Characteristics of study participants
A total of 1586 pregnant women were included in this study and 741 were diagnosed with GDM (Table 1). The mean maternal age was 30.24±4.18 years. Pre-pregnancy BMI (mean) was 21.74±4.51 kg/m2 and blood collection gestational age (mean) was 11.05±1.73 weeks for early pregnancy and 24.22±2.81 weeks for mid-gestation. In our study, none of the women smoke or drink for at least three months before enrolment, thus we did not show the data on smoking and drinking status. Women with GDM were more likely to be older and have a higher pre-pregnancy BMI compared with those without GDM. In early and mid-gestation, pregnant women with GDM had higher levels of TC, triglyceride glucose index, TG/HDL-c, and TC/HDL-c than those in the non-GDM group, whereas HDL-c, on the contrary, had lower levels than those in the non-GDM group. TG, TC, ApoA1, ApoB, HDL-c, LDL-c, triglyceride glucose index, TG/HDL-c, and TC/HDL-c tended to increase from early to mid-pregnancy regardless of GDM status (Figure S1).
Table1 Baseline Characteristics of Subjects in Study.
Characteristics
|
Non-GDM (n=845)
|
GDM (n=741)
|
P value
|
Maternal age, year
|
29.82±4.12
|
30.65±4.24
|
<0.001
|
Parity
|
|
|
0.185
|
0
|
527(62.4)
|
438(59.1)
|
|
≥1
|
318(37.6)
|
303(40.9)
|
|
Educational Level
|
|
|
0.321
|
Middle school or below
|
68(8.00)
|
66(8.90)
|
|
High school or technical
|
98(11.60)
|
102(13.80)
|
|
Undergraduate or above
|
679(80.40)
|
573(77.30)
|
|
Marital status
|
|
|
<0.001
|
Married
|
827(0.98)
|
700(0.94)
|
|
Other
|
18(0.02)
|
41(0.06)
|
|
Mode of conception
|
|
|
0.012
|
Natural conception
|
814(96.30)
|
729(98.40)
|
|
Assisted reproductive conception
|
31(3.70)
|
12(1.60)
|
|
Pre-pregnancy BMI, kg/m2
|
21.23±5.51
|
22.25±3.50
|
<0.001
|
Fertile season
|
|
|
<0.001
|
Spring
|
448(53.00)
|
248(33.50)
|
|
Summer
|
249(29.50)
|
60(8.10)
|
|
Autumn
|
15(1.80)
|
5(0.70)
|
|
Winter
|
133(15.70)
|
428(46.70)
|
|
OGTT fasting, mmol/L
|
4.41±0.28
|
4.82±0.45
|
<0.001
|
OGTT 1-hour, mmol/L
|
7.51±1.38
|
9.93±1.46
|
<0.001
|
OGTT 2-hour, mmol/L
|
6.41±1.08
|
8.40±1.43
|
<0.001
|
Early pregnancy
|
|
|
|
Gestational age at blood collection, weeks
|
11.01±1.76
|
11.09±1.70
|
0.362
|
TC (mmol/L)
|
4.59±0.76
|
4.67±0.77
|
0.032
|
TG (mmol/L)
|
1.36±0.77
|
1.52±0.63
|
<0.001
|
ApoA1
|
1.41±0.24
|
1.44±0.24
|
0.003
|
ApoB
|
0.73±0.17
|
0.77±0.17
|
<0.001
|
HDL-c (mmol/L)
|
1.69±0.30
|
1.62±0.31
|
<0.001
|
LDL-c (mmol/L)
|
2.36±0.61
|
2.43±0.62
|
0.018
|
Triglyceride glucose index
|
1.1±0.37
|
1.23±0.40
|
<0.001
|
TG/HDL-c ratio
|
0.84±0.55
|
0.99±0.55
|
<0.001
|
TC/HDL-c ratio
|
2.77±0.54
|
2.95±0.58
|
<0.001
|
Middle pregnancy
|
|
|
|
Gestational age at blood collection, weeks
|
24.30±2.89
|
24.13±2.72
|
0.377
|
TC (mmol/L)
|
6.19±1.01
|
6.11±1.10
|
0.299
|
TG (mmol/L)
|
2.42±0.98
|
2.55±1.13
|
0.107
|
ApoA1
|
1.80±0.27
|
1.77±0.31
|
0.228
|
ApoB
|
1.04±0.24
|
1.08±0.24
|
0.029
|
HDL-c (mmol/L)
|
2.00±0.38
|
1.88±0.37
|
<0.001
|
LDL-c (mmol/L)
|
3.36±0.82
|
3.37±0.89
|
0.891
|
Triglyceride glucose index
|
1.64±0.37
|
1.75±0.4
|
0.002
|
TG/HDL-c ratio
|
1.31±0.83
|
1.47±1.13
|
0.031
|
TC/HDL-c ratio
|
3.22±1.42
|
3.32±0.66
|
0.251
|
Abbreviations: BMI, body mass index; GDM, gestational diabetes mellitus; TC, total cholesterol; TG, triglyceride; LDL-c, low-density lipoprotein cholesterol; HDL-c, high-density lipoprotein cholesterol; ApoA1, Apolipoprotein AI; ApoB, Apolipoprotein B; OGTT, oral glucose tolerance test.
