Group-based trajectory modelling (GBTM)
GBTM was conducted to determine the best-fitting models for two to six groups. In the best-fitting models for each number of groups, we found no higher-order polynomial effects of time. All groups followed a linear or intercept-only trajectory. Each best-fitting model per group number is detailed in Table 1.
Table 1
GBTM fit statistics for two to six group models.
k
|
Trajectories1
|
Log-likelihood
|
AIC2
|
BIC3
|
ΔBIC4
|
Entropy
|
2
|
1 1
|
-7880.60
|
-7887.60
|
-7906.10
|
|
0.93
|
3
|
1 1 2
|
-7692.99
|
-7704.99
|
-7736.70
|
169.40
|
0.86
|
4
|
1 0 0 1
|
-7729.85
|
-7742.85
|
-7777.20
|
-40.50
|
0.91
|
5
|
1 0 1 1 1
|
-7596.91
|
-7614.91
|
-7662.48
|
114.72
|
0.83
|
6
|
1 1 1 0 1 0
|
-7589.02
|
-7610.02
|
-7665.51
|
-3.03
|
0.43
|
k = Number of groups |
1 Listed as order of powers (0 = intercept, 1 = linear, 2 = quadratic) |
2 AIC: Akaike Information Criterion |
3 BIC: Bayesian Information Criterion |
4 ΔBIC: Difference in BIC from previous model (k − 1) |
Comparatively, the AIC and BIC showed that the five- and six-group models were a better fit than models with four groups or less. Optimal BIC was found in the five-group model. This is illustrated by the difference in BIC: ΔBIC was 114.72 for the five-group model whereas the six-group model showed poorer fit with a ΔBIC of -3.03. Entropy was best in two- and four-group models, but acceptable (> 0.8) in models with five or less groups. With all fit information considered, the five-group model was determined the best fit (Supp Fig. 1).
However, the GBTM found no mathematical difference between pregnancy abstainers and women who had consumed some alcohol at the ‘prior to pregnancy recognition’ time point. To accommodate this theoretical implication, whole of pregnancy abstainers were ‘forced’ into their own group, resulting in a final six group solution. Trajectories of the five-group model with abstainer/control separation (group six) are illustrated in Fig. 1. Group means with 95% confidence intervals (CI – shaded area) are presented on a logarithmic scale to illustrate group differences at the lower end of alcohol consumption. The same six alcohol trajectories are also presented on a normal y-scale in Supp Fig. 2.
Examination of the trajectories, which are summarised in Table 2, resulted in six groups which have been named as follows:
Table 2
Alcohol consumption characteristics by trajectory
|
Abstained/
control
|
Low
discontinued
|
Moderate discontinued
|
Low
sustained
|
Moderate
sustained
|
High
sustained
|
N (%)
|
493
|
(33.8)
|
210
|
(14.4)
|
170
|
(11.7)
|
190
|
(13.0)
|
343
|
(23.5)
|
52
|
(3.6)
|
Trimester 1, pre pregnancy recognition
|
Mean gAA1/week (SD)
|
0
|
|
7.8
|
(12.1)
|
77.7
|
(105.9)
|
4.9
|
(9.4)
|
50.0
|
(52.9)
|
226.5
|
(174.8)
|
Median gAA/week (IQR)
|
0
|
|
3.1
|
(1.9–7.2)
|
34.4
|
(13.5–91.0)
|
0.9
|
(0.0-4.7)
|
33.8
|
(8.4–76.9)
|
183.8
|
(118.2-279.1)
|
Binge episode N (%)
|
0
|
|
0
|
|
113
|
(66.5)
|
0
|
|
143
|
(41.7)
|
34
|
(65.4)
|
Trimester 1, post pregnancy recognition
|
Mean gAA/week (SD)
|
0
|
|
0
|
|
0.0
|
(0.1)
|
0.8
|
(1.4)
|
2.6
|
(5.2)
|
18.0
|
(27.8)
|
Median g AA/week (IQR)
|
0
|
|
0
|
|
0.0
|
(0.0–0.0)
|
0.0
|
(0.0-0.9)
|
0.9
|
(0.0-2.8)
|
5.6
|
(0.0-22.5)
|
Binge episode N (%)
|
0
|
|
0
|
|
0
|
|
0
|
|
2
|
(0.6)
|
4
|
(7.7)
|
Trimester 2
|
Mean gAA/week (SD)
|
0
|
|
0
|
|
0
|
|
1.3
|
(1.8)
|
5.3
|
(8.7)
|
46.0
|
(61.0)
|
Median gAA/week (IQR)
|
0
|
|
0
|
|
0
|
|
0.7
|
(0.0-1.6)
|
1.9
|
(0.6–5.6)
|
33.8
|
(6.5–61.7)
|
Binge episode N (%)
|
0
|
|
0
|
|
0
|
|
0
|
|
4
|
(1.2)
|
7
|
(13.5)
|
Trimester 3
|
Mean gAA/week (SD)
|
0
|
|
0
|
|
0
|
|
1.7
|
(4.5)
|
4.8
|
(9.1)
|
27.9
|
(38.9)
|
Median gAA/week (IQR)
|
0
|
|
0
|
|
0
|
|
0.6
|
(0.0-1.3)
|
1.6
|
(0.6–5.3)
|
16.9
|
(3.4–35.5)
|
Binge episode N (%)
|
0
|
|
0
|
|
0
|
|
0
|
|
1
|
(0.3)
|
1
|
(1.9)
|
1 gAA: grams of absolute alcohol |
Abstained/control – no alcohol consumption during pregnancy (33.8% of the total sample);
Low discontinued – median alcohol consumption of 3gAA per week and none post pregnancy recognition i.e. approximately one to two standard drinks, once or twice per month prior to pregnancy recognition (14.4%);
