Medication
|
|
|
Metformin
|
97 (28.8)
|
|
DPP-4i
|
57 (16.9)
|
|
SGLT-2i
|
15 (4.5)
|
|
α-GI
|
65 (19.3)
|
|
GLP-1ra
|
6 (1.8)
|
|
TZD
|
9 (2.7)
|
|
SU
|
78 (23.1)
|
|
Insulin threapy
|
137 (40.7)
|
|
Long-acting insulin
|
78 (23.1)
|
|
Premixed insulin
|
36 (10.7)
|
|
Short-acting insulin
|
58 (17.2)
|
Complication
|
|
|
DR
|
54 (16.0)
|
|
DPN
|
45 (13.4)
|
|
DPVD
|
17 (5.0)
|
|
DN
|
17 (5.0)
|
|
DF
|
2 (0.6)
|
CGM data
|
|
AH (%)
|
137 (39.0)
|
Level 1 AH (%)
|
84 (23.9)
|
Level 2 AH (%)
|
53 (15.1)
|
DAH (%)
|
112 (31.9)
|
NAH (%)
|
85 (24.2)
|
DAH*(%)
|
52 (14.8)
|
Hypoglycemic events
|
1.0 (1.0,4.0)
|
Hypoglycemic duration, min
|
12.0 (3.0,37.0)
|
NAH*(%)
|
25 (7.1)
|
Hypoglycemic events
|
1.0 (0.0,4.0)
|
Hypoglycemic duration, min
|
6.0 (0.0,37.0)
|
both DAH and NAH (%)
|
60 (17.1)
|
Hypoglycemic events
|
8.0 (3.0,14.5)
|
Hypoglycemic duration, min
|
68.0 (27.0,135.0)
|
|
MAGE, mmol/L
|
4.2±2.2
|
|
SD, mmol/L
|
1.9±0.9
|
|
MBG, mmol/L
|
8.2±2.1
|
|
Total TIR, %
|
84.9 (66.7,96.0)
|
|
Nighttime TIR, %
|
96.8 (76.0,100.0)
|
|
Daytime TIR, %
|
81.9 (61.6,95.6)
|
|
Total TAR, %
|
12.7 (2.2,32.1)
|
|
Nighttime TAR, %
|
0.0 (0.0,17.1)
|
|
Daytime TAR, %
|
16.2 (2.9,37.6)
|
Data were present as Mean±SG or Median (25th percentile, 75th percentile); SBP, systolic blood pressure; DBP, diastolic blood pressure; BMI, body mass index; HbA1c, glycated hemoglobin; C-P, C-Peptide; CCVD, cardio-cerebral vascular disease; DPP-4i, Dipeptidylpeptidase-4 inhibitors; SGLT-2i, sodium–glucose cotransporter 2 inhibitors; α-GI, alpha-glucosidase inhibitors; GLP-1RA, glucagon-like peptide-1 receptor agonists; TZD, thiazolidinedione; SU, sulfonylurea; DR, diabetic retinopathy; DPN, diabetic peripheral neuropathy; DPVD, diabetic peripheral vascular disease; DN, diabetic nephropathy; DF, diabetic foot; CGM, continuous glucose monitoring; Level 1 AH, BG 3.0-3.9 mmol/L; Level 2 AH, BG <3.0 mmol/L; DAH, daytime asymptomatic hypoglycemia (defined as 6:00-24:00 blood glucose <3.9mmol/L); DAH*, asymptomatic hypoglycemia occurred exclusively in the daytime; NAH, nocturnal asymptomatic hypoglycemia (defined as 0:00-6:00 blood glucose <3.9mmol/L); NAH*, asymptomatic hypoglycemia occurred exclusively at night; both DAH and NAH, asymptomatic hypoglycemia episodes occurred in both daytime and nighttime; MAGE, mean amplitude of plasma glucose excursion; SD, glucose standard deviation; MBG, mean blood glucose; TBR, time below range (<3.9 mmol/L); Total TIR, time in range (3.9–10.0 mmol/L); Nighttime TIR, time in range (3.9–10.0 mmol/L) during 0:00-6:00; Daytime TIR, time in range (3.9–10.0 mmol/L) during 6:00-24:00; Total TAR, time above range (>10.0 mmol/L); Nighttime TAR, time in range (>10.0 mmol/L) during 0:00-6:00; Daytime TAR, time above range (>10.0 mmol/L) during 6:00-24:00.