Association between lipid profiles and lipid ratios and GDM in early pregnancy
The mean of TC, TG, ApoA1, ApoB, HDL-c, LDL-c, triglyceride glucose index, TG/HDL-c and TC/HDL-c in early pregnancy were 4.63±1.73, 1.44±0.77, 1.43±0.70, 0.75±0.24, 1.66±0.17, 2.40±0.30, 1.17±0.61, 0.92±0.39, and 2.86±0.62, respectively. In univariate analyses, per 1-SD increments of TC, TG, ApoA1, ApoB, LDL-c, triglyceride glucose index, TG/HDL-c, and TC/HDL-c in early gestation were positively associated with the risk of GDM, except for per 1-SD increments of HDL-c, which were negatively associated with the risk of GDM (P < 0.05, Table 2). After adjusting for maternal age, pre-pregnancy BMI, gestational age at the time of blood collection in early pregnancy, educational level, parity, marital status, mode of conception and fertile season, significant associations between lipid profiles and lipid ratios and GDM remained (all P < 0.05). Compared with women in the lowest quartile, the OR (95% CIs) for GDM in women in the highest quartiles of TG, ApoA1, ApoB, HDL-c, LDL-c, triglyceride glucose index, TG/HDL-c, and TC/HDL-c were 2.040(1.468-2.843), 1.506(1.091-2.082), 1.529(1.110-2.107), 0.639(0.459-0.889), 1.504(1.086-2.086), 1.952(1.398-2.731), 2.127(1.526-2.971), and 2.370(1.700-3.312) (P-trend≤0.05, Table 2, Model 2). In the restricted cubic spline-based model, only the associations between TG, APOB, LDL-c, triglyceride glucose index, TG/HDL-c, and GDM were nonlinear (nonlinear P< 0.05; Figure 1), whereas evidence of nonlinear associations for other lipid parameters and indicators of islet resistance was lacking (nonlinear P≥0.05, Figure 1).
Table 2 ORs (95% CIs) for GDM according to quartiles of maternal lipid profile and lipid ratios in early pregnancy.
Maternal lipid profile and lipid ratios
|
OR (95%CI) for GDM
|
P for trend*
|
Per 1-SD increment
|
Quartile 1
|
Quartile 2
|
Quartile 3
|
Quartile 4
|
OR (95%CI)
|
Early pregnancy
|
|
|
|
|
|
|
TC (mmol/L)
|
3.80
[<4.11]
|
4.36
[4.11-<4.57]
|
4.81
[4.57-<5.09]
|
5.48
[≥5.09]
|
|
|
Model 1
|
Reference
|
0.985
(0.742-1.308)
|
1.356
(1.023-1.800)
|
1.275
(0.961-1.693)
|
0.021
|
1.155
(1.013-1.318)
|
Model 2
|
Reference
|
0.964
(0.700-1.325)
|
1.278
(0.930-1.757)
|
1.332
(0.968-1.834)
|
0.031
|
1.176
(1.014-1.366)
|
TG (mmol/L)
|
0.90
[<1.04]
|
1.17
[1.04-<1.30]
|
1.46
[1.30-<1.66]
|
2.00
[≥1.66]
|
|
|
Model 1
|
Reference
|
1.059
(0.797-1.407)
|
1.425
(1.071-1.900)
|
2.373
(1.783-3.168)
|
<0.001
|
1.509
(1.263-1.820)
|
Model 2
|
Reference
|
1.078
(0.784-1.483)
|
1.186
(0.860-1.644)
|
2.040
(1.468-2.843)
|
<0.001
|
1.291
(1.090-1.560)
|
ApoA1 (g/L)
|
1.17
[<1.26]
|
1.34
[1.26-<1.40]
|
1.48
[1.40-<1.55]
|
1.69
[≥1.55]
|
|
|
Model 1
|
Reference
|
0.867
(0.654-1.150)
|
1.179
(0.891-1.562)
|
1.460
(1.104-1.934)
|
0.002
|
1.885
(1.241-2.888)
|
Model 2
|
Reference
|
0.921
(0.670-1.265)
|
1.136
(0.825-1.565)
|
1.506
(1.091-2.082)
|
0.012
|
1.877
(1.179-3.015)
|
ApoB (g/L)
|
0.60
[<0.60]
|
0.70
[0.60-<0.70]
|
0.80
[0.70-<0.80]
|
0.90
[≥0.80]
|
|
|
Model 1
|
Reference
|
1.153
(0.875-1.521)
|
1.516
(1.139-2.021)
|
1.581
(1.198-2.091)
|
<0.001
|
3.032
(1.679-5.519)
|
Model 2
|
Reference
|
1.136
(0.834-1.548)
|
1.428
(1.033-1.976)
|
1.529
(1.110-2.107)
|
0.006
|
2.636
(1.334-5.243)
|
HDL-c (mmol/L)
|
1.31
[<1.45]
|
1.55
[1.45-<1.64]
|
1.74
[1.64-<1.86]
|
2.02
[≥1.86]
|
|
|
Model 1
|
Reference
|
0.784
(0.592-1.037)
|
0.673
(0.509-0.891)
|
0.545
(0.409-0.725)
|
<0.001
|
0.470
(0.337-0.652)
|
Model 2
|
Reference
|
0.713
(0.518-0.979)
|
0.731
(0.531-1.005)
|
0.639
(0.459-0.889)
|
0.007
|
0.589
(0.401-0.863)
|
LDL-c (mmol/L)
|
1.75
[<2.00]
|
2.19
[2.00-<2.34]
|
2.51
[2.34-<2.73]
|
3.11
[≥2.73]
|
|
|
Model 1
|
Reference
|
1.264
(0.954-1.677)
|
1.420
(1.070-1.887)
|
1.398
(1.053-1.858)
|
0.014
|
1.217
(1.035-1.433)
|
Model 2
|
Reference
|
1.288
(0.939-1.768)
|
1.425
(1.034-1.967)
|
1.504
(1.086-2.086)
|
0.013
|
1.281
(1.064-1.544)
|
Triglyceride glucose index
|
0.75
[<0.91]
|
1.03
[0.91-<1.13]
|
1.25
[1.13-<1.38]
|
1.57
[≥1.38]
|
|
|
Model 1
|
Reference
|
0.928
(0.694-1.241)
|
1.520
(1.142-2.027)
|
2.315
(1.736-3.095)
|
<0.001
|
2.470
(1.880-3.261)
|
Model 2
|
Reference
|
0.883
(0.636-1.223)
|
1.232
(0.892-1.701)
|
1.952
(1.398-2.731)
|
<0.001
|
2.024
(1.492- 2.761)
|
TG/HDL-c ratio
|
0.51
[<0.60]
|
0.70
[0.60-<0.79]
|
0.91
[0.79-<1.07]
|
1.34
[≥1.07]
|
|
|
Model 1
|
Reference
|
1.033
(0.774-1.380)
|
1.595
(1.200-2.124)
|
2.485
(1.865-3.321)
|
<0.001
|
1.881
(1.495-2.396)
|
Model 2
|
Reference
|
1.089
(0.791-1.501)
|
1.424
(1.036-1.960)
|
2.127
(1.526-2.971)
|
<0.001
|
1.513
(1.207-1.946)
|
TC/HDL-c ratio
|
2.29
[<2.48]
|
2.63
[2.48-<2.78]
|
2.92
[2.78-<3.11]
|
3.45
[≥3.11]
|
|
|
Model 1
|
Reference
|
1.461
(1.094-1.954)
|
1.903
(1.428-2.543)
|
2.603
(1.945-3.487)
|
<0.001
|
1.828
(1.519-2.209)
|
Model 2
|
Reference
|
1.417
(1.027-1.957)
|
1.786
(1.292-2.473)
|
2.370
(1.700-3.312)
|
<0.001
|
1.717
(1.381-2.141)
|
Model 1 was an unadjusted model. Model 2 was adjusted for maternal age, educational level, marital status, pre-pregnancy BMI, parity, mode of conception, gestational age at the time of blood collection in early pregnancy, and fertile season.
