Arch Gynecol Obstet (2014) 289:771–779 DOI 10.1007/s00404-013-3056-9
MATERNAL-FETAL MEDICINE
Determinants of alcohol cessation, reduction and no reduction during pregnancy Panagiota Kitsantas • Kathleen F. Gaffney Huichuan Wu • Jennifer C. Kastello
•
Received: 20 July 2013 / Accepted: 9 October 2013 / Published online: 23 October 2013 Ó Springer-Verlag Berlin Heidelberg 2013
Abstract Purpose Despite public health initiatives targeting the harmful effects of alcohol exposure on fetal growth, 12 % of pregnant women report current alcohol use. For women who reported drinking alcohol prior to pregnancy, we examined several factors as predictors of three alcohol use patterns during the third trimester of pregnancy: cessation, reduction and no reduction. Methods Using the 2002–2009 Pregnancy Risk Assessment Monitoring System (PRAMS) dataset (311,428 records), a multinomial logistic regression model was constructed to compare alcohol risk by category: (1) cessation vs. reduction (2) no reduction vs. reduction. Results In this sample, 49.4 % drank alcohol before pregnancy. Among those who drank before pregnancy, *87 % quit drinking during pregnancy, 6.6 % reduced, and about 6.4 % reported no reduction. Older women and those with higher education were more likely to reduce than quit their alcohol use. Conversely, women who were black or Hispanic, overweight, obese, or multiparas were more likely to quit than to reduce their prenatal alcohol consumption. Several stressors such as abuse during pregnancy increased their risk of not quitting or not reducing alcohol during the last trimester of pregnancy. Conclusions Differentiating prenatal alcohol use patterns can inform the design of targeted interventions and public P. Kitsantas (&) H. Wu Department of Health Administration and Policy, MS 1J3, College of Health and Human Services, George Mason University, 4400 University Drive, Fairfax, VA 22030, USA e-mail: [email protected] K. F. Gaffney J. C. Kastello School of Nursing, College of Health and Human Services, George Mason University, Fairfax, VA 22030, USA
health policies to meet the Healthy People 2020 objective for achieving a national rate of 98.3 % alcohol abstinence during pregnancy. Keywords Alcohol consumption Pregnancy Alcohol reduction Cessation
Introduction Alcohol use during pregnancy can lead to fetal alcohol spectrum disorders (FASD), a wide range of adverse effects that cause lifelong physical, neurological, and psychological impairments [1–4]. Even low levels of gestational alcohol exposure have been associated with persistent behavioral and emotional problems among affected children [5]. Prevalence rates of FASD in the US and Western Europe have been estimated as high as 2–5 % among young school children [6]. In addition to quality of life issues, the financial impact of prenatal alcohol consumption is high with health care costs estimated to be over $4 billion annually [7, 8]. Since 1981, US Surgeons General have advised that pregnant women and those who anticipate a pregnancy avoid drinking alcohol [9]. Despite longstanding public health initiatives, state and national surveys demonstrate a substantial proportion of women who continue to report alcohol consumption during pregnancy: 15.8 % ages 15–17, 9.8 % ages 18–25, and 12.5 % ages 26–44 years [10]. These rates are significant in light of the Healthy People 2020 national goal for achieving a rate of 98.3 % for alcohol abstinence during pregnancy by the end of this decade [11]. Among the challenges for meeting this goal is the difficulty of correctly identifying and treating women most at risk for prenatal alcohol use. Prior research has examined
123
772
factors that may inform clinical screening and intervention efforts, but findings have been inconsistent and sometimes contradictory. For instance, higher maternal age has been reported as a predictor in some studies, but no statistically significant association has been found in others [12–15]. In separate studies, women with higher maternal education were more likely to drink alcohol during pregnancy, while in other studies those with lower levels of education were more likely to do so [16, 17]. Number of prior pregnancies has been a significant correlate in some research, while others have found no association [15, 18, 19]. Smoking and exposure to abuse have been reported consistently to increase risk [13, 14, 20]. Another factor that has been consistently associated with alcohol use during pregnancy is drinking alcohol prior to pregnancy recognition [12, 13, 19]. Among studies that have examined prepregnancy alcohol use, however, we know of none that has examined predictors of subsequent patterns of prenatal consumption (e.g., cessation, reduction, no reduction). Unveiling differentiated risk profiles may inform the inconsistent findings of prior research and provide evidence for more targeted, sub-group interventions to improve health outcomes for women and infants who experience alcohol-exposed pregnancies. To provide empirical evidence in support of future clinical interventions and public health initiatives that reduce the prevalence of prenatal alcohol use, it is essential to differentiate groups of women based on patterns of change in alcohol consumption between preconception and pregnancy. The aim of this study was to examine several factors as predictors of three alcohol use patterns (cessation, reduction, no reduction) between preconception and the third trimester of pregnancy for women who report drinking alcohol prior to pregnancy.
Methods Sample PRAMS is coordinated by the Centers for Disease Control and Prevention (CDC) in conjunction with state health departments and is designed to gather data on maternal attitudes, behaviors, and experiences during and around the time of pregnancy. The PRAMS data set allows CDC and the states to monitor changes in maternal and child health indicators, and measure progress towards improving the health of mothers and their children. The PRAMS sample is based on all women in the US who had a live birth. Therefore, findings from these data can be generalized for the population of women with a recent live birth. The PRAMS sample selection is based on the annual birth certificate files of participating states and involves stratified
123
Arch Gynecol Obstet (2014) 289:771–779
selection of 1,300–3,400 cases from each participating state from the population of mothers who experienced a recent live birth. Surveys are distributed by mail *2–6 months following delivery. PRAMS questionnaire responses are linked with birth certificate data to allow for simultaneous examination of demographic and medical information collected through state vital records systems. To ensure that high risk populations are adequately represented, some groups of mothers in the PRAMS dataset are sampled at a higher rate. This discrepancy in sampling is adjusted using appropriate sampling weights. Further information about PRAMS design and methods is available at http://www.cdc.gov/prams. We used PRAMS data from the 37 participating states for the years 2002–2009. The sample consisted of 362,752 singleton births. From these records, 311,428 cases with singleton births and with no missing data on the study variables comprised the analytic sample. The distribution of sociodemographic characteristics between those cases that were included in the analytic sample and those that were excluded due to missing data was similar. This study has been approved by the Institutional Review Board at George Mason University. CDC approved this study and provided the data set for the analyses. Outcome variable Categorization of the pattern of alcohol consumption was based on responses to two items in the PRAMS survey: (1) During the 3 months before you got pregnant, how many alcoholic drinks did you have in an average week? and (2) During the last 3 months of your pregnancy, how many alcoholic drinks did you have in an average week?. To reduce the risk of recall bias that has been associated with self-reports of the specific number of drinks consumed, we used an approach from previous research that categorized this outcome variable [21]. The category used for women who reported no alcohol consumption before or during pregnancy was abstinence. The remaining respondents were women who reported at least weekly preconceptual alcohol consumption. Based on women’s responses to the two alcohol-related PRAMS items, their pattern of alcohol consumption between the preconception (3 months before pregnancy) and prenatal timeframes (last 3 months of pregnancy) was categorized as: cessation (consumed alcohol weekly before pregnancy, but not during pregnancy), reduction (continued the consumption of alcohol during the last 3 months of pregnancy at a lower amount compared to the last 3 months before pregnancy), or no reduction (continued alcohol consumption the same or more during the last 3 months of pregnancy compared to the preconceptual period). Potential predictors were selected from several categories,
Arch Gynecol Obstet (2014) 289:771–779
including sociodemographic, health behavior and psychosocial factors.
773
history,
health
Sociodemographic factors Women self-reported their age, race/ethnicity, education, and marital status. Race/ethnicity consisted of non-Hispanic black, non-Hispanic white, Hispanic of any race, and other which included Asian/Pacific Islander, and American Indian/Alaskan. Maternal education was classified as less than high school (B11 years), high school (12 years), some college (13–15 years), and 4 years of college or more (C16 years). The mother’s age (B19, 20–24, 25–29, 30–34, C35), and marital status (married, other) were extracted from the linked child birth certificate. Health history factors Self-reported prepregnancy height and weight were used to calculate body mass index (BMI). Mothers were classified as underweight (\19.8), normal weight (19.8–26.0), overweight ([26.0–29.0) and obese ([29.0) based on the Institute of Medicine BMI criteria [22]. Prenatal health problems (yes, no) were determined based on responses to a list of clinical experiences women may have encountered during pregnancy including gestational diabetes, pregnancy-induced hypertension, preeclampsia, toxemia and premature rupture of membranes. Previous live births were classified as none, 1, or C2. Health behavior factors Smoking during the last 3 months of pregnancy was a dichotomous variable (yes, no). Women were classified as having timely prenatal care if they initiated prenatal care during the first trimester and untimely if they began prenatal care after the first trimester or received no prenatal care. Psychosocial factors To assess for abuse during pregnancy, the PRAMS survey asked women to signify whether they had been ‘‘physically hurt in any way by your husband or partner’’ during their pregnancy (yes, no). To assess for other psychosocial factors, respondents were provided with a list of potential untoward events and asked to indicate (yes, no) whether they had experienced any during the 12 months before their baby was born. Examples included: ‘‘A close family member was very sick and had to go into the hospital’’ or ‘‘I got separated or divorced from my husband or partner’’, ‘‘I was homeless’’, ‘‘Someone close to me had a bad problem with drinking or drugs’’, and ‘‘I had a lot of bills I
couldn’t pay’’. Whether the current pregnancy was wanted or not was assessed based on an item that asked respondents to characterize how they felt about becoming pregnant during the immediate time just before the current pregnancy (not wanted now, wanted). Statistical analysis Bivariate analyses were conducted to compare alcohol use or nonuse during the third trimester of pregnancy by several factors (Table 1). Specifically, in this combined sample (311, 428), we first examined the prevalence of alcohol use before and during pregnancy as well as associations between alcohol consumption status during the last 3 months of pregnancy and several sociodemographic, behavioral, health and psychosocial related factors. Subsequently, alcohol use was further divided into categories of cessation, reduction, or no reduction to unveil differences in risk factors that may exist within these more discrete groups (Table 2). We used the Chi-square statistic to identify significant associations. Multinomial logistic regression was conducted to determine correlates of prenatal alcohol cessation and no reduction relative to alcohol reduction. These analyses were important for identifying which of the predictor variables we considered in this study favor alcohol cessation during pregnancy and which ones increase the risk of no reduction in alcohol consumption among mothers who drank alcohol prior to pregnancy (Table 3). Odds ratios (95 % confidence intervals) were reported to establish the magnitude and direction of any significant effects. In the multinomial logistic regression, reduction was the reference group, cessation was coded as one, and no reduction was coded as two. All analyses were weighted to account for the complex survey design of PRAMS.
