A Cross‑Section Study of Relationship Characteristics and Smoking Cessation During Pregnancy in a Sample of Romanian Pregnant Women

Abstract

Introduction

Partner support and relationship characteristics may be important factors in effective couple-based pregnancy smoking cessation programs. Research is needed to investigate the links between couple relationship characteristics and maternal smoking cessation to inform the development of such interventions.

Methods

This paper relies on cross-section data collected during the formative phase in the development of an ongoing couple-focused pregnancy tobacco cessation trial. Data (n = 143 pregnant women) were collected from two large public and one private obstetrics and gynecology clinics located in Cluj-Napoca, Romania. Multiple logistic regression was used to identify correlates of smoking cessation during pregnancy.

Results

Higher education was a significant correlate of smoking cessation during pregnancy. Women with a college degree or more had higher odds to quit smoking during pregnancy in comparison to their counterparts who graduated high school or less (OR 14.3, 95% CI 2.75–74.28). In addition, women with increased positive partner interactions related to their smoking cessation efforts correlated with higher odds of quitting smoking during pregnancy (OR 1.48, 95% CI 1.15–1.91).

Discussion

While pregnancy tobacco cessation interventions with partner support do exist, most were not successful and did not focus on couple-related concepts such as partner interactions, dyadic coping, and dyadic efficacy. The findings of the study are important because they bring new insights regarding the potential role of relationship characteristics to inform future cessation programs focused on pregnant smokers and their life partners.

Introduction

Worldwide, tobacco smoking is the main cause of preventable death. Over one billion people smoke tobacco worldwide and there are over 7 million deaths per year (Zafeiridou et al. 2018). The burden of tobacco smoking is disproportionately high in the low-and middle-income countries (World Health Organization 2019).

Smoking during pregnancy is particularly harmful as it has negative health effects both on the pregnant woman and on the baby, with lifetime implications. Approximately 15%of Romanian women continue smoking during pregnancy with negative effects on maternal, pregnancy, and infant health, emphasizing the need for pregnancy smoking cessation programs (Meghea et al. 2012; Blaga et al. 2017). The vast majority of women are aware of the negative health effects of pregnancy smoking and want a healthy pregnancy, birth, and baby (Meghea et al. 2012). The pregnancy provides additional motivation for quitting smoking (Lange et al. 2018).

Maternal smoking is linked to poor birth outcomes such as stillbirth, miscarriage, low birthweight and sudden infant death syndrome (Fitzpatrick et al. 2016). No level of tobacco smoking during pregnancy is considered to be safe, greater smoking intensity calculated by the number of cigarettes used each day is associated with more severe effects on the fetus compared with lighter smoking (Curtin and Mathews2016).

Younger age, lack of prenatal care, not having a life partner, or having a smoker partner are risk factors for continued smoking during pregnancy (Širvinskienė et al. 2016). Other risk factors include low socioeconomic status, depressive symptoms, and high perceived stress (Lauria et al. 2012). Improved understanding of risk factors for maternal pregnancy smoking can help public health professionals design and target interventions for this vulnerable population of pregnant smokers. Identifying risk factors and subgroups of women who are more likely to smoke during pregnancy is critical (McDonnell and Regan 2019). Tobacco cessation interventions represent an immediate need for smokers anthem pregnancy period, and are critical for providing support and adequate resources (Bérard et al. 2016), especially in low-resource settings such as Romania.

Evidence exists in the general population suggesting that relationship characteristics, including dyadic efficacy, dyadic coping, and partner interactions related to smoking are associated with smoking cessation (Sterba et al. 2011).However, the data were from a high-income country and were not specific to pregnant women, a population in need of tailored tobacco cessation interventions (Diamanti et al.2019). Prenatal and postnatal support, mainly received from the spouse or life partner, is correlated with pregnancy smoking cessation (Meghea et al. 2012; Blaga et al. 2017).While partner and peer support may be important factors influencing smoking behavior, eliciting effective support can be difficult (Chamberlain et al. 2017). Research is needed to investigate the links between couple relationship characteristics and maternal smoking cessation to inform the development of pregnancy tobacco cessation interventions with partner support

In the context of the formative phase of the Quit Together(QT) couple-focused pregnancy smoking cessation intervention (Meghea et al. 2018a), the objective of this paper is to identify predictors of smoking cessation during pregnancy ina sample of Romanian women. The main contribution of this paper is to bring evidence on the link between a variety of relationship characteristics and maternal pregnancy smoking cessation, while accounting for relevant covariates, in an Eastern European middle-income country.