*Tests for trend were assessed by modeling median values of quartiles of maternal lipid profile and lipid ratios as continuous variables.
Abbreviations: BMI, body mass index; GDM, gestational diabetes mellitus; TC, total cholesterol; TG, triglyceride; LDL-c, low-density lipoprotein cholesterol; HDL-c, high-density lipoprotein cholesterol; ApoA1, Apolipoprotein AI; ApoB, Apolipoprotein B.
Figure1 Restricted cubic splines-based modeling for the associations of maternal lipid profile and lipid ratios in early pregnancy with GDM. The model was adjusted for maternal age, educational level, marital status, pre-pregnancy BMI, parity, mode of conception, gestational age at the time of blood collection in early pregnancy, and fertile season. Abbreviations: BMI, body mass index; GDM, gestational diabetes mellitus; TC, total cholesterol; TG, triglyceride; LDL-c, low-density lipoprotein cholesterol; HDL-c, high-density lipoprotein cholesterol; ApoA1, Apolipoprotein AI; ApoB, Apolipoprotein B.
Association between lipid profiles and lipid ratios and GDM in middle pregnancy
The mean values of TC, TG, ApoA1, ApoB, HDL-c, LDL-c, triglyceride glucose index, TG/HDL-c and TC/HDL-c at mid-pregnancy were 6.15±1.06, 2.49±1.06, 1.79±0.29, 1.06±0.24, 1.94±0.38, 3.37±0.86, 1.70±0.39, 1.39±0.98, 3.27±1.04. Similar trends were observed only between triglyceride glucose index, HDL-c (mmol/L) and risk of GDM in the second trimester of pregnancy, and associations between the rest of the lipid profiles and indicators of pancreatic resistance and GDM have not been found (Table 3). As the 1-SD of Triglyceride glucose index, and HDL-c (mmol/L) increased, the OR (95% CIs) of GDM in model 2 (Table 3) were 1.884(1.076- 3.358) and 0.537(0.336-0.852), respectively. In the fully corrected model, the OR (95% CI) for extreme interquartile comparisons of HDL-c, triglyceride glucose index, TG/HDL-c ratio, and TC/HDL-c ratio were 0.455(0.280-0.734), 2.318(1.278-4.262), respectively, 1.825(1.125-2.973) and 2.037(1.270-3.287) (P<0.05 for trend). In addition, there was a lack of evidence for a non-linear association of TG, TC, ApoA1, ApoB, HDL-c, LDL-c, triglyceride glucose index, TG/HDL-c and TC/HDL-c with GDM in mid-pregnancy (non-linear P>0.005, Figure S2).
Table 3 ORs (95% CIs) for GDM according to quartiles of maternal lipid profile and lipid ratios in middle pregnancy.