Results In this sample, 10.6 % of the women were 19 years old or younger, 75.7 % were between the ages of 20 and 34 years old and 13.7 % were 35 years old or older. Approximately 62 % were white non-Hispanic, 16 % were black nonHispanic and 13.3 % were Hispanic women. About 63 % were married; 18.8 % had less than high school education and 28.1 % had at least 16 years of education. Weighted analyses revealed that approximately half (49.4 %, unweighted n = 152,203) of the respondents reported that they consumed alcoholic beverages in a typical week during the 3 months prior to pregnancy. For the majority of this portion of the sample (unweighted n = 152,203), the pattern of change in alcohol consumption during the last trimester of pregnancy was categorized
123
774
Arch Gynecol Obstet (2014) 289:771–779
Table 1 Prenatal alcohol use for woman who drank alcohol prior to pregnancy (weighted estimates) Life course factors
No alcohol use during third trimester % (SE)
Table 1 continued Life course factors
No alcohol use during third trimester % (SE)
Some alcohol use during third trimester % (SE)
No
90.9 (0.1)
91.4 (0.3)
Yes
9.1 (0.1)
8.6 (0.3)
96.6 (0.1)
97.1 (0.2)
3.4 (0.1)
2.9 (0.2)
No
75.6 (0.1)
78.1 (0.5)
Yes
24.4 (0.1)
21.9 (0.5)
Some alcohol use during third trimester % (SE) Divorce
Sociodemographic Age B19
10.1 (0.1)
3.8 (0.2)
20–24
25.1 (0.1)
14.9 (0.4)
25–29
28.6 (0.1)
24.1 (0.5)
30–34
23.2 (0.1)
32.9 (0.5)
C35
13.1 (0.1)
24.3 (0.5)
64.6 (0.1)
75.3 (0.3)
Black (non-Hispanic)
16.0 (0.1)
10.3 (0.3)
No
74.5 (0.1)
75.7 (0.5)
Hispanic
10.8 (0.1)
8.5 (0.4)
Yes
25.5 (0.1)
24.3 (0.5)
8.6 (0.1)
5.9 (0.3)
Less than high school
16.0 (0.1)
8.9 (0.3)
High school
30.6 (0.1)
21.4 (0.5)
Some college
24.5 (0.1)
21.3 (0.5)
29.0 (0.1)
48.4 (0.6)
Married
64.5 (0.1)
74.1 (0.5)
Other
35.5 (0.1)
25.9 (0.5)
Under weight
13.0 (0.1)
13.3 (0.4)
Normal weight
50.8 (0.2)
59.9 (0.6)
Overweight
13.9 (0.1)
11.8 (0.4)
Obese
23.2 (0.1)
15.0 (0.4)
Bivariate analyses Table 1 displays findings related to the bivariate analyses which show associations between selected factors and prenatal alcohol use. The sociodemographic factors of being over 30 years of age, non-Hispanic white, having attained higher education, and being married were associated with an increased likelihood of alcohol consumption during pregnancy. Health status related factors such as having a normal BMI, experiencing health problems during pregnancy and an obstetrical history of one prior live birth were associated with prenatal alcohol use. Smoking during pregnancy was a negative health behavior linked with alcohol use. Psychosocial factors associated with alcohol consumption were the experiences of abuse during pregnancy and having someone close to you who has drinking and/or drug problems. To further determine whether these factors may be differentially associated with specific patterns of alcohol use during the last trimester of pregnancy, Chi-square analyses
Race/ethnicity White (non-Hispanic)
Other Education
4 years college or more Marital status
Health status Prepregnancy BMI
Homeless No Yes Could not pay bills
Family member ill
Someone close has drinking/drug problems No
87.3 (0.1)
86.0 (0.4)
Yes
12.7 (0.1)
14.0 (0.4)
Wanted
89.8 (0.1)
89.9 (0.3)
Did not want
10.2 (0.1)
10.1 (0.3)
Current pregnancy
as cessation (87 %, unweighted n = 133,861). However, a portion of the women in this study who were drinking before pregnancy indicated persistent alcohol use (6.6 % reduction, unweighted n = 8,909; 6.4 % no reduction, unweighted n = 9,433) while pregnant.
Health problems during pregnancy No
29.9 (0.1)
26.6 (0.5)
Yes
70.1 (0.1)
73.4 (0.5)
None
41.8 (0.1)
39.0 (0.5)
One C2
32.1 (0.1) 26.0 (0.1)
35.0 (0.5) 26.0 (0.5)
No
87.1 (0.1)
84.5 (0.4)
Yes
12.9 (0.1)
15.5 (0.4)
95.4 (0.1)
95.8 (0.2)
4.6 (0.1)
4.2 (0.2)
No
96.2 (0.1)
94.8 (0.3)
Yes
3.8 (0.1)
5.2 (0.3)
Previous live births
Health behaviors Smoked last trimester
Prenatal care initiation Timely (1st trimester) Untimely (2nd, 3rd, none) Psychosocial Abuse during pregnancy
123
Arch Gynecol Obstet (2014) 289:771–779
775
Table 2 Patterns of alcohol use during the last trimester of pregnancy by several factors for women who drank alcohol prior to pregnancy (weighted estimates) Life course factors
Cessation % (SE)
Reduction % (SE)
No reduction % (SE)
Sociodemographic Age B19
6.6 (0.1)
1.7 (0.2)
5.9 (0.4)
20–24
24.8 (0.2)
10.8 (0.5)
19.1 (0.7)
25–29
30.1 (0.2)
22.7 (0.7)
25.5 (0.1)
30–34
24.9 (0.2)
37.2 (0.8)
28.6 (0.7)
C35
13.5 (0.1)
27.6 (0.7)
20.9 (0.7)
White (non-Hispanic)
76.7 (0.2)
86.1 (0.6)
64.1 (0.8)
Black (non-Hispanic)
11.1 (0.1)
5.4 (0.3)
15.3 (0.5)
Hispanic
6.9 (0.1)
4.7 (0.4)
12.5 (0.6)
Other
5.4 (0.1)
3.8 (0.3)
8.0 (0.4)
Race/ethnicity
Education \High school High school
9.9 (0.1) 27.9 (0.2)
4.9 (0.4) 15.9 (0.6)
13.0 (0.6) 27.1 (0.7)
Some college
26.9 (0.2)
19.6 (0.6)
22.9 (0.7)
C4 years of college
35.3 (0.2)
59.6 (0.8)
37.0 (0.8)
Married
66.5 (0.2)
79.5 (0.7)
68.7 (0.8)
Not married
33.5 (0.2)
20.5 (0.7)
31.3 (0.8)
Under weight
12.3 (0.1)
13.5 (0.5)
13.2 (0.6)
Normal weight
52.6 (0.2)
64.5 (0.8)
55.0 (0.8)
Overweight
12.8 (0.1)
10.3 (0.5)
13.4 (0.6)
Obese
22.2 (0.2)
11.7 (0.5)
18.5 (0.6)
29.2 (0.2)
26.1 (0.7)
27.0 (0.7)
70.8 (0.2)
73.9 (0.7)
73.0 (0.7)
0
46.4 (0.2)
44.5 (0.8)
33.2 (0.7)
1
32.2 (0.2)
34.1 (0.7)
36.1 (0.7)
C2
21.3 (0.2)
21.5 (0.6)
30.6 (0.7)
No
83.8 (0.2)
83.6 (0.6)
85.5 (0.6)
Yes
16.2 (0.2)
16.4 (0.6)
14.5 (0.6)
Table 2 continued Life course factors
Cessation % (SE)
Reduction % (SE)
No reduction % (SE)
No
90.0 (0.1)
92.5 (0.4)
90.3 (0.5)
Yes
10.0 (0.1)
7.5 (0.4)
9.7 (0.5)
No
97.3 (0.1)
97.7 (0.2)
96.4 (0.3)
Yes
2.7 (0.1)
2.3 (0.2)
3.6 (0.3)
73.9 (0.2)
82.0 (0.6)
74.2 (0.7)
26.1 (0.2)
18.0 (0.6)
25.8 (0.7)
No
72.2 (0.2)
75.9 (0.7)
75.6 (0.7)
Yes
27.8 (0.2)
24.1 (0.7)
24.4 (0.7)
Divorce
Homelessness
Inability to pay bills No Yes Ill family member
Drinking/drug problem of close associate No
84.6 (0.2)
85.8 (0.6)
86.3 (0.6)
Yes
15.4 (0.2)
14.2 (0.6)
13.7 (0.6)
Wanted now
90.0 (0.1)
91.1 (0.5)
88.6 (0.5)
Not wanted now
10.0 (0.1)
8.9 (0.5)
11.4 (0.5)
Unwanted pregnancy
Marital status
Health history Prepregnancy BMI
Prenatal health problems No Yes Previous live births
Health behavior Smoked last trimester
Prenatal care initiation Timely
96.2 (0.1)
96.8 (0.3)
94.8 (0.4)
3.8 (0.1)
3.2 (0.2)
5.2 (0.4)
No
96.0 (0.1)
95.9 (0.3)
93.7 (0.4)
Yes
4.0 (0.1)
4.1 (0.3)
6.3 (0.4)
Untimely Psychosocial Abuse during pregnancy
were conducted between our reference group, that is, those whose alcohol pattern was categorized as reduction, and both the cessation and no reduction groups (Table 2). Those in the no reduction category were significantly more likely than those in the reduction group to be \24 years of age, non-white, have a high school education or less, be non-married, be overweight or obese, or have an obstetrical history of one or more prior live births. In addition, they were more likely to report that they had experienced physical abuse during their pregnancy and had bills they could not pay. Analyses comparing women in the reduction and cessation groups also revealed some significant differences (Table 2). Those who had quit their alcohol use during pregnancy (cessation group) were more likely to be younger than 29 years old, black or Hispanic, not married or obese. They were also more likely to report that they could not pay their bills. Multinomial logistic regression analyses for prenatal alcohol use Cessation versus reduction Multinomial logistic regression was used to identify whether women who reported cessation were distinct from women who reported alcohol reduction. These
123
776
Arch Gynecol Obstet (2014) 289:771–779
Table 3 Multinomial logistic regression of prenatal alcohol use: cessation vs. reduction and no reduction vs. reduction (weighted odds ratios, 95 % CI) Life course factors
Cessation vs. reduction OR (95 % CI)
No reduction vs. reduction OR (95 % CI)
Sociodemographic Age 25–29
Reference
Reference
B19
3.01 (2.29, 3.97)
2.99 (2.17, 4.11)
20–24
1.60 (1.40, 1.82)
1.40 (1.18, 1.65)
30–34
0.55 (0.50, 0.60)
0.74 (0.65, 0.84)
C35
0.39 (.35, 0.43)
0.69 (0.60, 0.79)
Race/ethnicity White (non-Hispanic)
Reference
Reference
Black (non-Hispanic) Hispanic
1.46 (1.28, 1.67) 1.18 (0.98, 1.42)
2.70 (2.30, 3.17) 2.59 (2.10, 3.18)
Other
1.43 (1.19, 1.72)
2.58 (2.09, 3.19)
Education Some college
Reference
Reference
Less than high school
1.13 (0.94, 1.36)
1.73 (1.39, 2.16)
High school
1.16 (1.03, 1.30)
1.37 (1.18, 1.58)
4 years college or more
0.58 (0.53, 0.64)
0.64 (0.57, 0.73)
Married
Reference
Reference
Not married
0.99 (0.89, 1.09)
0.86 (0.76, 0.99)
Normal
Reference
Reference
Underweight
0.99 (0.90, 1.10)
1.09 (0.95, 1.25)
Overweight Obese
1.41 (1.26, 1.57) 1.99 (1.78, 2.21)