In Table 2, compared to outpatients without AH, outpatients in the AH group showed younger age (52.1 ± 14.6 years old vs. 55.9 ± 14.1 years old, P = 0.019) and lower HbA1c (6.8 ± 1.3% vs. 7.2 ± 1.6%, P = 0.016). According to the data of CGM, glucose SD (2.2 ± 1.0mmol/L vs. 1.7 ± 0.9mmol/L, P < 0.001) and MAGE (4.8 ± 2.3mmol/L vs. 3.8 ± 2.0mmol/L, P < 0.001) in the AH group were significantly higher than those without AH, while MBG (7.5 ± 1.4mmol/L vs. 8.7 ± 2.2mmol/L, P < 0.001) was significantly lower. Nighttime TIR in the AH group was significantly lower than that in the non-AH group [0.920 (0.758,1.000) vs. 1.000 (0.764,1.000), P = 0.002]. Patients had AH episodes exhibited lower Total TAR [0.092 (0.014,0.250) vs. 0.147 (0.030,0.387), P = 0.009], lower Daytime TAR [0.123 (0.020,0.316) vs. 0.185 (0.041,0.446), P = 0.008], and lower Nighttime TAR [0.000 (0.000,0.100) vs. 0.000 (0.000,0.236), P = 0.018]. In order to determine the potential risk factors of AH, we conducted a series of analyses. After controlling for all factors identified through univariate analyses (Supplementary Table 1), multivariate analysis (Table 3) demonstrated that patients with younger age [0.975 (0.955–0.997), P = 0.023], lower HbA1c levels [0.759 (0.620–0.929), P = 0.008], and higher SBP [1.024 (1.002–1.047), P = 0.033] had a significantly higher risk of experiencing overall AH episodes.
Table 2
Clinical characteristics and CGM data of non-AH and AH patients.
|
Overall AH
|
|
Level 1 AH
|
|
Level 2 AH
|
|
|
non-AH(n = 214)
|
AH(n = 137)
|
P
|
non-AH(n = 267)
|
AH(n = 84)
|
P
|
non-AH(n = 298)
|
AH(n = 53)
|
P
|
Sex
|
|
|
0.260
|
|
|
0.401
|
|
|
0.011
|
Male
|
124(57.9)
|
71(51.8)
|
|
145(54.3)
|
50(59.5)
|
|
174(58.4)
|
21(39.6)
|
|
Female
|
90(42.1)
|
66(48.2)
|
|
122(45.7)
|
34(40.5)
|
|
124(41.6)
|
32(60.4)
|
|
Age, years
|
55.9 ± 14.1
|
52.1 ± 14.6
|
0.019
|
54.2 ± 14.9
|
55.3 ± 12.9
|
0.537
|
55.7 ± 13.7
|
47.0 ± 15.8
|
< 0.001
|
SBP, mmHg
|
125.2 ± 15.9
|
126.0 ± 16.0
|
0.683
|
124.6 ± 15.9
|
128.1 ± 16.0
|
0.100
|
126.1 ± 15.9
|
122.5 ± 15.7
|
0.158
|
DBP, mmHg
|
76.3 ± 10.6
|
75.3 ± 9.0
|
0.430
|
75.8 ± 10.5
|
76.0 ± 8.3
|
0.885
|
76.2 ± 10.0
|
74.2 ± 10.0
|
0.214
|
BMI, kg/m2
|
23.1 ± 3.0
|
22.5 ± 3.1
|
0.089
|
23.1 ± 3.1
|
22.4 ± 2.9
|
0.083
|
22.9 ± 3.0
|
22.8 ± 3.4
|
0.822
|
Duration, years
|