Maternal lipid profile and lipid ratios
|
OR (95%CI) for GDM
|
P for trend*
|
Per 1-SD increment
|
Quartile 1
|
Quartile 2
|
Quartile 3
|
Quartile 4
|
OR (95%CI)
|
Middle pregnancy
|
|
|
|
|
|
|
TC (mmol/L)
|
5.03
[<5.49]
|
5.83
[5.49-<6.14]
|
6.44
[6.14-<6.80]
|
7.24
[≥6.80]
|
|
|
Model 1
|
Reference
|
0.613
(0.402-0.931)
|
0.843
(0.554-1.280)
|
0.844
(0.556-1.280)
|
0.762
|
0.928
(0.805-1.068)
|
Model 2
|
Reference
|
0.574
(0.359-0.913)
|
0.834
(0.524-1.327)
|
0.805
(0.503-1.286)
|
0.720
|
0.970
(0.826-1.139)
|
TG (mmol/L)
|
1.60
[<1.85]
|
2.05
[1.85-<2.27]
|
2.55
[2.27-<2.86]
|
3.49
[≥2.86]
|
|
|
Model 1
|
Reference
|
1.268
(0.831-1.936)
|
1.580
(1.041-2.407)
|
1.510
(0.994-2.302)
|
0.031
|
1.125
(0.976-1.308)
|
Model 2
|
Reference
|
1.231
(0.767-1.978)
|
1.532
(0.961-2.449)
|
1.301
(0.812-2.085)
|
0.111
|
1.031
(0.882- 1.214)
|
ApoA1 (g/L)
|
1.47
[<1.57]
|
1.68
[1.57-<1.76]
|
1.86
[1.76-<1.98]
|
2.15
[≥1.98]
|
|
|
Model 1
|
Reference
|
1.230
(0.808-1.874)
|
0.954
(0.627-1.452)
|
0.744
(0.486-1.136)
|
0.099
|
0.727
(0.428-1.217)
|
Model 2
|
Reference
|
1.109
(0.696-1.769)
|
0.823
(0.514-1.317)
|
0.708
(0.439-1.140)
|
0.089
|
0.906
(0.495-1.632)
|
ApoB (g/L)
|
0.80
[<0.90]
|
1.10
[0.90-<1.10]
|
1.20
[1.10-<1.20]
|
1.30
[≥1.20]
|
|
|
Model 1
|
Reference
|
0.875
(0.612-1.252)
|
0.874
(0.546-1.394)
|
0.798
(0.505-1.257)
|
0.337
|
0.496
(0.261-0.928)
|
Model 2
|
Reference
|
0.800
(0.538-1.188)
|
0.851
(0.505-1.427)
|
0.832
(0.501-1.377)
|
0.495
|
0.589
(0.288-1.194)
|
HDL-c (mmol/L)
|
1.53
[<1.68]
|
1.83
[1.68-<1.95]
|
2.06
[1.95-<2.18]
|
2.40
[≥2.18]
|
|
|
Model 1
|
Reference
|
0.626
(0.412-0.947)
|
0.543
(0.356-0.825)
|
0.414
(0.268-0.635)
|
<0.001
|
0.415
(0.274-0.620)
|
Model 2
|
Reference
|
0.590
(0.371-0.935)
|
0.511
(0.319-0.815)
|
0.455
(0.280-0.734)
|
<0.001
|
0.537
(0.336-0.852)
|
LDL-c (mmol/L)
|
2.48
[<2.82]
|
3.08
[2.82-<3.34]
|
3.58
[3.34-<3.84]
|
4.27
[≥3.84]
|
|
|
Model 1
|
Reference
|
0.822(0.539-1.250)
|
0.998(0.657-1.517)
|
1.267
(0.836-1.925)
|
0.180
|
1.012
(0.851-1.203)
|
Model 2
|
Reference
|
0.865
(0.542-1.380)
|
1.106
(0.697-1.758)
|
1.295
(0.811-2.073)
|
0.154
|
1.076
(0.884-1.31)
|
Triglyceride glucose index
|
1.28
[<1.44]
|
1.57
[1.44-<1.67]
|
1.80
[1.67-<1.93]
|
2.13
[≥1.93]
|
|
|
Model 1
|
Reference
|
0.915
(0.547-1.529)
|
1.942
(1.162-3.268)
|
1.765
(1.063-2.950)
|
0.003
|
2.092
(1.300-3.436)
|
Model 2
|
Reference
|
0.976
(0.537-1.776)
|
1.616
(0.894-2.941)
|
2.318
(1.278-4.262)
|
0.008
|
1.884
(1.076- 3.358)
|
TG/HDL-c ratio
|
0.71
[<0.89]
|
1.04
[0.89-<1.19]
|
1.38
[1.19-<1.62]
|
2.01
[≥1.62]
|
|
|
Model 1
|
Reference
|
1.795
(1.170-2.767)
|
2.174
(1.413-3.668)
|
2.146
(1.397-3.318)
|
<0.001
|
1.215
(1.023-1.481)
|
Model 2
|
Reference
|
1.756
(1.091-2.842)
|
2.149
(1.327-3.504)
|
1.825
(1.125-2.973)
|
0.005
|
1.096
(0.917- 1.345)
|
TC/HDL-c ratio
|
2.57
[<2.81]
|
2.97
[2.81-<3.17]
|
3.35
[3.17-<3.58]
|
2.01
[≥3.58]
|
|
|
Model 1
|
Reference
|
1.636
(1.068-2.517)
|
1.519
(0.995-2.328)
|
2.260
(1.475-3.484)
|
<0.001
|
1.094
(0.949-1.336)
|
Model 2
|
Reference
|
1.300
(0.811-2.088)
|
1.322(0.827-2.117)
|
2.037
(1.270-3.287)
|
0.003
|
1.063
(0.915- 1.244)
|
Model 1 was unadjusted. Model 2 was adjusted for maternal age, educational level, marital status, pre-pregnancy BMI, parity, mode of conception, gestational age at the time of blood collection in early pregnancy, and fertile season.
*Tests for trend were assessed by modeling median values of quartiles of maternal lipid profile and lipid ratios as continuous variables.
Abbreviations: BMI, body mass index; GDM, gestational diabetes mellitus; TC, total cholesterol; TG, triglyceride; LDL-c, low-density lipoprotein cholesterol; HDL-c, high-density lipoprotein cholesterol; ApoA1, Apolipoprotein AI; ApoB, Apolipoprotein B.
Association of Patterns of Change in Lipid Profiles and Lipid Ratios with GDM Risk in Early to Middle Pregnancy
In univariate analyses, stable high levels of TG, triglyceride glucose index, TG/HDL-c, and TC/HDL-c were positively associated with the risk of GDM compared with stable low levels, whereas the opposite was true for HDL-c, whose stable high levels were negatively associated with the risk of GDM (Table 4). Except for TG, the above correlations remained significant after adjustment for confounders, with OR(95% CI) of 2.369(1.438-3.940), 1.588(1.077-2.341), 1.921(1.309-2.829), and 0.599 (0.405-0.883) for GDM, respectively (Table 4).
Table 4 Multivariable-adjusted ORs (95% CIs) for the associations between the change patterns of maternal lipid profile and lipid ratios from early to middle pregnancy and GDM.
