Comparison of health‐promoting behaviours, eating behaviour patterns and perceived social support in normal‐weight and overweight pregnant women: An unmatched case–control study

Abstract Aim The interventions based on adopting a healthy lifestyle during pregnancy have conflicting results. This study aimed to compare health‐promoting, dietary patterns and social support in normal and overweight pregnant women. Design An unmatched case–control design was used. Methods A total of 360 pregnant women were selected using multistage cluster sampling and divided into two groups of normal and overweight cases. Data were collected using demographic and obstetrics characteristics, health‐promoting lifestyle, perceived social support and eating behaviour questionnaires. Results The evaluation of the health‐promoting behaviours and dietary patterns demonstrated a significant difference between the mean of total scores and their subdomains including self‐actualization, nutrition, consumption of healthy and low‐fat foods, fast food and sweets, as well as emotional eating and accidental planning. There was no significant difference between the two groups about social support.


Reports have indicated the effects of non-genetic factors on
the prevalence of obesity, and lifestyle is a considerable environmental factor in this regard (Ordovas, 2018). An unhealthy diet, immobility, socio-economic factors and unfavourable social support are a part of an unhealthy lifestyle (Chen, Kuo, Chou, & Chen, 2007;Johnson & Schoeni, 2011;Shojaeezadeh, Estebsari, Azam, Batebi, & Mostafaee, 2008). Meanwhile, adopting health-promoting behaviours and healthy lifestyle are considered as the determinants of individual and social health, factors for disease prevention and weight control. In this regard, social support can have a significant impact on the quality of life and adoption of health-related behaviours due to a moderate effect on stressful events. In addition, it is a facilitating factor for continuing healthy behaviours Kazemi, Hajian, Ebrahimi-Mameghani, & Khob, 2018;Susan, Mallan, Callaway, Daniels, & Nicholson, 2017b).
Despite the desire of pregnant women to show hygienic behaviours, the interventions based on adopting a healthy lifestyle, especially in terms of weight loss, have failed and yielded conflicting results (Susan et al., 2017a(Susan et al., , 2017b. So that previous studies have demonstrated a significant difference in understanding healthy lifestyle of pregnant women with high and normal weight that exerted an impact on weight gain during pregnancy (Susan et al., 2017a(Susan et al., , 2017b. Given the fact that pregnancy is an opportunity to affect the health of two generations, the necessary support must be provided through required services to improve the pregnant women's health and take appropriate measures for ideal weight gain during pregnancy (Johnson et al., 2006). Therefore, it is vital to realize the health-related behaviours and needs of pregnant women, especially overweight and obese cases to enhance their health. We hypothesized that aspects of health-promoting, nutritional behaviours and perceptions of social support could make a difference between the two groups. In this regard, the awareness about lifestyle would help design suitable interventions. With this background in mind, this study aimed to compare the health-promoting behaviours, nutritional-behavioural patterns and perceived social support among the two groups of overweight and normal-weight women.

| Study design and participants
This unmatched case-control study was conducted during the first 6 months of 2017 in Tabriz, Iran. The inclusion criteria included the Iranian nationality, Tabriz residency, singleton pregnancy, ability to read and write in Persian, age range of 18-40 years, BMI registered within the range of 18.5-24.9 as a control and 25-29.9 as a case groups before pregnancy based on medical records, no experience of severe psychological crises over the past 6 months (declared by the participants), no known medical disorders or problems and obstetric risk factors during and before pregnancy based on medical records approved by the physician or midwife at the centre.
The exclusion criteria were lack of willingness to participate in the study and incomplete questionnaires.

| Sample size and sampling
The power analysis method was used to calculate the sample size.
Since the largest sample size was obtained by considering healthpromoting behaviours, this result was applied to estimate the sample size. In this regard, considering the results of a study carried out by Al-Kandari, Vidal, and Thomas in Kuwait (2008) the mean scores of health-promoting behaviours in normal-weight and overweight cases were 2.8 (0.53) and 2.6 (0.49), respectively. The effect size was calculated as 0.38 according to the equation. However, the sample size was calculated as 120 for each group considering 80% test power and 0.05 Type I error. In the light of the effect size of 1.5, the final sample size was estimated as 180 for each group. It should be noted that the G*Power software was exploited to calculate the sample size.
After the approvals were obtained from the authorities of the healthcare centres, a number of centres by multistage cluster sampling were randomly selected from 11 branches existing in the city, including 20 healthcare complexes and 87 healthcare centres using the Randomizer software. In total, 36 healthcare centres were selected. Afterwards, the suitable sample size was calculated and determined for each centre or according to the main sample size (N = 360) using quota sampling method and based on demographic characteristics of the centres.
Then, the list of all qualified pregnant women covered by each unit was extracted. Moreover, the names of the people were put in columns with numbers and randomly selected using computer and the Randomizer software. The lack of meeting the criteria for entering the study led to the replacement of the participant with a person randomly chosen from the list. The sampling continued until reaching the estimated sample size for both groups.
It is noteworthy that the selection of the cases introduced as the main individuals on the list was prioritized based on the quota of the centre.