6.0(2.0,11.0)
|
6.0(2.3–11.0)
|
0.642
|
6.0(2.0,12.0)
|
6.0(3.0,11.0)
|
0.806
|
6.0(2.0,11.0)
|
6.0(2.0,14.0)
|
0.732
|
HbA1c, %
|
7.2 ± 1.6
|
6.8 ± 1.3
|
0.016
|
7.2 ± 1.6
|
6.7 ± 1.2
|
0.009
|
7.1 ± 1.5
|
7.0 ± 1.6
|
0.882
|
HbA1c, mmol/mol
|
56 ± 17
|
51 ± 15
|
0.016
|
55 ± 17
|
50 ± 13
|
0.009
|
54 ± 16
|
53 ± 17
|
0.882
|
C-P, ng/mL
|
1.5(0.7,2.2)
|
1.2(0.4,1.9)
|
0.135
|
1.3(0.4,2.0)
|
1.5(1.1,2.6)
|
0.094
|
1.5(0.9,2.2)
|
0.5(0.1,1.6)
|
< 0.001
|
Comorbidity
|
|
|
|
|
|
|
|
|
|
Hypertension
|
27(60.0)
|
18(40.0)
|
0.902
|
31(68.9)
|
14(31.1)
|
0.255
|
41(91.1)
|
4(8.9)
|
0.228
|
Hyperlipidemia
|
36(61.0)
|
23(39.0)
|
0.325
|
43(72.9)
|
16(27.1)
|
0.583
|
52(88.1)
|
7(11.9)
|
0.479
|
Medication
|
|
|
|
|
|
|
|
|
|
Metformin
|
63(64.9)
|
34(35.1)
|
0.325
|
72(74.2)
|
25(25.8)
|
0.695
|
88(90.7)
|
9(9.3)
|
0.068
|
DPP-4i
|
36(63.2)
|
21(36.8)
|
0.693
|
37(64.9)
|
20(35.1)
|
0.038
|
50(88.2)
|
1(1.8)
|
0.002
|
α-GI
|
42(64.6)
|
23(35.4)
|
0.810
|
47(72.3)
|
18(27.7)
|
0.482
|
60(92.3)
|
5(7.7)
|
0.071
|
SU
|
46(21.5)
|
32(23.4)
|
0.682
|
55(20.6)
|
23(27.4)
|
0.192
|
69(23.2)
|
9(17.0)
|
0.319
|
insulin
|
79(57.7)
|
58(42.3)
|
0.324
|
107(78.1)
|
30(21.9)
|
0.389
|
109(79.6)
|
28(20.4)
|
0.019
|
Complication
|
|
|
|
|
|
|
|
|
|
DR
|
32(59.3)
|
22(40.7)
|
0.796
|
43(79.6)
|
11(20.4)
|
0.459
|
43(79.6)
|
11(20.4)
|
0.212
|
DPN
|
31(68.9)
|
14(31.1)
|
0.234
|
37(82.2)
|
8(7.8)
|
0.351
|
39(86.7)
|
6(13.3)
|
0.761
|
DPVD
|
11(64.7)
|
6(35.3)
|
0.737
|
14(82.4)
|
3(17.6)
|
0.712
|
14(82.4)
|
3(17.6)
|
0.999
|
DN
|
11(64.7)
|
6(35.3)
|
0.737
|
13(76.5)
|
4(23.5)
|
0.999
|
15(88.2)
|
2(11.8)
|
0.988
|
DF
|
2(100.0)
|
0(0.0)
|
0.522
|
2(100.0)
|
0(0.0)
|
0.999
|
2(100.0)
|
0(0.0)
|
0.999
|
CGM data
|
|
|
|
|
|
|
|
|
|
MAGE, mmol/L
|
3.8 ± 2.0
|
4.8 ± 2.3
|
< 0.001
|
4.1 ± 2.2
|
4.4 ± 2.0
|
0.210
|
4.0 ± 2.0
|
5.2 ± 2.5
|
< 0.001
|
SD, mmol/L
|
1.7 ± 0.9
|
2.2 ± 1.0
|
< 0.001
|
1.9 ± 0.9
|
1.9 ± 0.9
|
0.702
|
1.8 ± 0.9
|
2.5 ± 1.0
|
< 0.001
|
MBG, mmol/L
|
8.7 ± 2.2
|
7.5 ± 1.4
|
< 0.001
|
8.5 ± 2.1
|
7.4 ± 1.3
|
< 0.001
|
8.4 ± 2.1
|
7.6 ± 1.6
|
0.005
|
Total TIR, %
|
85.3(61.3,96.9)
|
84.8(70.8,95.4)
|
0.906
|
83.4(62.9,94.7)
|
89.4(79.7,96.8)
|
0.006
|
86.6(66.7,96.9)
|
78.1(67.0,86.0)
|
< 0.001
|
Nighttime TIR, %
|
100.0(76.4,100.0)
|
92.0(75.8,100.0)
|
0.002
|
97.2(73.7,100.0)
|