Groups
|
Early pregnancy
|
Middle pregnancy
|
Model1
OR (95%CI)
|
Model2
OR (95%CI)
|
TC (mmol/L)
|
|
|
|
|
|
low
|
low
|
Reference
|
Reference
|
|
low
|
high
|
1.067(0.663-1.710)
|
0.955(0.558-1.629)
|
|
high
|
low
|
1.340(0.841-2.137)
|
1.140(0.680-1.909)
|
|
high
|
high
|
1.266(0.888-1.806)
|
1.272(0.860-1.887)
|
TG (mmol/L)
|
|
|
|
|
|
low
|
low
|
Reference
|
Reference
|
|
low
|
high
|
0.856(0.515-1.407)
|
0.912(0.516-1.591)
|
|
high
|
low
|
1.193(0.725-1.956)
|
0.957(0.550-1.654)
|
|
high
|
high
|
1.756(1.242-2.489)
|
1.443(0.977-2.131)
|
ApoA1(g/L)
|
|
|
|
|
|
low
|
low
|
Reference
|
Reference
|
|
low
|
high
|
0.669(0.417-1.064)
|
0.734(0.435-1.228)
|
|
high
|
low
|
1.464(0.917-2.349)
|
1.366(0.808-2.318)
|
|
high
|
high
|
0.929(0.647-1.333)
|
0.820(0.544-1.232)
|
ApoB(g/L)
|
|
|
|
|
|
low
|
low
|
Reference
|
Reference
|
|
low
|
high
|
0.741(0.398-1.347)
|
0.879(0.434-1.727)
|
|
high
|
low
|
1.499(1.023-2.203)
|
1.579(1.028-2.432)
|
|
high
|
high
|
1.153(0.789-1.685)
|
1.202(0.791-1.827)
|
HDL-c (mmol/L)
|
|
|
|
|
|
low
|
low
|
Reference
|
Reference
|
|
low
|
high
|
0.563(0.340-0.925)
|
0.566(0.323-0.981)
|
|
high
|
low
|
0.609(0.370-0.996)
|
0.667(0.381-1.159)
|
|
high
|
high
|
0.536(0.378-0.758)
|
0.599(0.405-0.883)
|
LDL-c (mmol/L)
|
|
|
|
|
|
low
|
low
|
Reference
|
Reference
|
|
low
|
high
|
0.920(0.579-1.456)
|
0.826(0.490-1.385)
|
|
high
|
low
|
0.832(0.516-1.331)
|
0.782(0.459-1.320)
|
|
high
|
high
|
1.379(0.965-1.973)
|
1.468(0.985-2.194)
|
Triglyceride glucose index
|
|
|
|
|
|
low
|
low
|
Reference
|
Reference
|
|
low
|
high
|
1.607(0.858-3.006)
|
2.053(0.995-4.255)
|
|
high
|
low
|
1.694(0.951-3.022)
|
1.366(0.695-2.686)
|
|
high
|
high
|
2.589(1.682-4.015)
|
2.369(1.438-3.940)
|
TG/HDL-c ratio
|
|
|
|
|
|
low
|
low
|
Reference
|
Reference
|
|
low
|
high
|
1.009(0.596-1.689)
|
1.289(0.720-2.291)
|
|
high
|
low
|
1.423(0.851-2.376)
|
1.206(0.679-2.134)
|
|
high
|
high
|
2.000(1.419-2.827)
|
1.588(1.077-2.341)
|
TC/HDL-c ratio
|
|
|
|
|
|
low
|
low
|
Reference
|
Reference
|
|
low
|
high
|
0.685(0.403-1.142)
|
0.683(0.380-1.202)
|
|
high
|
low
|
1.112(0.665-1.847)
|
1.049(0.594-1.844)
|
|
high
|
high
|
1.984(1.404-2.813)
|
1.921(1.309-2.829)
|
Multivariable-adjusted ORs (95% CIs) for the associations between the change patterns of maternal lipid profile and lipid ratios from early to middle pregnancy and GDM.
Model 1 was unadjusted. Model 2 was adjusted for maternal age, educational level, marital status, pre-pregnancy BMI, parity, mode of conception, gestational age at the time of blood collection in early pregnancy, and fertile season.
High means levels above the median; while low means levels below the median;
Abbreviations: BMI, body mass index; GDM, gestational diabetes mellitus; TC, total cholesterol; TG, triglyceride; LDL-c, low-density lipoprotein cholesterol; HDL-c, high-density lipoprotein cholesterol; ApoA1, Apolipoprotein AI; ApoB, Apolipoprotein B.
The combined effect of lipid profiles and lipid ratios and their corresponding patterns of change at different times and pre-pregnancy BMI on GDM risk
Elevated BMI is thought to be caused by poor lipid levels with high concentrations of TG, TC, ApoA1, ApoB, LDL-c and low concentrations of HDL-c[21]. We observed a positive association between pre-pregnancy BMI and GDM (Figure S3). Therefore, we investigated the combined effect of lipid profiles and lipid ratios and their corresponding patterns of change in early and mid-pregnancy and pre-pregnancy BMI on the risk of GDM.
Figures 2 and 3 show a heat map of the combined association of pre-pregnancy BMI (x-axis) (ranging from 14.95 kg/m2 to 37.97 kg/m2) and early/mid-pregnancy TC, TG, ApoA1, ApoB, HDL-c, LDL-c, triacylglycerol glucose index, TG/HDL-c, and TC/HDL-c (y-axis) with the incidence (%) of GDM (z-axis; red represents a higher incidence and blue a lower incidence). (z-axis; red represents higher incidence, blue represents lower incidence). A significant difference in the incidence of GDM was found between pre-pregnancy BMI and the combination of ApoA1, HDL-c, and TC/HDL-c based on the interaction terms of pre-pregnancy BMI and early gestation TC, TG, ApoA1, ApoB, HDL-c, LDL-c, triglyceride glucose index, TG/HDL-c, and TC/HDL-c (P for interaction < 0.05; Figure 2). The interaction effect showed that an increase in TC/HDL-c and pre-pregnancy body mass index in early pregnancy was associated with a higher risk of GDM. In contrast, a decrease in ApoA1 and HDL-c in early pregnancy and an increase in prepregnancy body mass index were associated with a higher risk of GDM. A similar effect was observed in the combination of pre-pregnancy BMI and mid-pregnancy TC/HDL-c levels (P for interaction = 0.019) (Figure 3). Testing of the interaction term between pre-pregnancy BMI and the pattern of change in lipid profile and lipid ratios from early to mid-pregnancy (Table 5) revealed that only the pattern of change in Triglyceride glucose index from early to mid-pregnancy with pre-pregnancy BMI also had a combined effect on GDM (P < 0.05).