Materials and Methods

This paper relies on cross-sectional data collected from convenience sample during the formative research phase in the development phase of an ongoing couple-focused pregnancy tobacco cessation randomized controlled trial (Meghea et al. 2018a). The study received approval from the Michigan State University Institutional Review Board and the Ethics Commission at Babes-Boylan University and was conducted in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments.

Study Setting

Data were collected at Novogyn, a private clinic of obstetrics and gynecology, and two large public obstetrics and gynecology clinics, Obstetrics and Gynecology Clinic I and II, allocated in Cluj-Napoca, Romania. The three clinics deliver around 6000 births annually, the vast majority of all births in Cluj-Napoca, a metropolitan area of over 420,000.

Participants

Data were collected from 143 pregnant women between April 2016 and January 2017. The study population included smoking (n = 62, 43.4%) and ex-smoking (n = 81, 56.6%)Romanian pregnant women. The study relies on this convenience sample with the sample size driven by the available time and funds to recruit. The eligibility criteria for inclusion in the study were: ≤ 28 weeks of gestation, age over 18, and having a stable life partner. Out of 693 women who went through the eligibility criteria, 143 were included in the research study. This resulted in an eligibility rate of20%. Reasons for exclusion were: being under 18, not being pregnant, not having a stable partner, not being a smoker or an ex-smoker, refusing to sign the informed consent, or refusing the measurement of the level of CO.

Data Collection

The enrollment and data collection took place in the waiting rooms of the three partner clinics and was conducted by two female data collectors trained to approach women, describe the study to them, invite them to participate, screen them against the pre-set eligibility criteria, and obtain their written informed consent. If eligible and interested to participate in the study, women signed two copies of an informed consent before completing the paper-based questionnaire. One copy of the informed consent was kept by each participant in the study. The women who were eligible but refused participation in the study were asked to answer five questions regarding their age, residence, marital status, ethnicity and education, in order to assess any existing differences between women who were willing to enroll and those who refused to further participate in the study.

Measures

The questionnaire consisted of five sections on women’ssocio-demographics, medical and reproductive history, alcohol and smoking consumption, emotional health and quality of relationship with the partner. The questionnaire was developed based on similar questionnaires developed with insight from stakeholders and used in our prior studies in Romania focused on pregnant and postpartum women(Meghea et al. 2012, 2018b). As before, the questionnaire included validated instruments that were adapted and translated in Romanian by the authors (MDD, CIM, OMB are all native Romanian speakers)

Smoking Status

Smoking status (smoker vs ex-smoker) was the outcome in our analyses. Women were asked which of the following statements would best describe their cigarette smoking: ‘1—I smoke every once in a while’, ‘2—I smoke regularly now but I cut down since I found out I was pregnant’, ‘3—I smoke regularly now, about the same amount as before finding out I was pregnant’, ‘4—I have quit smoking since finding out I was pregnant or in the 6 months before becoming pregnant’. The binary outcome was recoded as smoker (1,2, or 3) vs ex-smoker (4)

Relationship Variables

Teamwork standards (Sterba et al. 2011) measured on a scale from 1 to 10, assessed the extent the women considered that the problems related to smoking cessation should be managed alone or with the partner (1 meaning that they should manage them alone, while 10 meaning that they should manage them together as a team).

For dyadic coping, we used a modified and adapted form of the dyadic coping subscale of the BodenmannDyadic Coping instrument to the smoking cessation background (Bodenmann 1997). The variable ‘dyadic coping ‘was measured with a six-item scale. Women were asked to report how they and their partner have coped with the stress related to smoking and smoking cessation (‘My partner showed empathy’, ‘my partner expressed that he is on my side in my smoking cessation efforts’, ‘my partner and have tried to find solutions together regarding smoking cessation’, ‘my partner would engage in a serious discussion about smoking-related problems and think what haste be done by both of us’, ‘I was satisfied with the smoking cessation support I have received from my partner’, ‘I was satisfied with the way we dealt with smoking-related stress together’). The answer options were: never, hardly ever, rarely, sometimes, often, or very often (5). The scale was computed by summing up the responses for all the items. The dyadic coping score variable was measured on a continuous scale, higher scores meaning higher dyadic coping