1.34 (1.15, 1.55) 1.49 (1.30, 1.70)
No
Reference
Reference
Yes
0.87 (0.81, 0.94)
1.00 (0.90, 1.10)
0
Reference
Reference
1
1.11 (1.03, 1.20)
1.64 (1.47, 1.82)
C2
1.26 (1.14, 1.38)
2.06 (1.81, 2.34)
No
Reference
Reference
Yes
0.56 (0.51, 0.63)
0.53 (0.46, 0.62)
Timely
Reference
Reference
Untimely
0.85 (0.71, 1.02)
1.14 (0.90, 1.43)
No
Reference
Reference
Yes
0.60 (0.50, 0.73)
0.98 (0.78, 1.24)
Table 3 continued Life course factors
Cessation vs. reduction OR (95 % CI)
No reduction vs. reduction OR (95 % CI)
No
Reference
Reference
Yes
0.96 (0.83, 1.11)
0.87 (0.72, 1.04)
Reference
Reference
0.73 (0.57, 0.93)
0.90 (0.66, 1.21)
No
Reference
Reference
Yes
1.11 (1.01, 1.23)
1.13 (1.01, 1.30)
No
Reference
Reference
Yes
1.16 (1.03, 1.21)
0.98 (0.88, 1.09)
Divorce
Homeless No Yes Could not pay bills
Family member ill
Someone close has drinking/drug problems No
Reference
Reference
Yes
0.77 (0.69, 0.85)
0.70 (0.61, 0.81)
Pregnancy wanted Wanted now
Reference
Reference
Not wanted now
0.97 (0.85, 1.10)
0.86 (0.73, 1.01)
Marital status
Health history Prepregnancy BMI
Prenatal health problems
Previous live births
Health behavior Smoked last trimester
findings are shown in Table 3. Compared to women whose pattern of alcohol use was classified as reduction, those in the cessation group were more likely to be younger than 24 years of age, non-white, be overweight or obese, be multiparas and were more likely to report having a family member who was very sick (OR 1.16; 95 % CI 1.03, 1.21). Further, women who ceased drinking during pregnancy were less likely to have at least 4 years of college education (OR 0.58; 95 % CI 0.53, 0.64), have had a health problem during pregnancy (OR 0.87; 95 % CI 0.81, 0.94), to have smoked in the last trimester (OR 0.56; 95 % CI 0.51, 0.63), reported abuse during pregnancy (OR 0.60; 95 % CI 0.50, 0.73), be homeless (OR 0.73; 95 % CI 0.57, 0.93) or to have had someone close with drinking/drugs problems (OR 0.77; 95 % CI 0.69, 0.85) compared to those who reduced prenatal alcohol consumption. No reduction versus reduction
Prenatal care initiation
Psychosocial Abuse during pregnancy
123
In comparison with those in the reduction group, those in the no reduction group were more likely to be younger than 24 years of age, more likely to be black, Hispanic or other, more likely to have less than high school education (OR 1.73; 95 % CI 1.39, 2.16), more likely to be overweight (OR 1.34; 95 % CI 1.15, 1.55) or obese (OR 1.49; 95 % CI 1.30, 1.70), more likely to have experienced previous
Arch Gynecol Obstet (2014) 289:771–779
births, and more likely not to be able to pay bills (OR 1.13; 95 % CI 1.01, 1.30). Further, women who did not reduce prenatal alcohol were less likely to have smoked in the last 3 months of pregnancy (OR 0.53; 95 % CI 0.46, 0.62), and less likely to have someone close with drinking/drug problems (OR 0.70; 95 % CI 0.61, 0.81) than those in the reduction group.
Discussion This study revealed several predictors of alcohol use patterns during pregnancy among women who reported at least weekly alcohol consumption prior to conception. Differentiating predictors of cessation, reduction or no reduction in alcohol use unveiled pattern-specific risk factors that may allow for more targeted preventive interventions in the future. Certain sociodemographic factors such as advanced maternal age and higher education demonstrated an increased risk for reducing rather than quitting alcohol consumption during pregnancy, while Hispanic ethnicity favored quitting rather than reducing alcohol use. These findings extend previous research that has linked these factors with prenatal alcohol use by specifying the pattern as reduction, rather using the broader distinction of use vs. non-use of alcohol during pregnancy [22–24]. In the current study, health history also differentiated patterns of prenatal alcohol use. Women who were overweight or obese prior to pregnancy or were multiparas were unlikely to reduce alcohol use, but significantly more likely to either quit or not reduce alcohol consumption between the time of preconception and the third trimester of pregnancy. While replication of these findings is a necessary next step, a potential explanation is that these two health history factors mirror the cumulative effect of risk or protective factors on a woman’s health trajectory throughout their life course. Further study may reveal whether these factors link with untested long-term characteristics that support cessation or increase the likelihood of no reduction. For women with elevated BMIs or those who were multiparas, the identification of sub-groups who are at higher risk for no reduction in alcohol use has implications for clinical practice and policy. For instance, supportive programs for these women and their offspring may be both more time-intensive and costly than those designed for women who have quit or reduced their alcohol intake during pregnancy. In our study, women with prenatal health problems, such as gestational diabetes, pregnancy-induced hypertension, and preeclampsia were found to be less likely to quit than to reduce alcohol consumption. This is particularly alarming since the risk for poor maternal and child health outcomes are associated with these diagnoses and the continued use of alcohol during pregnancy adds additional
777
risk [25, 26]. While we are not aware of previous research that has examined these prenatal health diagnoses in light of alcohol use, the finding raises questions about what processes may be occurring for women that lead to their continued alcohol use. One potential explanation is that alcohol serves as source of stress relief that enhances coping with difficult life problems. Others have postulated that some women, especially those with a history of both prepregnancy alcohol and tobacco use, may have difficulty formulating strategies for coping with the cessation of multiple substances at once [21]. This explanation fits with our finding that smoking during pregnancy was a negative health behavior associated with greater risk for not quitting but only reducing alcohol consumption. Previous research also has linked tobacco and alcohol use during pregnancy but not specified the pattern as reduction [21, 27, 28]. Three psychosocial factors also significantly favored the risk of reduction rather than cessation of alcohol use during pregnancy: abuse during pregnancy, being homeless, and having someone close to you who has a drug or drinking problem. A systematic review of predictors of drinking alcohol during pregnancy consistently identified the experience of having been abused as a predictor of prenatal alcohol use [28]. The use of alcohol as a coping behavior for dealing with difficult life circumstances also may explain the identified risk for not quitting when pregnant. A limitation of this study is the potential bias associated with self-reports of alcohol use during pregnancy. As previously reported, pregnant women are often reluctant to report the amount of alcohol they consume [1]. It can be a bias for any study examining alcohol consumption that is measured indirectly. However, this threat to validity is lessened by the confidentiality of the PRAMS survey approach when compared to face-to-face clinical assessments. Also, data were collected after the birth of the child rather than during pregnancy, an approach that has been found to result in more truthful and accurate assessments [6]. Other limitations in the present study include selfreported data on several other variables such as prepregnancy BMI. Recall bias might also influence the reporting of certain information such as BMI, alcohol consumption, health problems during pregnancy, and several others. Despite these limitations, our findings indicate that differences exist in alcohol consumption patterns across several sociodemographic and psychosocial factors as well as health status during pregnancy. The importance of distinguishing patterns of alcohol use lies in the fact that effective clinical interventions for those who have quit, reduced, or not reduced their alcohol use are likely substantially different in both approach and health care costs. For instance, strategies of reinforcement and positive feedback during routine obstetric appointments may prove to be most appropriate for women in the cessation pattern, while more
123
778
expensive and time-intensive approaches, such as motivational interviewing and referral for multi-service interventions may be more appropriate for those who have reduced but not quit. Programs for women who have not reduced will need to account for lifelong factors that may require intensive, extensive, and expensive treatment options.