Figure 2 Combined effects of maternal pre-pregnancy BMI and maternal lipid profile and lipid ratios in early pregnancy on the incidence of GDM. Heat map for the correlation of incidence of GDM (red represents increased risks of GDM, blue represents decreased risks of GDM) according to the interaction of pre-pregnancy BMI and (A)TC, (B)TG, (C)ApoA1, (D)ApoB, (E)HDL-c, (F) LDL-c, (G)Triglyceride glucose index, (H)TG/HDL-c ratio, or(I)TC/HDL-c ratio. Analyses were adjusted for maternal age, educational level, marital status, pre-pregnancy BMI, parity, mode of conception, gestational age at the time of blood collection in early pregnancy, and fertile season.
Figure3 Combined effects of maternal pre-pregnancy BMI and maternal lipid profile and lipid ratios in middle pregnancy on the incidence of GDM. Heat map for the correlation of incidence of GDM (red represents increased risks of GDM, blue represents decreased risks of GDM) according to the interaction of pre-pregnancy BMI and (A)TC, (B)TG, (C)ApoA1, (D)ApoB, (E)HDL-c, (F) LDL-c, (G)Triglyceride glucose index, (H)TG/HDL-c ratio, or(I)TC/HDL-c ratio. Analyses were adjusted for maternal age, educational level, marital status, pre-pregnancy BMI, parity, mode of conception, gestational age at the time of blood collection in early pregnancy, and fertile season.
Table 5 Combined effects of maternal pre-pregnancy BMI and the change patterns of maternal lipid profile and lipid ratios on the incidence of GDM.
Groups
|
Early pregnancy
|
Middle pregnancy
|
Model
OR (95%CI)
|
P for interaction
|
TC (mmol/L)
|
|
|
|
0.946
|
|
low
|
low
|
Reference
|
|
|
low
|
high
|
1.060(0.854-1.328)
|
|
|
high
|
low
|
1.025(0.872-1.208)
|
|
|
high
|
high
|
1.030(0.904-1.173)
|
|
TG (mmol/L)
|
|
|
|
0.069
|
|
low
|
low
|
Reference
|
|
|
low
|
high
|
0.953(0.765- 1.190)
|
|
|
high
|
low
|
1.296(1.053-1.640)
|
|
|
high
|
high
|
1.042(0.919-1.183)
|
|
ApoA1(g/L)
|
|
|
|
0.411
|
|
low
|
low
|
Reference
|
|
|
low
|
high
|
0.965(0.817-1.145)
|
|
|
high
|
low
|
1.034(0.872-1.237)
|
|
|
high
|
high
|
0.910(0.796- 1.039)
|
|
ApoB(g/L)
|
|
|
|
0.732
|
|
low
|
low
|
Reference
|
|
|
low
|
high
|
0.925(0.702-1.182)
|
|
|
high
|
low
|
1.029(0.888-1.197)
|
|
|
high
|
high
|
1.054(0.921-1.212)
|
|
HDL-c (mmol/L)
|
|
|
|
0.642
|
|
low
|
low
|
Reference
|
|
|
low
|
high
|
0.951(0.785- 1.158)
|
|
|
high
|
low
|
0.947(0.804- 1.119)
|
|
|
high
|
high
|
0.917(0.802- 1.048)
|
|
LDL-c (mmol/L)
|
|
|
|
0.946
|
|
low
|
low
|
Reference
|
|
|
low
|
high
|
0.988(0.805- 1.214)
|
|
|
high
|
low
|
1.014(0.859- 1.201)
|
|
|
high
|
high
|
1.037(0.904- 1.190)
|
|
Triglyceride glucose index
|
|
|
|
0.022
|
|
low
|
low
|
Reference
|
|
|
low
|
high
|
0.900(0.682-1.147)
|
|
|
high
|
low
|
1.343(1.061-1.765)
|
|
|
high
|
high
|
1.115(0.949-1.316)
|
|
TG/HDL-c ratio
|
|
|
|
0.238
|
|
low
|
low
|
Reference
|
|
|
low
|
high
|
1.129(0.912-1.432)
|
|
|
high
|
low
|
1.223(0.997-1.533)
|
|
|
high
|
high
|
1.052(0.926-1.196)
|
|
TC/HDL-c ratio
|
|
|
|
0.707
|
|
low
|
low
|
Reference
|
|
|
low
|
high
|
1.060(0.860-1.319)
|
|
|
high
|
low
|
1.117(0.898-1.408)
|
|
|
high
|
high
|
1.062(0.936- 1.204)
|
|
Multivariable-adjusted ORs (95% CIs) for the associations between the change patterns of maternal lipid profile and lipid ratios from early to middle pregnancy and GDM.
The model was adjusted for maternal age, educational level, marital status, pre-pregnancy BMI, parity, mode of conception, gestational age at the time of blood collection in early pregnancy, and fertile season. High means the levels above the median; while low means the levels below the median.
Association between lipid profiles and lipid ratios and plasma glucose values in the OGTT
In early pregnancy, the levels of triglyceride glucose index, TG/HDL-c ratio, and TC/HDL-c ratio are positively correlated with fasting, 1-hour, and 2-hour OGTT glucose levels. The increase of TG and ApoA1 are positively correlated with 1-hour and 2-hour OGTT glucose levels. The increase of HDL-c is negatively correlated with fasting and 1-hour OGTT glucose levels. In mid-pregnancy, the levels of triglyceride glucose index and TG/HDL-c ratio increase, and HDL-c decreases are positively correlated with fasting, 1-hour, and 2-hour OGTT glucose levels. The increase of TG and ApoA1 is positively correlated with 1-hour and 2-hour OGTT glucose levels. Observing the changes of blood lipid spectrum and lipid ratios from early to mid-pregnancy, it is found that compared with the stable low level, the stable high level of triglyceride glucose index, TG/HDL-c ratio, and TC/HDL-c ratio are positively correlated with fasting, 1-hour, and 2-hour OGTT glucose levels (Table 6).