96.8(80.6,100.0)
|
0.640
|
99.4(79.7,100.0)
|
81.5(65.5,95.8)
|
< 0.001
|
Daytime TIR, %
|
81.1(55.4.95.9)
|
84.7(67.8,95.3)
|
0.391
|
79.7(57.5,94.4)
|
89.3(73.7,96.8)
|
0.004
|
84.4(60.9,96.1)
|
76.0(63.3,85.9)
|
0.024
|
Total TAR, %
|
14.7(3.0,38.7)
|
9.2(1.4,25.0)
|
0.009
|
15.1(3.1,36.6)
|
6.2(1.2,20.1)
|
0.001
|
12.0(2.3,32.9)
|
18.2(3.9,28.6)
|
0.794
|
Nighttime TAR, %
|
0.0(0.0,23.6)
|
0.0(0.0,10.0)
|
0.018
|
0.0(0.0,20.4)
|
0.0(0.0,2.5)
|
0.001
|
0.0(0.0,15.5)
|
0.0(0.0,19.9)
|
0.383
|
Daytime TAR, %
|
18.5(4.1,44.6)
|
12.3(2.0,31.6)
|
0.008
|
18.8(4.1,41.2)
|
8.2(1.6,26.3)
|
0.002
|
15.1(2.9,38.5)
|
19.0(5.3,34.8)
|
0.885
|
Data were present as Mean ± SG or Median (25th percentile, 75th percentile). P < 0.05 was considered statistically significant. |
Table 3
Predictors of AH, Level 1 AH, and Level 2 AH on the basis of logistic regression models.
|
|
|
Multivariate analysis
|
|
|
|
AH
|
Level 1 AH
|
Level 2 AH
|
Predictor
|
OR (95%Cl)
|
P
|
OR (95%Cl)
|
P
|
OR (95%Cl)
|
P
|
Sex
|
0.822(0.494–1.368)
|
0.450
|
1.610(0.886–2.923)
|
0.118
|
0.391(0.191–0.799)
|
0.010
|
Age
|
0.975(0.955–0.997)
|
0.023
|
0.996(0.971–1.021)
|
0.746
|
0.962(0.936–0.990)
|
0.008
|
Duration
|
1.016(0.969–1.066)
|
0.506
|
0.986(0.931–1.044)
|
0.626
|
1.041(0.978–1.108)
|
0.211
|
Metformin
|
0.816(0.440–1.513)
|
0.519
|
0.943(0.492–1.883)
|
0.867
|
0.715(0.274–1.862)
|
0.492
|
DPP-4i
|
0.937(0.460–1.908)
|
0.858
|
2.011(0.927–4.361)
|
0.077
|
0.110(0.014–0.855)
|
0.035
|
α-GI
|
0.993(0.517–1.906)
|
0.982
|
1.272(0.627–2.577)
|
0.505
|
0.647(0.225–1.859)
|
0.419
|
SU
|
1.218(0.635–2.334)
|
0.553
|
1.112(0.545–2.272)
|
0.770
|
1.361(0.511–3.624)
|
0.538
|
HbA1c
|
0.759(0.620–0.929)
|
0.008
|
0.683(0.527–0.887)
|
0.004
|
0.940(0.727–1.216)
|
0.639
|
insulin
|
1.261(0.682–2.333)
|
0.460
|
1.197(0.593–2.414)
|
0.616
|
1.198(0.510–2.813)
|
0.679
|
SBP
|
1.024(1.002–1.047)
|
0.033
|
1.033(1.008–1.059)
|
0.010
|
1.000(0.969–1.032)
|
0.999
|
DBP
|
0.981(0.950–1.013)
|
0.238
|
0.980(0.946–1.016)
|
0.272
|
0.991(0.946–1.038)
|
0.692
|
BMI
|
0.961(0.882–1.047)
|
0.365
|
0.920(0.831–1.018)
|
0.108
|
1.063(0.944–1.198)
|
0.310
|
OR: odds ratio; CI: confidence interval. |