Conclusion Alcohol is a teratogen that affects fetal development with irreversible and lifelong outcomes [1, 2]. Since evidence regarding the safe timing and dose of alcohol during pregnancy has not been determined, the CDC recommends that no alcohol be consumed during pregnancy [29]. Our findings are consistent with prior research in demonstrating that the US is far from reaching the Healthy People 2020 goal of 98.3 % alcohol abstinence during pregnancy. The current study offers new evidence for understanding this public health problem by differentiating prenatal alcohol use patterns and related predictors. Pattern-specific findings regarding the socioeconomic factors of age, education, and race/ethnicity may contribute to the identification of target populations for preventive interventions. Further, findings that link specific alcohol use patterns to either abuse during pregnancy or coping with prenatal health problems, such as gestational diabetes and pregnancy-induced hypertension, point to a need for our health care delivery system to identify and treat these stressful exposures as modifiable risk factors for prenatal alcohol use. By establishing clinical programs and public health policies that empower women’s coping mechanisms during pregnancy, it is anticipated that future patterns of reducing or quitting alcohol use during pregnancy may demonstrate both an improvement in health behaviors and a decrease in poor maternal and child health outcomes. Acknowledgments This work has been funded by ABRMF/The Foundation for Alcohol Research. We would like to thank the PRAMS Working Group and the Centers for Disease Control and Prevention for providing the PRAMS data set for this study. Conflict of interest The authors declare that they have no conflict of interest. The authors have full control of all primary data and that they agree to allow the journal to review their data if requested.
References 1. Ismail S, Buckley S, Budacki R, Jabbar A, Gallicano GI (2010) Screening, diagnosing and prevention of fetal alcohol syndrome: is this syndrome treatable? Dev Neurosci 32:91–100 2. Riley EP, Infante MA, Warren KR (2011) Fetal alcohol spectrum disorders: an overview. Neuropsychol Rev 21:73–80
123
Arch Gynecol Obstet (2014) 289:771–779 3. Bertrand J, Floyd RL, Weber MK (2005) Guidelines for identifying and referring persons with fetal alcohol syndrome. MMWR Recomm Rep 54:1–14 4. Sokol RJ, Delaney-Black V, Nordstrom B (2003) Fetal alcohol spectrum disorder. JAMA 290:2996–2999 5. Sayal K, Heron J, Golding J, Emond A (2007) Prenatal alcohol exposure and gender differences in childhood mental health problems: a longitudinal population-based study. Pediatrics 119:e426–e434 6. May PA, Gossage JP, Kalberg WO, Robinson LK, Buckley D, Manning M et al (2009) Prevalence and epidemiologic characteristics of FASD from various research methods with an emphasis on recent in-school studies. Dev Disabil Res Rev 15:176–192 7. Centers for Disease Control and Prevention (CDC) (2002) Fetal alcohol syndrome-Alaska, Arizona, Colorado, New York, 1995–1997. MMWR Recomm Rep 51:433–435 8. Olson HC, Ohlemiller MM, O’Connor MJ, Brown CW, Morris CA, Damus K (2009) National task force on fetal alcohol syndrome and fetal alcohol effect. A call to action: advancing essential services and research on fetal alcohol spectrum disorders: a report of the National Task Force on fetal alcohol syndrome and fetal alcohol effect 9. United States Department of Health and Human Services (DHHS) (2005) U.S. Surgeon General releases advisory on alcohol use in pregnancy: Urges women who are pregnant or who may become pregnant to abstain from alcohol. Available at http://www.surgeongeneral.gov/pressreleases/sg02222005.html. Accessed 30 July 2011 10. Substance abuse and Mental Health Services Administration, office of applied studies (2008) The NSDUH Report: alcohol use among pregnant women and recent mothers: 2002 to 2007. Rockville, MD 11. United States Department of Health and Human Services (2010) Healthy people 2020. Washington, DC 12. Alvik A, Heyerdahl S, Haldorsen T, Lindemann R (2006) Alcohol use before and during pregnancy: a population-based study. Acta Obstet Gynecol Scand 85:1292–1298 13. Harrison PA, Sidebottom AC (2009) Alcohol and drug use before and during pregnancy: an examination of use patterns and predictors of cessation. Maternal Child Health J 13:386–394 14. Fotham E, Ali R, White J, Robinson J (2008) Pregnancy-related changes in tobacco, alcohol and cannabis use reported by antenatal patients at two public hospitals in South Australia. Aust NZ J Obstet Gynaecol 48:248–254 15. Leonardson GR, Loudenburg R, Struck J (2007) Factors predictive of alcohol use during pregnancy in three rural states. Behav Brain Funct 3:8 16. Palma S, Prado-Crespo R, Mariscal M, Perez-Iglesias R, Llorca J, Delgado-Rodriguez M (2007) Weekday but not weekend alcohol consumption before pregnancy influences alcohol cessation during pregnancy. Eur J Public Health 17:394–399 17. Cheng D, Kettinger L, Uduhiri K, Hurt L (2011) Alcohol consumption during pregnancy: prevalence and provider assessment. Obstet Gynecol 117:212–217 18. McLeod D, Pullon S, Cookson T, Cornford E (2002) Factors influencing alcohol consumption during pregnancy and after giving birth. NZ Med J 115:U29 19. O’Connor MJ, Whaley SE (2003) Alcohol use in pregnant lowincome women. J Stud Alcohol 64:773–783 20. Meschke LL, Hellerstedt W, Holl JA, Messelt S (2008) Correlates of prenatal alcohol use. Maternal Child Health J 12:442–451 21. Pirie PL, Lando H, Curry SJ, McBride CM, Grothaus LC (2000) Tobacco, alcohol, and caffeine use and cessation in early pregnancy. Am J Prev Med 18:54–61
Arch Gynecol Obstet (2014) 289:771–779 22. Institute of Medicine (1991) Nutrition during lactation: report of the subcommittee on nutrition during lactation. Committee on nutritional status during pregnancy and lactation: Food and Nutrition Board. Washington DC, National Academy Press 23. Centers for Disease Control and Prevention (CDC) (2012) Alcohol use and binge drinking among women of childbearing age—United States, 2006–2010. MMWR Recomm Rep 61:534–538 24. Harrison PA, Sidebottom AC (2008) Alcohol and drug use before and during pregnancy: an examination of use patterns and predictors of cessation. Matern Child Health J 13:386–394 25. Rosenberg TJ, Garbers S, Lipkind H, Chiasson MA (2005) Maternal obesity and diabetes as risk factors for adverse pregnancy outcomes: differences among 4 racial/ethnic groups. Am J Public Health 95:1545–1551
779 26. Tanaka M, Jaamaa G, Kaiser M, Hillis E, Soim A, Zhu M, Shcherbtykh I, Samelson R, Bell E, Zdeb M, McNuff L-A (2005) Racial disparity in hypertensive disorders of pregnancy in New York state: a 10-year longitudinal population-based study. Am J Public Health 97:163–170 27. Ethen MK, Ramadhani TA, Scheuerle AE, Canfield AM, Wyszynski DF, Druschel AE et al (2009) Alcohol consumption by women before and during pregnancy. Matern Child Health J 13:274–285 28. Skagerstrom J, Chang G, Nilsen P (2011) Predictors of drinking during pregnancy: a systematic review. J Womens Health 20:901–913 29. Centers for disease control and prevention (2009) National Center on birth defects and developmental disabilities. Available at http://www.cdc.gov/ncbddd/index.html
123
MATERNAL-FETAL MEDICINE
Determinants of alcohol cessation, reduction and no reduction during pregnancy Panagiota Kitsantas • Kathleen F. Gaffney Huichuan Wu • Jennifer C. Kastello
•
Received: 20 July 2013 / Accepted: 9 October 2013 / Published online: 23 October 2013 Ó Springer-Verlag Berlin Heidelberg 2013
Abstract Purpose Despite public health initiatives targeting the harmful effects of alcohol exposure on fetal growth, 12 % of pregnant women report current alcohol use. For women who reported drinking alcohol prior to pregnancy, we examined several factors as predictors of three alcohol use patterns during the third trimester of pregnancy: cessation, reduction and no reduction. Methods Using the 2002–2009 Pregnancy Risk Assessment Monitoring System (PRAMS) dataset (311,428 records), a multinomial logistic regression model was constructed to compare alcohol risk by category: (1) cessation vs. reduction (2) no reduction vs. reduction. Results In this sample, 49.4 % drank alcohol before pregnancy. Among those who drank before pregnancy, *87 % quit drinking during pregnancy, 6.6 % reduced, and about 6.4 % reported no reduction. Older women and those with higher education were more likely to reduce than quit their alcohol use. Conversely, women who were black or Hispanic, overweight, obese, or multiparas were more likely to quit than to reduce their prenatal alcohol consumption. Several stressors such as abuse during pregnancy increased their risk of not quitting or not reducing alcohol during the last trimester of pregnancy. Conclusions Differentiating prenatal alcohol use patterns can inform the design of targeted interventions and public P. Kitsantas (&) H. Wu Department of Health Administration and Policy, MS 1J3, College of Health and Human Services, George Mason University, 4400 University Drive, Fairfax, VA 22030, USA e-mail: [email protected] K. F. Gaffney J. C. Kastello School of Nursing, College of Health and Human Services, George Mason University, Fairfax, VA 22030, USA
health policies to meet the Healthy People 2020 objective for achieving a national rate of 98.3 % alcohol abstinence during pregnancy. Keywords Alcohol consumption Pregnancy Alcohol reduction Cessation