We have used three items from the Partner Interaction Questionnaire (PIQ) (Cohen and Lichtenstein 1990) focusing on partner positive behaviors and three items focusing on negative behaviors in order to assess partner behavior related to the women’s smoking cessation process. Women were asked to mention how often did their partner engaged in positive or negative activities in relation to their smoking behavior (positive: ‘congratulated you for your decision to quit smoking’, ‘expressed confidence in your ability to quit/remain quit’, and ‘expressed pleasure at your efforts to quit’; and negative: ‘talked you out of smoking cigarette’, ‘commented that smoking is a bad habit’, ‘mentioned being bothered by smoke’). Possible answer alternatives were: never, almost never, sometimes, fairly often, very often. The 5-point Likert scale was recoded from 0(never) to 4 (very often). The negative partner interactions subscale consisted of items 1, 2, and 3, whereas the positive behaviors subscale consisted of items 4, 5, and 6.Separate scores were calculated for negative and positive behaviors by summing up responses on each subscale. The score for positive partner interactions and negative partner interactions were measured from 0 to 12, higher scores meaning more negative/positive partner interactions.

Dyadic efficacy was measured through a 8-item tool (Sterna et al. 2011) answered on a scale from 0 to 100(0 = not at all confident; 100 = very confident) regarding how confident was the participant that she could work together with her partner in different smoking-related situations to: ‘make a decision about medication to help me quit smoking’, ‘focus on the benefits of quitting smoking’, ‘make me feel better when I am feeling down about quitting’, ‘manage the daily smoking-related problems that came up’, ‘deal with ups and downs of trying to quit’, ‘deal with negative emotions I have experienced when trying to quit’ and ‘do what it takes to quit smoking for good’. We averaged the response to each item and generated the overall score, with higher scores reflecting a higher dyadic efficacy score.

Partner smoking was initially categorized in being a smoker, being an ex-smoker or never been a smoker. The variable was recoded in smoker or non-smoker.

Risk Factors

Socio‑demographics

Women were asked to answer a general question regarding their age which was further categorized in three main categories (18–25, > 25, missing). Their highest level of attained education was categorized in high school or less and college or more. Regarding ethnicity, the variable was categorized Romanian vs other (which included other minority ethnicities such as Hungarian, Roma). The marital status was coded binary married or not married but living with a stable partner; the residence was grouped in rural or urban (including per urban/suburban); household income was categorized in less than 1500 RON (315.23€), 1501–5000 RON (315.44€–1,051€), and > 5001 RON (1,051€); indoor smoking was categorized in: not allowed/ allowed; not allowed but with exceptions/ missing.

BMI

The BMI of the pregnant women was calculated based on their height and pre-pregnancy weight and was further classified in three categories: underweight (BMI ≤ 18.5), normal weight (18.5 < BMI ≤ 24.9), pre-obesity; obese (BMI > 25) or missing.

Medical History

We assessed the presence of various chronic conditions such as asthma, diabetes, hypertension, obesity, anemia, thrombophilia, lupus, epilepsy, cancer, or other medical conditions. We generated a binary variable coded 1 (yes) for women who had at least one medical condition and no (0) for women who did not have any medical condition.

Pregnancy History

Prior birth outcomes were categorized into three mutually exclusive categories: prior poor pregnancy outcomes, no prior poor pregnancy outcomes, and first time pregnant. Prior poor pregnancy outcomes included miscarriages, still births, low birth weight (< 2500 g), preterm (< 37 weeks gestation age at birth), baby stayed in hospital after the mother went home, baby weight > 4000 g, and babies born with congenital defects or who died in the first year of life.

Anxiety and Depression Symptoms

Anxiety and depression symptoms were assessed with the validated 4-item scale of the ‘Patient Health Questionnaire-4 (PHQ-4)’ instrument (Kroenke et al. 2009). Women were asked to report on a 4-point Likert scale whether they experienced feeling nervous, anxious or on edge, not being able to stop or control worrying, little interest or pleasure in doing things and feeling down, depressed or hopeless, having the possibility to answer not at all, several days, more than half of the days and nearly every day. The total score was calculated by summing up the scores for each of the four items. The PHQ4 score was further categorized in normal (0–2), mild (3–5) and moderate to severe (6–12) anxiety and depression symptoms. A new variable was created and categorized intro normal, mild + moderate to severe and missing.

Data Analysis

Descriptive analyses were conducted to report the sociodemographic characteristics (age, gender, ethnicity, education level, current residence, monthly household income, marital status) and other variables of interest for the study population. Frequencies and percentages were used to describe ordinal and nominal data. Mean, standard deviation, and frequency distribution were used to describe continuous variables.