As for Level 1 AH, patients with lower HbA1c exhibited a higher prevalence of AH (6.7 ± 1.2% vs. 7.2 ± 1.6%, P = 0.009). There was significant statistical difference in the use of Dipeptidylpeptidase-4 inhibitors (DPP4i) (35.1% vs. 64.9%) between AH group and non-AH group. The MBG of the AH group was substantially lower (7.4 ± 1.3 vs. 8.5 ± 2.1, P < 0.001) while the SD and MBG showed no association. Total TIR [0.894 (0.797,0.968) vs. 0.834 (0.629,0.947), P = 0.006] and Daytime TIR [0.893 (0.737,0.968) vs. 0.797 (0.575,0.944), P = 0.004] were corelated with AH episodes. Patients with lower Total TAR [0.062 (0.012,0.201) vs. 0.051 (0.031,0.366), P = 0.001], lower Nighttime TAR [0.000 (0.000,0.250) vs. 0.000 (0.000,0.204), P = 0.001], and lower Daytime TAR [0.082 (0.016,0.263) vs. 0.188 (0.041,0.412), P = 0.002] were more vulnerable to AH. Multivariate analyses (Table 3) controlled for factors associated with high regimen distress in univariate analyses (Supplementary Table 1) and found that lower HbA1c levels [0.683 (0.527–0.887), P = 0.004] and higher SBP [1.033 (1.008–1.059), P = 0.010] were also associated with Level 1 AH.
There was greater percentage of female in the Level 2 AH group (60.4% vs. 39.6%, P = 0.011). The age (47.0 ± 15.8 vs. 55.7 ± 13.7, P < 0.001) of the AH group was significantly lower and the C-Peptide (C-P) [0.5 (0.1,1.6) vs. 1.5 (0.9,2.2), P < 0.001] level was lower. Significantly, fewer patients in the AH group received DPP4i than the non-AH group (1.8% vs. 88.2%, P = 0002). The MAGE (5.2 ± 2.5mmol/L vs. 4.0 ± 2.0mmol/L, P < 0.001) and glucose SD (2.5 ± 1.0mmol/L vs. 1.8 ± 0.9mmol/L, P < 0.001) in the AH group were significantly higher than those without AH, while MBG (7.6 ± 1.6mmol/L vs. 8.4 ± 2.1mmol/L, P = 0.005) was significantly lower. There were significant statistical differences in Total TIR [0.781 (0.670,0.860) vs. 0.866 (0.667,0.969), P < 0.001], Nighttime TIR [0.815 (0.655,0.958) vs. 0.994 (0.797,1.000), P < 0.001], and Daytime TIR [0.760 (0.633,0.859) vs. 0.844 (0.609,0.961), P = 0.024] between two groups. After controlling for all factors identified through univariate analyses (Supplementary Table 1), multivariate analyses found that female sex [0.391 (0.191–0.799), P = 0.010] and younger age [0.962 (0.936–0.990), P = 0.008] were independent predictors of Level 2 AH episodes. In addition, patients treated with DPP4i tended to have lower risk of Level 2 AH [0.110 (0.014–0.855), P = 0.035]. (Table 3).