Introduction Alcohol use during pregnancy can lead to fetal alcohol spectrum disorders (FASD), a wide range of adverse effects that cause lifelong physical, neurological, and psychological impairments [1–4]. Even low levels of gestational alcohol exposure have been associated with persistent behavioral and emotional problems among affected children [5]. Prevalence rates of FASD in the US and Western Europe have been estimated as high as 2–5 % among young school children [6]. In addition to quality of life issues, the financial impact of prenatal alcohol consumption is high with health care costs estimated to be over $4 billion annually [7, 8]. Since 1981, US Surgeons General have advised that pregnant women and those who anticipate a pregnancy avoid drinking alcohol [9]. Despite longstanding public health initiatives, state and national surveys demonstrate a substantial proportion of women who continue to report alcohol consumption during pregnancy: 15.8 % ages 15–17, 9.8 % ages 18–25, and 12.5 % ages 26–44 years [10]. These rates are significant in light of the Healthy People 2020 national goal for achieving a rate of 98.3 % for alcohol abstinence during pregnancy by the end of this decade [11]. Among the challenges for meeting this goal is the difficulty of correctly identifying and treating women most at risk for prenatal alcohol use. Prior research has examined
123
772
factors that may inform clinical screening and intervention efforts, but findings have been inconsistent and sometimes contradictory. For instance, higher maternal age has been reported as a predictor in some studies, but no statistically significant association has been found in others [12–15]. In separate studies, women with higher maternal education were more likely to drink alcohol during pregnancy, while in other studies those with lower levels of education were more likely to do so [16, 17]. Number of prior pregnancies has been a significant correlate in some research, while others have found no association [15, 18, 19]. Smoking and exposure to abuse have been reported consistently to increase risk [13, 14, 20]. Another factor that has been consistently associated with alcohol use during pregnancy is drinking alcohol prior to pregnancy recognition [12, 13, 19]. Among studies that have examined prepregnancy alcohol use, however, we know of none that has examined predictors of subsequent patterns of prenatal consumption (e.g., cessation, reduction, no reduction). Unveiling differentiated risk profiles may inform the inconsistent findings of prior research and provide evidence for more targeted, sub-group interventions to improve health outcomes for women and infants who experience alcohol-exposed pregnancies. To provide empirical evidence in support of future clinical interventions and public health initiatives that reduce the prevalence of prenatal alcohol use, it is essential to differentiate groups of women based on patterns of change in alcohol consumption between preconception and pregnancy. The aim of this study was to examine several factors as predictors of three alcohol use patterns (cessation, reduction, no reduction) between preconception and the third trimester of pregnancy for women who report drinking alcohol prior to pregnancy.
Methods Sample PRAMS is coordinated by the Centers for Disease Control and Prevention (CDC) in conjunction with state health departments and is designed to gather data on maternal attitudes, behaviors, and experiences during and around the time of pregnancy. The PRAMS data set allows CDC and the states to monitor changes in maternal and child health indicators, and measure progress towards improving the health of mothers and their children. The PRAMS sample is based on all women in the US who had a live birth. Therefore, findings from these data can be generalized for the population of women with a recent live birth. The PRAMS sample selection is based on the annual birth certificate files of participating states and involves stratified
123
Arch Gynecol Obstet (2014) 289:771–779
selection of 1,300–3,400 cases from each participating state from the population of mothers who experienced a recent live birth. Surveys are distributed by mail *2–6 months following delivery. PRAMS questionnaire responses are linked with birth certificate data to allow for simultaneous examination of demographic and medical information collected through state vital records systems. To ensure that high risk populations are adequately represented, some groups of mothers in the PRAMS dataset are sampled at a higher rate. This discrepancy in sampling is adjusted using appropriate sampling weights. Further information about PRAMS design and methods is available at http://www.cdc.gov/prams. We used PRAMS data from the 37 participating states for the years 2002–2009. The sample consisted of 362,752 singleton births. From these records, 311,428 cases with singleton births and with no missing data on the study variables comprised the analytic sample. The distribution of sociodemographic characteristics between those cases that were included in the analytic sample and those that were excluded due to missing data was similar. This study has been approved by the Institutional Review Board at George Mason University. CDC approved this study and provided the data set for the analyses. Outcome variable Categorization of the pattern of alcohol consumption was based on responses to two items in the PRAMS survey: (1) During the 3 months before you got pregnant, how many alcoholic drinks did you have in an average week? and (2) During the last 3 months of your pregnancy, how many alcoholic drinks did you have in an average week?. To reduce the risk of recall bias that has been associated with self-reports of the specific number of drinks consumed, we used an approach from previous research that categorized this outcome variable [21]. The category used for women who reported no alcohol consumption before or during pregnancy was abstinence. The remaining respondents were women who reported at least weekly preconceptual alcohol consumption. Based on women’s responses to the two alcohol-related PRAMS items, their pattern of alcohol consumption between the preconception (3 months before pregnancy) and prenatal timeframes (last 3 months of pregnancy) was categorized as: cessation (consumed alcohol weekly before pregnancy, but not during pregnancy), reduction (continued the consumption of alcohol during the last 3 months of pregnancy at a lower amount compared to the last 3 months before pregnancy), or no reduction (continued alcohol consumption the same or more during the last 3 months of pregnancy compared to the preconceptual period). Potential predictors were selected from several categories,
Arch Gynecol Obstet (2014) 289:771–779
including sociodemographic, health behavior and psychosocial factors.
773
history,
health
Sociodemographic factors Women self-reported their age, race/ethnicity, education, and marital status. Race/ethnicity consisted of non-Hispanic black, non-Hispanic white, Hispanic of any race, and other which included Asian/Pacific Islander, and American Indian/Alaskan. Maternal education was classified as less than high school (B11 years), high school (12 years), some college (13–15 years), and 4 years of college or more (C16 years). The mother’s age (B19, 20–24, 25–29, 30–34, C35), and marital status (married, other) were extracted from the linked child birth certificate. Health history factors Self-reported prepregnancy height and weight were used to calculate body mass index (BMI). Mothers were classified as underweight (\19.8), normal weight (19.8–26.0), overweight ([26.0–29.0) and obese ([29.0) based on the Institute of Medicine BMI criteria [22]. Prenatal health problems (yes, no) were determined based on responses to a list of clinical experiences women may have encountered during pregnancy including gestational diabetes, pregnancy-induced hypertension, preeclampsia, toxemia and premature rupture of membranes. Previous live births were classified as none, 1, or C2. Health behavior factors Smoking during the last 3 months of pregnancy was a dichotomous variable (yes, no). Women were classified as having timely prenatal care if they initiated prenatal care during the first trimester and untimely if they began prenatal care after the first trimester or received no prenatal care. Psychosocial factors To assess for abuse during pregnancy, the PRAMS survey asked women to signify whether they had been ‘‘physically hurt in any way by your husband or partner’’ during their pregnancy (yes, no). To assess for other psychosocial factors, respondents were provided with a list of potential untoward events and asked to indicate (yes, no) whether they had experienced any during the 12 months before their baby was born. Examples included: ‘‘A close family member was very sick and had to go into the hospital’’ or ‘‘I got separated or divorced from my husband or partner’’, ‘‘I was homeless’’, ‘‘Someone close to me had a bad problem with drinking or drugs’’, and ‘‘I had a lot of bills I
couldn’t pay’’. Whether the current pregnancy was wanted or not was assessed based on an item that asked respondents to characterize how they felt about becoming pregnant during the immediate time just before the current pregnancy (not wanted now, wanted). Statistical analysis Bivariate analyses were conducted to compare alcohol use or nonuse during the third trimester of pregnancy by several factors (Table 1). Specifically, in this combined sample (311, 428), we first examined the prevalence of alcohol use before and during pregnancy as well as associations between alcohol consumption status during the last 3 months of pregnancy and several sociodemographic, behavioral, health and psychosocial related factors. Subsequently, alcohol use was further divided into categories of cessation, reduction, or no reduction to unveil differences in risk factors that may exist within these more discrete groups (Table 2). We used the Chi-square statistic to identify significant associations. Multinomial logistic regression was conducted to determine correlates of prenatal alcohol cessation and no reduction relative to alcohol reduction. These analyses were important for identifying which of the predictor variables we considered in this study favor alcohol cessation during pregnancy and which ones increase the risk of no reduction in alcohol consumption among mothers who drank alcohol prior to pregnancy (Table 3). Odds ratios (95 % confidence intervals) were reported to establish the magnitude and direction of any significant effects. In the multinomial logistic regression, reduction was the reference group, cessation was coded as one, and no reduction was coded as two. All analyses were weighted to account for the complex survey design of PRAMS.