A multiple logistic regression was used to identify correlates of continuous smoking during pregnancy. The smoking status was the dependent variable and age, level of education, ethnicity, residence, marital status, household income, indoor smoking, BMI, medical history, pregnancy outcomes, partner smoking, anxiety and depression symptoms, dyadic coping, negative/positive partner interactions, teamwork standards and dyadic efficacy were the independent variables considered in the analysis.

To maximize useable sample size, we coded missing values as a separate category for each variable included in the multivariate regression. Specifically, a “missing value” category was created for each categorical variable when there were more than 10 missing values. This strategy reduces potential bias that may be induced by excluding from the multivariate analyses the observations with missing data in any of the analyzed variables.

All the analyses were carried out with the Statistical Package for the Social Sciences software, SPSS version 23 (SPSS 2016). The level of statistical significance was set at 95% (p < 0.05).

Results

Descriptive Statistics

Out of the total sample of participants (n = 143), 62 (43.4%) participants were current smokers and 81 (56.6%) were ex-smokers. The descriptive variables of interest are displayed in Table 1. The age of participants in the research study ranged from 19 to 41 years, with a mean age of 28.90 and a SD of 5.345. Over 86% (n = 123) of the women were of Romanian ethnicity, close to half of the participants (49%, n = 70) had college studies or more, over 76% (n = 107) of the participants were married, and approximately 61% of the women were first-time pregnant.

Table 1.

Descriptive statistics on variables of interest

Variable	Measurement	N	%
Dependent variable
Smoking status	Smoker	81	56.6
	Ex-smoker	62	43.4
Independent variables
Age	18–25	28	19.6
	> 25	105	73.4
	Missing	10	7
Ethnicity	Romanian	123	86
	Other	20	14
Level of education	High school or less	73	51
	College studies or more	70	49
Marital status	Married	107	74.8
	Not married but living with a stable partner	33	23.1
	Missing	3	2.1
Residence	Urban	81	56.6
	Rural	60	42
	Missing	2	1.4
Indoor smoking	Not allowed	52	36.4
	Allowed; not allowed but with exceptions	77	53.8
	Missing	14	9.8
Household income	< 315.23€	46	32.2
	315.44€−1,051€	71	49.7
	> 1,051€	22	15.4
	Missing	4	2.7
Chronic medical condition present	No	119	83.2
	Yes	20	14
	Missing	4	2.8
Pregnancy outcome new variable	First pregnancy	86	60.1
	Not first pregnancy, poor pregnancy outcome	28	19.6
	Not first pregnancy, not poor pregnancy outcome	26	18.2
	Missing	3	2.1
BMI	Underweight	22	15.4
	Normal weight	83	58
	Pre-obesity and obesity	29	20.3
	Missing	9	6.3
PHQ4	Normal	63	44.1
	Mild + moderate to severe	68	47.6
	Missing	12	8.3
Partner smoking	Yes	70	49
	No	65	45.5
	Missing	8	5.5
Variable name	Mean	SD	Range
Teamwork standards	7.23	3.27	1–10
Dyadic efficacy	68.49	28.82	0–100
PIQ positive	8.99	3.64	0–12
PIQ negative	5.38	3.52	0–12
Dyadic coping	29.81	6.42	0–40

Correlates of Smoking Cessation During Pregnancy in a Sample of Romanian Pregnant Women

Higher education was a significant correlate of smoking cessation during pregnancy (Table 2). Women with a college degree or more had higher odds to quit smoking during pregnancy in comparison to their counterparts who graduated high school or less (OR 14.30; 95% CI 2.75–74.28). In addition, pregnant women who reported positive interactions with their partners in what regards their smoking had higher odds to quit smoking during pregnancy (OR 1.48; 95% CI 1.15–1.91).

Table 2.

Multivariate logistic regression results: prenatal smoking cessation (vs continued pregnancy smoking)