Moderate discontinued – Median alcohol consumption of 34gAA per week (approximately three standard drinks per week) and none post first trimester (11.7%);
Low sustained – continued low median consumption post-awareness (13.0%);
Moderate sustained – Median consumption of 34gAA per week prior to pregnancy recognition and a continued low consumption trend post-awareness (23.5%).;
High sustained – Median consumption of 184gAA per week (approximately three to four standard drinks, three to four times a week) prior to pregnancy recognition and a continued moderate consumption trend post-awareness (3.6%).
Pre-pregnancy recognition binge episodes were confined to the three moderate and high drinking trajectories with the occasional post pregnancy-recognition binge episode in the moderate and high sustained trajectories only. (Table 2)
Both the mean posterior probabilities and odds of correct classification (OCC) were high: Abstained/control = 93.7% and 29.3 (respectively), Low discontinued = 80.9% and 25.1, Moderate discontinued = 80.4% and 31.0, Low sustained = 82.9% and 32.3, Moderate sustained = 85.1% and 18.6, High sustained = 76.6% and 88.4. Overall, the mean posterior probability was 86.2% and OCC was 18.59.
Association between group membership and non-pregnancy related alcohol use behaviour
Pregnant women with a moderate to high alcohol consumption trajectory were less likely to report that they felt the effects of alcohol quickly than controls or women with a low consumption trajectory. Women with moderate to high consumption were also more likely to have experienced their first alcohol intoxication before the Australian legal drinking age of 18 years and to have had at least one binge drinking episode in the three months before pregnancy. (Table 3).
Table 3
Pre-pregnancy binge drinking, drinking age and alcohol sensitivity by trajectory
|
Abstained/ control
|
Low discontinued
|
Moderate discontinued
|
Low sustained
|
Moderate sustained
|
High
sustained
|
|
|
|
n
|
(%)
|
n
|
(%)
|
n
|
(%)
|
n
|
(%)
|
n
|
(%)
|
n
|
(%)
|
p
|
ES
|
|
493
|
|
210
|
|
170
|
|
190
|
|
343
|
|
52
|
|
|
|
Pre-pregnancy binge episode
|
67
|
(13.6)
|
44
|
(21.0)
|
147
|
(86.5)
|
0
|
(0.0)
|
287
|
(83.7)
|
44
|
(84.6)
|
< 0.001
|
0.73
|
Age drinking regularly < 18y
|
133
|
(27.0)
|
64
|
(30.5)
|
65
|
(38.2)
|
45
|
(23.7)
|
165
|
(48.1)
|
17
|
(32.7)
|
0.004
|
0.12
|
Age first intoxication < 18y
|
163
|
(33.1)
|
79
|
(37.6)
|
93
|
(54.7)
|
57
|
(30.0)
|
229
|
(66.8)
|
30
|
(57.7)
|
< 0.001
|
0.25
|
Perceived sensitivity to effects of alcohol
|
|
|
|
|
|
|
|
|
|
|
|
|
Normal
|
187
|
(37.9)
|
93
|
(44.3)
|
100
|
(58.8)
|
81
|
(42.6)
|
226
|
(65.9)
|
32
|
(61.5)
|
< 0.001
|
0.19
|
Very/quickly
|
251
|
(50.9)
|
101
|
(48.1)
|
52
|
(30.6)
|
94
|
(49.5)
|
88
|
(25.7)
|
9
|
(17.3)
|
|
|
Very/slowly
|
45
|
(9.1)
|
14
|
(6.7)
|
17
|
(10.0)
|
11
|
(5.8)
|
29
|
(8.5)
|
11
|
(21.2)
|
|
|
p = p-value, ES = effect size (Cramer’s V), boldface = statistically significant difference between group and control |
Association between group membership and demographic and pregnancy-related characteristics
Multivariate analysis revealed no discernible difference between controls and the low discontinued trajectory in any of the characteristics investigated (Table 4). Compared to controls, women in all other alcohol consumption trajectory groups were two to seven times more likely to be Caucasian (e.g. low sustained (AOR 2.32 (95%CI 1.40–3.85)) and moderate sustained (AOR 7.13 (95%CI 3.99–12.73)). Cigarette smoking in pregnancy was associated with all moderate to high drinking trajectories, e.g. moderate sustained (AOR 4.05 (95%CI 2.60–6.31)) and high sustained (AOR 4.28 (95%CI 1.92–9.54)). Women in the moderate discontinued trajectory were more likely to have an unplanned pregnancy (AOR 2.99 (95%CI 1.92–4.67)) and be pregnant with their firstborn (AOR 2.09 (95%CI 1.38–3.16)). Women with low sustained group membership were less likely to be primipara (AOR 0.61 (95%CI 0.42–0.91)).