Results In this sample, 10.6 % of the women were 19 years old or younger, 75.7 % were between the ages of 20 and 34 years old and 13.7 % were 35 years old or older. Approximately 62 % were white non-Hispanic, 16 % were black nonHispanic and 13.3 % were Hispanic women. About 63 % were married; 18.8 % had less than high school education and 28.1 % had at least 16 years of education. Weighted analyses revealed that approximately half (49.4 %, unweighted n = 152,203) of the respondents reported that they consumed alcoholic beverages in a typical week during the 3 months prior to pregnancy. For the majority of this portion of the sample (unweighted n = 152,203), the pattern of change in alcohol consumption during the last trimester of pregnancy was categorized
123
774
Arch Gynecol Obstet (2014) 289:771–779
Table 1 Prenatal alcohol use for woman who drank alcohol prior to pregnancy (weighted estimates) Life course factors
No alcohol use during third trimester % (SE)
Table 1 continued Life course factors
No alcohol use during third trimester % (SE)
Some alcohol use during third trimester % (SE)
No
90.9 (0.1)
91.4 (0.3)
Yes
9.1 (0.1)
8.6 (0.3)
96.6 (0.1)
97.1 (0.2)
3.4 (0.1)
2.9 (0.2)
No
75.6 (0.1)
78.1 (0.5)
Yes
24.4 (0.1)
21.9 (0.5)
Some alcohol use during third trimester % (SE) Divorce
Sociodemographic Age B19
10.1 (0.1)
3.8 (0.2)
20–24
25.1 (0.1)
14.9 (0.4)
25–29
28.6 (0.1)
24.1 (0.5)
30–34
23.2 (0.1)
32.9 (0.5)
C35
13.1 (0.1)
24.3 (0.5)
64.6 (0.1)
75.3 (0.3)
Black (non-Hispanic)
16.0 (0.1)
10.3 (0.3)
No
74.5 (0.1)
75.7 (0.5)
Hispanic
10.8 (0.1)
8.5 (0.4)
Yes
25.5 (0.1)
24.3 (0.5)
8.6 (0.1)
5.9 (0.3)
Less than high school
16.0 (0.1)
8.9 (0.3)
High school
30.6 (0.1)
21.4 (0.5)
Some college
24.5 (0.1)
21.3 (0.5)
29.0 (0.1)
48.4 (0.6)
Married
64.5 (0.1)
74.1 (0.5)
Other
35.5 (0.1)
25.9 (0.5)
Under weight
13.0 (0.1)
13.3 (0.4)
Normal weight
50.8 (0.2)
59.9 (0.6)
Overweight
13.9 (0.1)
11.8 (0.4)
Obese
23.2 (0.1)
15.0 (0.4)
Bivariate analyses Table 1 displays findings related to the bivariate analyses which show associations between selected factors and prenatal alcohol use. The sociodemographic factors of being over 30 years of age, non-Hispanic white, having attained higher education, and being married were associated with an increased likelihood of alcohol consumption during pregnancy. Health status related factors such as having a normal BMI, experiencing health problems during pregnancy and an obstetrical history of one prior live birth were associated with prenatal alcohol use. Smoking during pregnancy was a negative health behavior linked with alcohol use. Psychosocial factors associated with alcohol consumption were the experiences of abuse during pregnancy and having someone close to you who has drinking and/or drug problems. To further determine whether these factors may be differentially associated with specific patterns of alcohol use during the last trimester of pregnancy, Chi-square analyses
Race/ethnicity White (non-Hispanic)
Other Education
4 years college or more Marital status
Health status Prepregnancy BMI
Homeless No Yes Could not pay bills
Family member ill
Someone close has drinking/drug problems No
87.3 (0.1)
86.0 (0.4)
Yes
12.7 (0.1)
14.0 (0.4)
Wanted
89.8 (0.1)
89.9 (0.3)
Did not want
10.2 (0.1)
10.1 (0.3)
Current pregnancy
as cessation (87 %, unweighted n = 133,861). However, a portion of the women in this study who were drinking before pregnancy indicated persistent alcohol use (6.6 % reduction, unweighted n = 8,909; 6.4 % no reduction, unweighted n = 9,433) while pregnant.
Health problems during pregnancy No
29.9 (0.1)
26.6 (0.5)
Yes
70.1 (0.1)
73.4 (0.5)
None
41.8 (0.1)
39.0 (0.5)
One C2
32.1 (0.1) 26.0 (0.1)
35.0 (0.5) 26.0 (0.5)
No
87.1 (0.1)
84.5 (0.4)
Yes
12.9 (0.1)
15.5 (0.4)
95.4 (0.1)
95.8 (0.2)
4.6 (0.1)
4.2 (0.2)
No
96.2 (0.1)
94.8 (0.3)
Yes
3.8 (0.1)
5.2 (0.3)
Previous live births
Health behaviors Smoked last trimester
Prenatal care initiation Timely (1st trimester) Untimely (2nd, 3rd, none) Psychosocial Abuse during pregnancy
123
Arch Gynecol Obstet (2014) 289:771–779
775
Table 2 Patterns of alcohol use during the last trimester of pregnancy by several factors for women who drank alcohol prior to pregnancy (weighted estimates) Life course factors
Cessation % (SE)
Reduction % (SE)
No reduction % (SE)
Sociodemographic Age B19
6.6 (0.1)
1.7 (0.2)
5.9 (0.4)
20–24
24.8 (0.2)
10.8 (0.5)
19.1 (0.7)
25–29
30.1 (0.2)
22.7 (0.7)
25.5 (0.1)
30–34
24.9 (0.2)
37.2 (0.8)
28.6 (0.7)
C35
13.5 (0.1)
27.6 (0.7)
20.9 (0.7)
White (non-Hispanic)
76.7 (0.2)
86.1 (0.6)
64.1 (0.8)
Black (non-Hispanic)
11.1 (0.1)
5.4 (0.3)
15.3 (0.5)
Hispanic
6.9 (0.1)
4.7 (0.4)
12.5 (0.6)
Other
5.4 (0.1)
3.8 (0.3)
8.0 (0.4)
Race/ethnicity
Education \High school High school
9.9 (0.1) 27.9 (0.2)
4.9 (0.4) 15.9 (0.6)
13.0 (0.6) 27.1 (0.7)
Some college
26.9 (0.2)
19.6 (0.6)
22.9 (0.7)
C4 years of college
35.3 (0.2)
59.6 (0.8)
37.0 (0.8)
Married
66.5 (0.2)
79.5 (0.7)
68.7 (0.8)
Not married
33.5 (0.2)
20.5 (0.7)
31.3 (0.8)
Under weight
12.3 (0.1)
13.5 (0.5)
13.2 (0.6)
Normal weight
52.6 (0.2)
64.5 (0.8)
55.0 (0.8)
Overweight
12.8 (0.1)
10.3 (0.5)
13.4 (0.6)
Obese
22.2 (0.2)
11.7 (0.5)
18.5 (0.6)
29.2 (0.2)
26.1 (0.7)
27.0 (0.7)
70.8 (0.2)
73.9 (0.7)
73.0 (0.7)
0
46.4 (0.2)
44.5 (0.8)
33.2 (0.7)
1
32.2 (0.2)
34.1 (0.7)
36.1 (0.7)
C2
21.3 (0.2)
21.5 (0.6)
30.6 (0.7)
No
83.8 (0.2)
83.6 (0.6)
85.5 (0.6)
Yes
16.2 (0.2)
16.4 (0.6)
14.5 (0.6)
Table 2 continued Life course factors
Cessation % (SE)
Reduction % (SE)
No reduction % (SE)
No
90.0 (0.1)
92.5 (0.4)
90.3 (0.5)
Yes
10.0 (0.1)
7.5 (0.4)
9.7 (0.5)
No
97.3 (0.1)
97.7 (0.2)
96.4 (0.3)
Yes
2.7 (0.1)
2.3 (0.2)
3.6 (0.3)
73.9 (0.2)
82.0 (0.6)
74.2 (0.7)
26.1 (0.2)
18.0 (0.6)
25.8 (0.7)
No
72.2 (0.2)
75.9 (0.7)
75.6 (0.7)
Yes
27.8 (0.2)
24.1 (0.7)
24.4 (0.7)
Divorce
Homelessness
Inability to pay bills No Yes Ill family member
Drinking/drug problem of close associate No
84.6 (0.2)
85.8 (0.6)
86.3 (0.6)
Yes
15.4 (0.2)
14.2 (0.6)
13.7 (0.6)
Wanted now
90.0 (0.1)
91.1 (0.5)
88.6 (0.5)
Not wanted now
10.0 (0.1)
8.9 (0.5)
11.4 (0.5)
Unwanted pregnancy
Marital status
Health history Prepregnancy BMI
Prenatal health problems No Yes Previous live births
Health behavior Smoked last trimester
Prenatal care initiation Timely
96.2 (0.1)
96.8 (0.3)
94.8 (0.4)
3.8 (0.1)
3.2 (0.2)
5.2 (0.4)
No
96.0 (0.1)
95.9 (0.3)
93.7 (0.4)
Yes
4.0 (0.1)
4.1 (0.3)
6.3 (0.4)
Untimely Psychosocial Abuse during pregnancy
were conducted between our reference group, that is, those whose alcohol pattern was categorized as reduction, and both the cessation and no reduction groups (Table 2). Those in the no reduction category were significantly more likely than those in the reduction group to be \24 years of age, non-white, have a high school education or less, be non-married, be overweight or obese, or have an obstetrical history of one or more prior live births. In addition, they were more likely to report that they had experienced physical abuse during their pregnancy and had bills they could not pay. Analyses comparing women in the reduction and cessation groups also revealed some significant differences (Table 2). Those who had quit their alcohol use during pregnancy (cessation group) were more likely to be younger than 29 years old, black or Hispanic, not married or obese. They were also more likely to report that they could not pay their bills. Multinomial logistic regression analyses for prenatal alcohol use Cessation versus reduction Multinomial logistic regression was used to identify whether women who reported cessation were distinct from women who reported alcohol reduction. These
123
776
Arch Gynecol Obstet (2014) 289:771–779
Table 3 Multinomial logistic regression of prenatal alcohol use: cessation vs. reduction and no reduction vs. reduction (weighted odds ratios, 95 % CI) Life course factors
Cessation vs. reduction OR (95 % CI)
No reduction vs. reduction OR (95 % CI)
Sociodemographic Age 25–29
Reference
Reference
B19
3.01 (2.29, 3.97)
2.99 (2.17, 4.11)
20–24
1.60 (1.40, 1.82)
1.40 (1.18, 1.65)
30–34
0.55 (0.50, 0.60)
0.74 (0.65, 0.84)
C35
0.39 (.35, 0.43)
0.69 (0.60, 0.79)
Race/ethnicity White (non-Hispanic)
Reference
Reference
Black (non-Hispanic) Hispanic
1.46 (1.28, 1.67) 1.18 (0.98, 1.42)
2.70 (2.30, 3.17) 2.59 (2.10, 3.18)
Other
1.43 (1.19, 1.72)
2.58 (2.09, 3.19)
Education Some college
Reference
Reference
Less than high school
1.13 (0.94, 1.36)
1.73 (1.39, 2.16)
High school
1.16 (1.03, 1.30)
1.37 (1.18, 1.58)
4 years college or more
0.58 (0.53, 0.64)
0.64 (0.57, 0.73)
Married
Reference
Reference
Not married
0.99 (0.89, 1.09)
0.86 (0.76, 0.99)
Normal
Reference
Reference
Underweight
0.99 (0.90, 1.10)
1.09 (0.95, 1.25)
Overweight Obese
1.41 (1.26, 1.57) 1.99 (1.78, 2.21)