Outcome: prenatal smoking cessation (vs continued pregnancy smoking)	Odds ratio	95% CI
			Lower	Upper
College education or higher (vs ≤highschool or less)	14.30*	2.75	74.28
Household income €315–€1,051(vs ≤ less than €315)	.55	.13	2.28
Household income > €1,051(vs ≤ less than €315)	.22	.03	1.60
No chronic medical condition present (vs ≤ at least one medical condition)	.23	.02	2.32
Romanian ethnicity (vs ≤ other)	1.64	.21	12.56
Married (vs ≤ not married but living with a partner)	1.39	.33	5.89
Urban residence (vs ≤ rural)	.41	.12	1.41
Previous poor pregnancy outcome (vs first pregnancy)	1.20	.23	6.31
No previous poor pregnancy outcome (vs first pregnancy)	.48	.08	2.86
Mild + moderate to severe anxiety and depression symptoms (vs ≤ normal)	.59	.16	2.13
Older than 25 (vs ≤ 18–25 years old)	.20	.04	1.08
Allowed; not allowed but with exceptions indoor smoking (vs ≤ allowed)	.76	.22	2.60
Normal BMI (vs ≤ underweight)	1.34	.20	8.84
Pre-obese or obese	1.72	.20	14.83
Dyadic coping score	1.08	.95	1.23
Dyadic efficacy	.98	.96	1.00
Teamwork standards	.93	.77	1.12
PIQ negative	.89	.71	1.12
PIQ positive	1.48*	1.15	1.91

Missing value categories were coded and used for some independent variables to maximize the useable sample size in the multivariate regression. The missing value category estimated coefficients are not reported in this table

^*p < 0.05.

Although not reaching statistical significance, being married (OR 1.39) and having a previous poor pregnancy outcome (OR 1.20) are associated with higher odds of quitting smoking during pregnancy, while having a higher income and mild to moderate anxiety and depression symptoms (OR .591) are associated with lower odds of smoking cessation during pregnancy.

Discussion

The primary objective of this paper was to bring evidence on the link between a variety of relationship characteristics and maternal pregnancy smoking cessation, while accounting for relevant covariates, in an Eastern European middle-income country. The main findings of the study may be summarized succinctly. First, and the main contribution of the study, positive spouse or life partner interactions regarding maternal smoking was associated with significantly higher odds of maternal tobacco smoking cessation during pregnancy. Second, confirming findings in other, mainly high-income, populations in resource-rich settings, higher education was positively associated with pregnancy smoking cessation. Finally, although not reaching statistical significance, the study found that negative partner interactions regarding maternal smoking and increased anxiety and depression symptoms were negatively associated and dyadic coping was positively associated with maternal tobacco smoking cessation during pregnancy. These findings are clinically relevant for future couple-based tobacco cessation interventions.

The findings are important because they bring new insights regarding the potential role of relationship characteristics to inform future cessation programs focused on pregnant smokers and their life partners. While pregnancy tobacco cessation interventions with partner support do exist, most were not successful and did not focus on couple related concepts such as partner interactions, dyadic coping, and dyadic efficacy. The associations of couple-based concepts with pregnancy smoking cessation, although not statically significant, provide insight for future couple-based tobacco cessation interventions.

Previous studies in other pregnant populations and this team in the same target population showed that pregnancy smoking cessation was correlated with lower anxiety and depression symptoms, social support (mainly from the life partner), being married, living in an urban dwelling (vs rural), carrying a wanted pregnancy, and having better knowledge about tobacco smoking risks (Meghea et al. 2012; Blaga et al. 2017). In general, the findings of this study align with the previous research literature on pregnancy tobacco smoking and cessation.

Limitations of the Study

One of the study limitation was the fact that the relationship characteristics related to tobacco smoking were assessed retrospectively for those who quit smoking and at the time of assessment for the continuous smokers. As this study relied on cross-sectional data, it cannot determine causal effects of various characteristics and risk factors on tobacco smoking cessation during pregnancy. Another limitation was the convenience nature of the sample, mostly urban, not representative of the Romanian population of pregnant women. The magnitude and direction of the potential sampling bias cannot be accurately estimated. The results of this study are based on a convenience sample, therefore cannot be generalized to the pregnant population in Romania.

Implications of the Findings for Future Research and for Clinical Practice

Despite such limitations, the study findings are of interest for clinical practice because they provide insight for future couple-based tobacco cessation interventions. Such interventions will need to consider relationship characteristics and partner interactions. Future research needs to further investigate various relationship characteristics and the interactions between them, especially before and at the time of quit attempts. This will help shed light on relationship characteristics and partner behaviors that may be promising to be addressed in smoking cessation interventions. For example, couples’ interventions could focus on enhancing the positive partner interactions supporting smoking cessation. At the same time, like prior interventions building self-efficacy among smokers, couples’ interventions can focus on improving dyadic efficacy for smoking cessation in both partners.

Significance.

While pregnancy tobacco cessation interventions with partner support do exist, most were not successful and did not focus on couple-related concepts such as partner interactions, dyadic coping, and dyadic efficacy. The findings of this study may provide insight for future relationship based therapeutic cessation efforts. Future such interventions will need to consider relationship characteristics and partner interactions.