Table 4
Demographic and pregnancy-related characteristics predictive of trajectory
Exposure trajectory
|
Abstinent
|
Low discontinued
|
Moderate discontinued
|
Low sustained
|
Moderate sustained
|
High sustained
|
|
N
|
N
|
AOR
|
95% CI
|
p
value
|
N
|
AOR
|
95% CI
|
p
value
|
N
|
AOR
|
95% CI
|
p
value
|
N
|
AOR
|
95% CI
|
p
value
|
N
|
AOR
|
95% CI
|
p
value
|
Participants
|
493
|
210
|
|
|
|
170
|
|
|
|
190
|
|
|
|
343
|
|
|
|
52
|
|
|
|
Maternal age
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
< 30 years
|
179
|
70
|
Ref
|
73
|
Ref
|
42
|
Ref
|
96
|
Ref
|
9
|
Ref
|
30–34 years
|
184
|
82
|
1.20
|
(0.79–1.81)
|
0.40
|
52
|
0.96
|
(0.59–1.56)
|
0.87
|
94
|
1.71
|
(1.07–2.72)
|
0.02
|
160
|
1.63
|
(1.11–2.40)
|
0.01
|
23
|
2.98
|
(1.23–7.20)
|
0.02
|
>=35 years
|
130
|
58
|
1.35
|
(0.86–2.12)
|
0.20
|
45
|
1.16
|
(0.70–1.94)
|
0.57
|
54
|
1.39
|
(0.83–2.32)
|
0.21
|
87
|
1.13
|
(0.74–1.74)
|
0.57
|
20
|
3.70
|
(1.49–9.19)
|
0.01
|
Maternal education
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Secondary
|
98
|
38
|
Ref
|
|
|
32
|
Ref
|
|
|
18
|
Ref
|
|
|
67
|
Ref
|
|
|
11
|
Ref
|
|
|
Trade/Diploma
|
140
|
45
|
0.78
|
(0.46–1.33)
|
0.36
|
69
|
2.05
|
(1.16–3.63)
|
0.01
|
47
|
1.92
|
(1.01–3.66)
|
0.05
|
71
|
0.78
|
(0.48–1.27)
|
0.32
|
12
|
0.78
|
(0.31–1.97)
|
0.60
|
Tertiary
|
255
|
126
|
1.14
|
(0.69–1.86)
|
0.61
|
69
|
1.12
|
(0.61–2.04)
|
0.72
|
124
|
3.00
|
(1.59–5.64)
|
< 0.01
|
207
|
1.43
|
(0.90–2.26)
|
0.13
|
29
|
0.85
|
(0.35–2.05)
|
0.72
|
Household income
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
<$40,000
|
76
|
31
|
Ref
|
|
|
26
|
Ref
|
|
|
31
|
Ref
|
|
|
25
|
Ref
|
|
|
5
|
Ref
|
|
|
$40–70,000
|
112
|
37
|
0.89
|
(0.50–1.59)
|
0.69
|
19
|
0.54
|
(0.26–1.13)
|
0.10
|
36
|
0.80
|
(0.44–1.46)
|
0.46
|
52
|
2.21
|
(1.14–4.30)
|
0.02
|
4
|
0.66
|
(0.16–2.67)
|
0.56
|
$70,000-$100,000
|
125
|
54
|
1.06
|
(0.61–1.85)
|
0.84
|
44
|
1.19
|
(0.63–2.24)
|
0.59
|
40
|
0.62
|
(0.34–1.11)
|
0.11
|
92
|
3.00
|
(1.59–5.68)
|
< 0.01
|
13
|
1.68
|
(0.54–5.25)
|
0.37
|
>$100,000
|
155
|
80
|
1.21
|
(0.70–2.08)
|
0.50
|
74
|
1.81
|
(0.97–3.37)
|
0.06
|
76
|
0.96
|
(0.55–1.67)
|
0.88
|
162
|
4.27
|
(2.28–7.99)
|
< 0.001
|
29
|
3.27
|
(1.08–9.84)
|
0.04
|
Missing
|
25
|
8
|
0.72
|
(0.27–1.92)
|
0.51
|
7
|
0.88
|
(0.30–2.52)
|
0.81
|
7
|
0.65
|
(0.21–1.96)
|
0.44
|
12
|
2.19
|
(0.83–5.77)
|
0.11
|
1
|
0.89
|
(0.09–8.68)
|
0.92
|
Caucasian
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
No
|
120
|
47
|
Ref
|
|
|
15
|
Ref
|
|
|
26
|
Ref
|
|
|
17
|
Ref
|
|
|
3
|
Ref
|
|
|
Yes
|
372
|
163
|
1.23
|
(0.81–1.88)
|
0.33
|
155
|