1.34 (1.15, 1.55) 1.49 (1.30, 1.70)
No
Reference
Reference
Yes
0.87 (0.81, 0.94)
1.00 (0.90, 1.10)
0
Reference
Reference
1
1.11 (1.03, 1.20)
1.64 (1.47, 1.82)
C2
1.26 (1.14, 1.38)
2.06 (1.81, 2.34)
No
Reference
Reference
Yes
0.56 (0.51, 0.63)
0.53 (0.46, 0.62)
Timely
Reference
Reference
Untimely
0.85 (0.71, 1.02)
1.14 (0.90, 1.43)
No
Reference
Reference
Yes
0.60 (0.50, 0.73)
0.98 (0.78, 1.24)
Table 3 continued Life course factors
Cessation vs. reduction OR (95 % CI)
No reduction vs. reduction OR (95 % CI)
No
Reference
Reference
Yes
0.96 (0.83, 1.11)
0.87 (0.72, 1.04)
Reference
Reference
0.73 (0.57, 0.93)
0.90 (0.66, 1.21)
No
Reference
Reference
Yes
1.11 (1.01, 1.23)
1.13 (1.01, 1.30)
No
Reference
Reference
Yes
1.16 (1.03, 1.21)
0.98 (0.88, 1.09)
Divorce
Homeless No Yes Could not pay bills
Family member ill
Someone close has drinking/drug problems No
Reference
Reference
Yes
0.77 (0.69, 0.85)
0.70 (0.61, 0.81)
Pregnancy wanted Wanted now
Reference
Reference
Not wanted now
0.97 (0.85, 1.10)
0.86 (0.73, 1.01)
Marital status
Health history Prepregnancy BMI
Prenatal health problems
Previous live births
Health behavior Smoked last trimester
findings are shown in Table 3. Compared to women whose pattern of alcohol use was classified as reduction, those in the cessation group were more likely to be younger than 24 years of age, non-white, be overweight or obese, be multiparas and were more likely to report having a family member who was very sick (OR 1.16; 95 % CI 1.03, 1.21). Further, women who ceased drinking during pregnancy were less likely to have at least 4 years of college education (OR 0.58; 95 % CI 0.53, 0.64), have had a health problem during pregnancy (OR 0.87; 95 % CI 0.81, 0.94), to have smoked in the last trimester (OR 0.56; 95 % CI 0.51, 0.63), reported abuse during pregnancy (OR 0.60; 95 % CI 0.50, 0.73), be homeless (OR 0.73; 95 % CI 0.57, 0.93) or to have had someone close with drinking/drugs problems (OR 0.77; 95 % CI 0.69, 0.85) compared to those who reduced prenatal alcohol consumption. No reduction versus reduction
Prenatal care initiation
Psychosocial Abuse during pregnancy
123
In comparison with those in the reduction group, those in the no reduction group were more likely to be younger than 24 years of age, more likely to be black, Hispanic or other, more likely to have less than high school education (OR 1.73; 95 % CI 1.39, 2.16), more likely to be overweight (OR 1.34; 95 % CI 1.15, 1.55) or obese (OR 1.49; 95 % CI 1.30, 1.70), more likely to have experienced previous
Arch Gynecol Obstet (2014) 289:771–779
births, and more likely not to be able to pay bills (OR 1.13; 95 % CI 1.01, 1.30). Further, women who did not reduce prenatal alcohol were less likely to have smoked in the last 3 months of pregnancy (OR 0.53; 95 % CI 0.46, 0.62), and less likely to have someone close with drinking/drug problems (OR 0.70; 95 % CI 0.61, 0.81) than those in the reduction group.
Discussion This study revealed several predictors of alcohol use patterns during pregnancy among women who reported at least weekly alcohol consumption prior to conception. Differentiating predictors of cessation, reduction or no reduction in alcohol use unveiled pattern-specific risk factors that may allow for more targeted preventive interventions in the future. Certain sociodemographic factors such as advanced maternal age and higher education demonstrated an increased risk for reducing rather than quitting alcohol consumption during pregnancy, while Hispanic ethnicity favored quitting rather than reducing alcohol use. These findings extend previous research that has linked these factors with prenatal alcohol use by specifying the pattern as reduction, rather using the broader distinction of use vs. non-use of alcohol during pregnancy [22–24]. In the current study, health history also differentiated patterns of prenatal alcohol use. Women who were overweight or obese prior to pregnancy or were multiparas were unlikely to reduce alcohol use, but significantly more likely to either quit or not reduce alcohol consumption between the time of preconception and the third trimester of pregnancy. While replication of these findings is a necessary next step, a potential explanation is that these two health history factors mirror the cumulative effect of risk or protective factors on a woman’s health trajectory throughout their life course. Further study may reveal whether these factors link with untested long-term characteristics that support cessation or increase the likelihood of no reduction. For women with elevated BMIs or those who were multiparas, the identification of sub-groups who are at higher risk for no reduction in alcohol use has implications for clinical practice and policy. For instance, supportive programs for these women and their offspring may be both more time-intensive and costly than those designed for women who have quit or reduced their alcohol intake during pregnancy. In our study, women with prenatal health problems, such as gestational diabetes, pregnancy-induced hypertension, and preeclampsia were found to be less likely to quit than to reduce alcohol consumption. This is particularly alarming since the risk for poor maternal and child health outcomes are associated with these diagnoses and the continued use of alcohol during pregnancy adds additional
777
risk [25, 26]. While we are not aware of previous research that has examined these prenatal health diagnoses in light of alcohol use, the finding raises questions about what processes may be occurring for women that lead to their continued alcohol use. One potential explanation is that alcohol serves as source of stress relief that enhances coping with difficult life problems. Others have postulated that some women, especially those with a history of both prepregnancy alcohol and tobacco use, may have difficulty formulating strategies for coping with the cessation of multiple substances at once [21]. This explanation fits with our finding that smoking during pregnancy was a negative health behavior associated with greater risk for not quitting but only reducing alcohol consumption. Previous research also has linked tobacco and alcohol use during pregnancy but not specified the pattern as reduction [21, 27, 28]. Three psychosocial factors also significantly favored the risk of reduction rather than cessation of alcohol use during pregnancy: abuse during pregnancy, being homeless, and having someone close to you who has a drug or drinking problem. A systematic review of predictors of drinking alcohol during pregnancy consistently identified the experience of having been abused as a predictor of prenatal alcohol use [28]. The use of alcohol as a coping behavior for dealing with difficult life circumstances also may explain the identified risk for not quitting when pregnant. A limitation of this study is the potential bias associated with self-reports of alcohol use during pregnancy. As previously reported, pregnant women are often reluctant to report the amount of alcohol they consume [1]. It can be a bias for any study examining alcohol consumption that is measured indirectly. However, this threat to validity is lessened by the confidentiality of the PRAMS survey approach when compared to face-to-face clinical assessments. Also, data were collected after the birth of the child rather than during pregnancy, an approach that has been found to result in more truthful and accurate assessments [6]. Other limitations in the present study include selfreported data on several other variables such as prepregnancy BMI. Recall bias might also influence the reporting of certain information such as BMI, alcohol consumption, health problems during pregnancy, and several others. Despite these limitations, our findings indicate that differences exist in alcohol consumption patterns across several sociodemographic and psychosocial factors as well as health status during pregnancy. The importance of distinguishing patterns of alcohol use lies in the fact that effective clinical interventions for those who have quit, reduced, or not reduced their alcohol use are likely substantially different in both approach and health care costs. For instance, strategies of reinforcement and positive feedback during routine obstetric appointments may prove to be most appropriate for women in the cessation pattern, while more
123
778
expensive and time-intensive approaches, such as motivational interviewing and referral for multi-service interventions may be more appropriate for those who have reduced but not quit. Programs for women who have not reduced will need to account for lifelong factors that may require intensive, extensive, and expensive treatment options.