| Socio-demographic and obstetrics characteristics
It consists of the demographic variables of pregnant women containing age, educational level, occupational status of pregnant women and their spouses, self-assessment of household economic status, as well as obstetrics characteristics, including the first day of the last menstruation, probable due date, gestational age based on first trimester ultrasound, number of pregnancies and childbirths, as well as height and weight before pregnancy.

| Health-promoting Lifestyle-II Questionnaire
It contains 52 items assessing six dimensions of nutrition (nine items), exercise (eight items), accountability regarding health (nine items), stress management (eight items), interpersonal support (nine items) and self-actualization (nine items). All the items are scored based on a four-point Likert scale (1 = never, 2 = sometimes, 3 = often, 4 = always). The total score for these behaviours is within the range of 52-208 (Walker, Sechrist, & Pender, 1987). The Persian version of this tool, on the population as a whole (i.e. men and women), has been evaluated in previous studies in terms of validity and reliability, and the Cronbach's alpha coefficients for the total tool and its dimensions were obtained as 0.82 within the range of 0.64-0.91, respectively. In addition, the questionnaire had sufficient stability (0.89) (Khazaeian, Kariman, Ebadi, & Nasiri, 2018;Zeidi, Hajiagha, & Zeidi, 2012).

| The multidimensional scale of perceived social support
This is a social support questionnaire designed by Zimet et al. that encompasses 12 items scored based on a Likert scale. The questionnaire evaluates three domains of perceived support from the family (four items), perceived support from family members and acquaintances (four items), and perceived support from friends (four items). The items are scored based on a seven-point Likert scale from "completely disagree" (score: 1) to "completely agree" (degree: 7) where the minimum and maximum scores are 12 and 84, respectively (Zimet, Dahlem, Zimet, & Farley, 1988). Its validity and reliability were confirmed in Iran; its validity was con-

| Eating behaviour pattern questionnaire
It contains six dimensions of low-fat eating (11 items), convenience snack foods (fast food) and sweets (10 items), emotional eating (eight items), accidental planning (six items), meal skipping (seven items) and cultural/lifestyle behaviours (nine items). All the items were scored based on a five-point Likert scale from completely agree to completely disagree (Schlundt, Hargreaves, & Buchowski, 2003). According to a study, the Persian version of the tool in the women has an appropriate validity and reliability (Dehghan, Asghari-Jafarabadi, & Salekzamani, 2015).

| Data collection
The subjects were chosen after the referral to healthcare centres and investigation of the pregnant women's medical files. The cases with BMI 18.5-24.9 before pregnancy and a group of women with BMI within the range of 25-29.9 during the same period were contacted through phone calls or in-person consultation. First, the researcher explained the objectives of the study and requested the women to determine a time and date for referring to the healthcare centre to complete the questionnaire in case of willingness to participate in the project. On referral to the centres, in addition to the primary evaluations by the researcher, a written informed consent was obtained from the subjects. Following that, the study questionnaires were completed by each participant in one of the empty rooms of the centre.

| Analysis
Data analysis was performed in SPSS software (version 21) using descriptive statistics to adjust the frequency tables and determine the central indexes, as well as the distribution of study variables to describe the features of the research units, health-promoting behaviours, social support and nutritional behaviours. Furthermore, the data were analysed using analytical statistics, including chi-square and independent t test (to compare the quantitative variables) and logistic regression analysis. The normality of quantitative data was measured based on kurtosis and skewness, all of which were normal. All the statistical tests were two-sided, using a significance level of p < .05. It should be noted that the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) standard was used to report this article.

| Ethics approval and consent to participate
The University Research Ethics Committee (ID.1395.498. as part of a PhD dissertation) approved this study. After the researchers had explained the purpose and content of the study, written informed consent was obtained from all participants.

| RE SULTS
In the present study, 17 women had no desire to participate in the study and 25 questionnaires were incomplete all of which were excluded from the study and the sampling continued until reaching 180 participants.

| Characteristics of participants
The obtained results of the present study indicated that the mean age of participants was 27.56 (SD 5.09) years and most participants (76.1%) had diploma or lower educational level and 90% were housewives.
According to the number of pregnancies, most pregnant women were nulliparous 172 (48.6%) and 126 (35%) second pregnancy. Data analysis demonstrated that there was no significant difference in the demographic characteristics of the two groups, with the exception of maternal age, number of pregnancies and educational level of the spouse (Table 1).