Table 6 Multivariable-adjusted regression coefficients for the association between maternal lipid profile and lipid ratios and plasma glucose values in 75-g OGTT
Maternal lipid profile and lipid ratios
|
OGTT-fasting
|
OGTT-1 hour
|
OGTT-2 hour
|
β (95%CI)
|
P value
|
β (95%CI)
|
P value
|
β (95%CI)
|
P value
|
Early pregnancy
|
|
|
|
|
|
|
TC (mmol/L)
|
-0.001
(-0.028-0.0269)
|
0.964
|
0.051
(-0.073-0.174)
|
0.420
|
0.062
(-0.044-0.168)
|
0.249
|
TG (mmol/L)
|
0.0645
(0.031-0.0980)
|
<0.001
|
0.139
(-0.010-0.288)
|
0.069
|
0.200
(0.071-0.328)
|
0.002
|
ApoA1(g/L)
|
-0.0200
(-0.107-0.067)
|
0.654
|
0.445
(0.059-0.832)
|
0.024
|
0.513
(0.181-0.846)
|
0.003
|
ApoB(g/L)
|
0.123
(-0.003-0.250)
|
0.056
|
0.378
(-0.186-0.941)
|
0.189
|
0.588
(0.103-1.073)
|
0.018
|
HDL-c (mmol/L)
|
-0.169
(-0.239--0.098)
|
<0.001
|
-0.330
(-0.646--0.014)
|
0.041
|
-0.172
(-0.445-0.100)
|
0.215
|
LDL-c (mmol/L)
|
0.034
(-0.001-0.069)
|
0.055
|
0.107
(-0.048-0.261)
|
0.177
|
0.085
(-0.048-0.218)
|
0.211
|
Triglyceride glucose index
|
0.205
(0.148-0.263)
|
<0.001
|
0.416
(0.158-0.674)
|
0.002
|
0.486
(0.257-0.715)
|
<0.001
|
TG/HDL-c ratio
|
0.108
(0.064-0.152)
|
<0.001
|
0.231
(0.036-0.427)
|
0.021
|
0.255
(0.086-0.424)
|
0.003
|
TC/HDL-c ratio
|
0.095
(0.056-0.135)
|
<0.001
|
0.268
(0.092-0.444)
|
0.003
|
0.202
(0.050-0.355)
|
0.010
|
Middle pregnancy
|
|
|
|
|
|
|
TC (mmol/L)
|
-0.033
(-0.063 -0.0023)
|
0.035
|
-0.095
(-0.227-0.038)
|
0.162
|
-0.021
(-0.137-0.094)
|
0.720
|
TG (mmol/L)
|
0.049
(0.019- 0.079)
|
0.001
|
0.073
(-0.057- 0.203)
|
0.270
|
0.171
(0.060- 0.282)
|
0.003
|
ApoA1(g/L)
|
-0.089
(-0.199- 0.021)
|
0.115
|
-0.177
(-0.661- 0.307)
|
0.484
|
0.004
(-0.418-0.425)
|
0.986
|
ApoB(g/L)
|
-0.097
(-0.229- 0.035)
|
0.152
|
-0.363
(-0.936- 0.209)
|
0.214
|
-0.173
(-0.675-0.328)
|
0.499
|
HDL-c (mmol/L)
|
-0.227
(-0.311- -0.143)
|
<0.001
|
-0.591
(-0.964--0.218)
|
0.002
|
-0.485
(-0.810--0.160)
|
0.004
|
LDL-c (mmol/L)
|
-0.015
(-0.052- 0.022)
|
0.436
|
-0.039
(-0.201- 0.124)
|
0.641
|
-0.002
(-0.144-0.141)
|
0.982
|
Triglyceride glucose index
|
0.294
(0.196-0.393)
|
<0.001
|
0.605
(0.167- 1.043)
|
0.007
|
0.710
(0.331-1.089)
|
0.040
|
TG/HDL-c ratio
|
0.085
(0.055- 0.116)
|
<0.001
|
0.131
(-0.005- 0.268)
|
0.060
|
0.182
(0.064- 0.299)
|
0.003
|
TC/HDL-c ratio
|
0.034
(0.008- 0.060)
|
0.011
|
0.068
(-0.046- 0.182)
|
0.242
|
0.063
(-0.036-0.162)
|
0.211
|
Changes from early to middle pregnancy
|
|
|
|
|
|
|
TC (mmol/L)
|
|
|
|
|
|
|
Low, low
|
Reference
|
|
Reference
|
|
Reference
|
|
Low, high
|
-0.026
(-0.124- 0.073)
|
0.611
|
-0.376
(-0.807-0.0558)
|
0.088
|
0.023
(-0.354-0.400)
|
0.906
|
High, low
|
-0.011
(-0.111- 0.089)
|
0.824
|
-0.251
(-0.686- 0.184)
|
0.259
|
0.018
(-0.361-0.397)
|
0.925
|
High, high
|
-0.012
(-0.087- 0.062)
|
0.747
|
-0.051
(-0.378- 0.275)
|
0.758
|
0.069
(-0.216-0.354)
|
0.636
|
TG (mmol/L)
|
|
|
|
|
|
|
Low, low
|
Reference
|
|
Reference
|
|
Reference
|
|
Low, high
|
0.074
( -0.028- 0.176)
|
0.154
|
0.190
(-0.257- 0.637)
|
0.406
|
0.240
(-0.148-0.627)
|
0.226
|
High, low
|
0.050
(-0.053- 0.152)
|
0.342
|
0.066
(-0.392- 0.524)
|
0.777
|
-0.018
(-0.410-0.374)
|
0.930
|
High, high
|
0.046
(-0.029- 0.121)
|
0.233
|
0.216
(-0.113-0.545)
|
0.199
|
0.426
(0.141-0.711)
|
0.004
|
ApoA1(g/L)
|
|
|
|
|
|
|
Low, low
|
Reference
|
|
Reference
|
|
Reference
|
|
Low, high
|
-0.067