Conclusion Alcohol is a teratogen that affects fetal development with irreversible and lifelong outcomes [1, 2]. Since evidence regarding the safe timing and dose of alcohol during pregnancy has not been determined, the CDC recommends that no alcohol be consumed during pregnancy [29]. Our findings are consistent with prior research in demonstrating that the US is far from reaching the Healthy People 2020 goal of 98.3 % alcohol abstinence during pregnancy. The current study offers new evidence for understanding this public health problem by differentiating prenatal alcohol use patterns and related predictors. Pattern-specific findings regarding the socioeconomic factors of age, education, and race/ethnicity may contribute to the identification of target populations for preventive interventions. Further, findings that link specific alcohol use patterns to either abuse during pregnancy or coping with prenatal health problems, such as gestational diabetes and pregnancy-induced hypertension, point to a need for our health care delivery system to identify and treat these stressful exposures as modifiable risk factors for prenatal alcohol use. By establishing clinical programs and public health policies that empower women’s coping mechanisms during pregnancy, it is anticipated that future patterns of reducing or quitting alcohol use during pregnancy may demonstrate both an improvement in health behaviors and a decrease in poor maternal and child health outcomes. Acknowledgments This work has been funded by ABRMF/The Foundation for Alcohol Research. We would like to thank the PRAMS Working Group and the Centers for Disease Control and Prevention for providing the PRAMS data set for this study. Conflict of interest The authors declare that they have no conflict of interest. The authors have full control of all primary data and that they agree to allow the journal to review their data if requested.
References 1. Ismail S, Buckley S, Budacki R, Jabbar A, Gallicano GI (2010) Screening, diagnosing and prevention of fetal alcohol syndrome: is this syndrome treatable? Dev Neurosci 32:91–100 2. Riley EP, Infante MA, Warren KR (2011) Fetal alcohol spectrum disorders: an overview. Neuropsychol Rev 21:73–80
123
Arch Gynecol Obstet (2014) 289:771–779 3. Bertrand J, Floyd RL, Weber MK (2005) Guidelines for identifying and referring persons with fetal alcohol syndrome. MMWR Recomm Rep 54:1–14 4. Sokol RJ, Delaney-Black V, Nordstrom B (2003) Fetal alcohol spectrum disorder. JAMA 290:2996–2999 5. Sayal K, Heron J, Golding J, Emond A (2007) Prenatal alcohol exposure and gender differences in childhood mental health problems: a longitudinal population-based study. Pediatrics 119:e426–e434 6. May PA, Gossage JP, Kalberg WO, Robinson LK, Buckley D, Manning M et al (2009) Prevalence and epidemiologic characteristics of FASD from various research methods with an emphasis on recent in-school studies. Dev Disabil Res Rev 15:176–192 7. Centers for Disease Control and Prevention (CDC) (2002) Fetal alcohol syndrome-Alaska, Arizona, Colorado, New York, 1995–1997. MMWR Recomm Rep 51:433–435 8. Olson HC, Ohlemiller MM, O’Connor MJ, Brown CW, Morris CA, Damus K (2009) National task force on fetal alcohol syndrome and fetal alcohol effect. A call to action: advancing essential services and research on fetal alcohol spectrum disorders: a report of the National Task Force on fetal alcohol syndrome and fetal alcohol effect 9. United States Department of Health and Human Services (DHHS) (2005) U.S. Surgeon General releases advisory on alcohol use in pregnancy: Urges women who are pregnant or who may become pregnant to abstain from alcohol. Available at http://www.surgeongeneral.gov/pressreleases/sg02222005.html. Accessed 30 July 2011 10. Substance abuse and Mental Health Services Administration, office of applied studies (2008) The NSDUH Report: alcohol use among pregnant women and recent mothers: 2002 to 2007. Rockville, MD 11. United States Department of Health and Human Services (2010) Healthy people 2020. Washington, DC 12. Alvik A, Heyerdahl S, Haldorsen T, Lindemann R (2006) Alcohol use before and during pregnancy: a population-based study. Acta Obstet Gynecol Scand 85:1292–1298 13. Harrison PA, Sidebottom AC (2009) Alcohol and drug use before and during pregnancy: an examination of use patterns and predictors of cessation. Maternal Child Health J 13:386–394 14. Fotham E, Ali R, White J, Robinson J (2008) Pregnancy-related changes in tobacco, alcohol and cannabis use reported by antenatal patients at two public hospitals in South Australia. Aust NZ J Obstet Gynaecol 48:248–254 15. Leonardson GR, Loudenburg R, Struck J (2007) Factors predictive of alcohol use during pregnancy in three rural states. Behav Brain Funct 3:8 16. Palma S, Prado-Crespo R, Mariscal M, Perez-Iglesias R, Llorca J, Delgado-Rodriguez M (2007) Weekday but not weekend alcohol consumption before pregnancy influences alcohol cessation during pregnancy. Eur J Public Health 17:394–399 17. Cheng D, Kettinger L, Uduhiri K, Hurt L (2011) Alcohol consumption during pregnancy: prevalence and provider assessment. Obstet Gynecol 117:212–217 18. McLeod D, Pullon S, Cookson T, Cornford E (2002) Factors influencing alcohol consumption during pregnancy and after giving birth. NZ Med J 115:U29 19. O’Connor MJ, Whaley SE (2003) Alcohol use in pregnant lowincome women. J Stud Alcohol 64:773–783 20. Meschke LL, Hellerstedt W, Holl JA, Messelt S (2008) Correlates of prenatal alcohol use. Maternal Child Health J 12:442–451 21. Pirie PL, Lando H, Curry SJ, McBride CM, Grothaus LC (2000) Tobacco, alcohol, and caffeine use and cessation in early pregnancy. Am J Prev Med 18:54–61
Arch Gynecol Obstet (2014) 289:771–779 22. Institute of Medicine (1991) Nutrition during lactation: report of the subcommittee on nutrition during lactation. Committee on nutritional status during pregnancy and lactation: Food and Nutrition Board. Washington DC, National Academy Press 23. Centers for Disease Control and Prevention (CDC) (2012) Alcohol use and binge drinking among women of childbearing age—United States, 2006–2010. MMWR Recomm Rep 61:534–538 24. Harrison PA, Sidebottom AC (2008) Alcohol and drug use before and during pregnancy: an examination of use patterns and predictors of cessation. Matern Child Health J 13:386–394 25. Rosenberg TJ, Garbers S, Lipkind H, Chiasson MA (2005) Maternal obesity and diabetes as risk factors for adverse pregnancy outcomes: differences among 4 racial/ethnic groups. Am J Public Health 95:1545–1551
779 26. Tanaka M, Jaamaa G, Kaiser M, Hillis E, Soim A, Zhu M, Shcherbtykh I, Samelson R, Bell E, Zdeb M, McNuff L-A (2005) Racial disparity in hypertensive disorders of pregnancy in New York state: a 10-year longitudinal population-based study. Am J Public Health 97:163–170 27. Ethen MK, Ramadhani TA, Scheuerle AE, Canfield AM, Wyszynski DF, Druschel AE et al (2009) Alcohol consumption by women before and during pregnancy. Matern Child Health J 13:274–285 28. Skagerstrom J, Chang G, Nilsen P (2011) Predictors of drinking during pregnancy: a systematic review. J Womens Health 20:901–913 29. Centers for disease control and prevention (2009) National Center on birth defects and developmental disabilities. Available at http://www.cdc.gov/ncbddd/index.html
123