2.87
|
(1.52–5.42)
|
< 0.01
|
164
|
2.32
|
(1.40–3.85)
|
< 0.01
|
326
|
7.13
|
(3.99–12.73)
|
< 0.001
|
49
|
4.14
|
(1.22–14.1)
|
0.02
|
Pre-pregnancy BMI
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Normal/underweight
|
287
|
142
|
Ref
|
|
|
87
|
Ref
|
|
|
120
|
Ref
|
|
|
239
|
Ref
|
|
|
32
|
Ref
|
|
|
Overweight
|
95
|
32
|
0.71
|
(0.45–1.12)
|
0.14
|
44
|
1.58
|
(0.99–2.54)
|
0.06
|
38
|
0.91
|
(0.58–1.45)
|
0.69
|
57
|
0.68
|
(0.45–1.02)
|
0.06
|
13
|
1.16
|
(0.55–2.42)
|
0.70
|
Obese
|
93
|
29
|
0.62
|
(0.38–1.01)
|
0.06
|
33
|
0.89
|
(0.53–1.50)
|
0.67
|
24
|
0.62
|
(0.37–1.05)
|
0.07
|
35
|
0.36
|
(0.22–0.58)
|
< 0.001
|
7
|
0.41
|
(0.16–1.05)
|
0.06
|
Pregnancy planning
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
No
|
94
|
50
|
1.52
|
(1.00-2.32)
|
0.05
|
66
|
2.99
|
(1.92–4.67)
|
< 0.001
|
40
|
1.39
|
(0.88–2.22)
|
0.16
|
77
|
1.40
|
(0.93–2.09
|
0.10
|
10
|
1.06
|
(0.46–2.43)
|
0.89
|
Yes
|
398
|
159
|
Ref
|
|
|
104
|
Ref
|
|
|
149
|
Ref
|
|
|
266
|
Ref
|
|
|
42
|
Ref
|
|
|
Primipara
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
No
|
259
|
99
|
Ref
|
62
|
Ref
|
123
|
Ref
|
180
|
Ref
|
|
24
|
Ref
|
Yes
|
227
|
108
|
1.25
|
(0.88–1.79)
|
0.22
|
106
|
2.09
|
(1.38–3.16)
|
< 0.01
|
63
|
0.61
|
(0.42–0.91)
|
0.02
|
160
|
0.85
|
(0.61–1.19)
|
0.34
|
28
|
1.40
|
(0.73–2.67)
|
0.31
|
Smoking in pregnancy
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
No
|
429
|
187
|
Ref
|
127
|
Ref
|
173
|
Ref
|
248
|
Ref
|
37
|
Ref
|
Yes
|
63
|
23
|
0.87
|
(0.50–1.54)
|
0.64
|
43
|
1.71
|
(1.03–2.85)
|
0.04
|
17
|
1.04
|
(0.55–1.97)
|
0.90
|
94
|
4.05
|
(2.60–6.31)
|
< 0.001
|
15
|
4.28
|
(1.92–9.54)
|
< 0.001
|
N: Sample size for multivariate analysis (i.e. number of cases with a complete set of predictors, except for income, where a ‘missing’ category was included because ~ 4% of missing data)
Ref: Reference category
AOR (95% CI): Odds ratio and 95% confidence interval adjusted for all predictors shown in table. Control group is abstinence in pregnancy, but not lifetime abstainer. Boldface where a significant difference was found. For predictor variables with more than two categories (maternal age, income, BMI and education), p values from likelihood ratio tests (not shown) were used-evaluate the predictors. Likelihood ratio p values were <0.01 in all bolded results.
A sustained alcohol consumption pattern was more likely in women in their early to mid-thirties and a high sustained level was more likely in women aged 35 years or more. Increasing household income was associated with moderate to high sustained group membership.