(-0.164- 0.029)
|
0.173
|
-0.301
(-0.727-0.125)
|
0.167
|
-0.126
(-0.500-0.247)
|
0.507
|
High, low
|
-0.011
(-0.111- 0.088)
|
0.820
|
-0.013
(-0.448-0.422)
|
0.953
|
0.091
(-0.290-0.473)
|
0.640
|
High, high
|
-0.065
(-0.142- 0.012)
|
0.100
|
-0.039
(-0.380-0.301)
|
0.821
|
0.106
(-0.193-0.404)
|
0.488
|
ApoB(g/L)
|
|
|
|
|
|
|
Low, low
|
Reference
|
|
Reference
|
|
Reference
|
|
Low, high
|
-0.003
(-0.124- 0.119)
|
0.966
|
-0.043
(-0.583-0.497)
|
0.876
|
0.280
(-0.184-0.743)
|
0.237
|
High, low
|
-0.058
(-0.024- 0.140)
|
0.165
|
0.154
(-0.205-0.513)
|
0.402
|
0.376
(0.065-0.687)
|
0.018
|
High, high
|
-0.012
(-0.091- 0.068)
|
0.777
|
0.050
(-0.302-0.402)
|
0.780
|
0.044
(-0.263-0.350)
|
0.781
|
HDL-c (mmol/L)
|
|
|
|
|
|
|
Low, low
|
Reference
|
|
Reference
|
|
Reference
|
|
Low, high
|
-0.098
(-0.198-0.002)
|
0.055
|
-0.309
(-0.758-0.140)
|
0.177
|
-0.303
(-0.700-0.089)
|
0.130
|
High, low
|
-0.131
(-0.235- -0.028)
|
0.013
|
-0.147
(-0.602-0.309)
|
0.528
|
-0.071
(-0.46928034 0.327)
|
0.726
|
High, high
|
-0.151
(-0.225--0.076)
|
<0.001
|
-0.243
(-0.574-0.088)
|
0.151
|
-0.276
(-0.565-0.014)
|
0.062
|
LDL-c (mmol/L)
|
|
|
|
|
|
|
Low, low
|
Reference
|
|
Reference
|
|
Reference
|
|
Low, high
|
0.041
(-0.081- 0.109)
|
0.771
|
-0.214
(-0.628- 0.201)
|
0.313
|
-0.070
(-0.433-0.294)
|
0.708
|
High, low
|
0.030
(-0.071- 0.132)
|
0.556
|
-0.019
(-0.463- 0.424)
|
0.932
|
0.111
(-0.275-0.500)
|
0.574
|
High, high
|
0.023
(-0.053- 0.099)
|
0.550
|
0.148
(-0.185- 0.480)
|
0.385
|
0.131
(-0.161-0.423)
|
0.380
|
Triglyceride glucose index
|
|
|
|
|
|
|
Low, low
|
Reference
|
|
Reference
|
|
Reference
|
|
Low, high
|
0.128
(0.003- 0.252)
|
0.045
|
0.013
(-0.544- 0.569)
|
0.965
|
0.041
(-0.441-0.523)
|
0.867
|
High, low
|
0.032
(-0.085-0.148)
|
0.593
|
-0.167
(-0.689- 0.355)
|
0.531
|
0.030
(-0.424-0.483)
|
0.898
|
High, high
|
0.213
(0.123-0.303)
|
<0.001
|
0.468
(0.069-0.866)
|
0.022
|
0.575
(0.227-0.923)
|
0.001
|
TG/HDL-c ratio
|
|
|
|
|
|
|
Low, low
|
Reference
|
|
Reference
|
|
Reference
|
|
Low, high
|
-0.011
(-0.116-0.094)
|
0.834
|
0.203
(-0.263- 0.667)
|
0.394
|
0.165
(-0.239-0.569)
|
0.424
|
High, low
|
0.095
(-0.010-0.201)
|
0.077
|
0.261
(-0.214- 0.735)
|
0.282
|
0.035
(-0.375- 0.444)
|
0.868
|
High, high
|
0.096
(0.021- 0.172)
|
0.013
|
0.333
(0.003- 0.664)
|
0.048
|
0.437
(0.140-0.714)
|
0.004
|
TC/HDL-c ratio
|
|
|
|
|
|
|
Low, low
|
Reference
|
|
Reference
|
|
Reference
|
|
Low, high
|
0.003
(-0.097- 0.104)
|
0.949
|
0.009
(-0.436- 0.455)
|
0.968
|
0.149
(-0.239-0.537)
|
0.453
|
High, low
|
0.103
(-0.004-0.209)
|
0.589
|
0.280
(-0.189- 0.748)
|
0.243
|
0.024
(-0.387-0.435)
|
0.908
|
High, high
|
0.136
(0.062-0.210)
|
<0.001
|
0.362
(0.036- 0.688)
|
0.030
|
0.305
(0.021-0.589)
|
0.036
|
High means levels above median, while low means levels below median.
Abbreviations: BMI, body mass index; TC, total cholesterol; TG, triglyceride; LDL-c, low-density lipoprotein cholesterol; HDL-c, high-density lipoprotein cholesterol; ApoA1, Apolipoprotein AI; ApoB, Apolipoprotein B; OGTT, oral glucose tolerance test.
The β coefficients are derived from multiple linear regression models.
Models were adjusted for maternal age, educational level, marital status, pre-pregnancy BMI, parity, mode of conception, gestational age at the time of blood collection in early pregnancy, and fertile season.