| Comparison of Health-promoting Behaviours in Case and Control Groups
According to the results, the mean of total score of health-promoting lifestyle in women with normal BMI and overweight women was 371.51 and 132.89, respectively. The highest mean score was related to the flourishment and nutrition field, and the lowest mean score was associated with stress management and exercise that was true in both groups. In sum, the comparison of scores of different dimensions indicated a significant difference between the two groups in terms of selfactualization and nutrition dimensions, in a way that the mean scores of these domains were lower in the group with overweight BMI, compared to that of the control group (p < .05) ( Table 2).
One-variable regression analysis showed that one-point increase in the mean score of self-actualization and nutrition led to the 5% reduction of overweight chance. In total, one-point score increase in all aspects of health promotion resulted in a 2% decrease in weight gain chance (Table 3).

| Comparison of nutritional-behavioural patterns between case and control groups
In addition, the results of data analysis indicated that the total score of nutritional-behavioural patterns between the two normal-weight and overweight groups was 156 and 160, respectively. While the highest score in the two groups was related to the consumption of low-fat and healthy foods, and cultural and lifestyle behaviours, the lowest score was related to accidental planning and skipping the meal. Regarding nutritional-behavioural patterns, the results were indicative of a significant difference between the groups considering the total score and score of dimensions of consuming low-fat and healthy foods, fast food and sweets, as well as emotional eating and accidental planning (p < .001) ( Table 4).
One-variable regression analysis demonstrated that the increase of one score in domain of fast food, sweets and emotional eating was associated with a probability of weight gain with the 1:11 ratio. In addition, the increase of one score in domain of accidental planning increased the chance of weight gain by 1.1 times (Table 5).

| Comparison of perceived social support in case and control groups
The evaluation of the overall score of perceived social support and the relevant domains in the two normal-weight and overweight groups showed no significant difference. In addition, the highest mean was related to the social protection of the family and special individuals, while the lowest mean score was related to the social support of friends (Table 6).

| D ISCUSS I ON
In the present study, the health-promoting behaviours, nutritional patterns and perceived social support were compared between the two groups of normal-weight and overweight pregnant women.
According to the obtained results, there was a significant difference between the subjects in terms of adopting a healthy lifestyle and performing health-promoting behaviours, including nutritional pattern. Nonetheless, no significant difference was observed between the two groups regarding perceived social support.
The evaluation of the mean scores of total health-promoting behaviours in both overweight (132.89) and normal-weight (137.51) groups showed that both groups had a moderate level in terms of adopting such behaviours, based on the results of other studies (Baheiraei et al., 2011;Gokyildiz, Alan, Elmas, Bostanci, & Kucuk, 2014;Malakouti, Sehhati, Mirghafourvand, & Nahangi, 2015). In this regard, the findings of this study are comparable to the results of similar studies carried out in Iran and other countries, with the exception of two studies conducted by Taopia in Thailand and Onat in Turkey that reported a better overall score, compared to those of other studies (Onat & Aba, 2014;Thaewpia, Howland, Clark, & James, 2013). This lack of consistency between the aforementioned studies and the present study might be attributed to the impact of the factors, such as cultural differences and pregnancy age, on the studied subjects. Generally, the assessed cases in the mentioned studies, all of the women were in their second trimesters of pregnancy but in our study only 182 (52.5%) were in their second trimesters. It seems that during this period, women had a more stable condition, compared to other pregnancy trimesters.
In the present study, it was concluded that overweight pregnant women obtained a lower overall score in adopting health-promoting behaviours, compared to the subjects in the control group. In addition, this difference was statistically considered significant. In this respect, the results of a study carried out by Cho et al. are in line with the findings of this study. In the aforementioned study, the status of health-promoting behaviours was evaluated in overweight and obese women within the age range of 18-65 years and it was reported that increased level of BMI was associated with the decreased total score of health-promoting behaviours (Cho, Jae, Choo, & Choo, 2014).
Moreover, the evaluation of status of health-promoting behaviours in other demographic groups, such as nursing students, suggested that fewer overweight people participated in such activities, compared to the individuals with normal weight that is in line with the findings of the present study (Al-Kandari et al., 2008;Chen et al., 2007). In this regard, according to a report by Susan et al. (2017a)  More than sufficient (ability to save money) 1 (0.6) 1 (0.6)  (Gharaibeh, Al-Ma'aitah, & Al Jada, 2005;Gokyildiz et al., 2014;Kavlak et al., 2013). This inconsistency between the results might be due to the environmental and cultural differences. Moreover, the results of a study carried out by Nies, Buffington, Cowan, and Hepworth (1998) in the United States are not consistent with the findings of the present study. In the aforementioned study, overweight non-pregnant women achieved lower scores in all aspects of health-promoting behaviours, compared to normal-weight individuals.

Dimensions of health-promoting behaviours
In addition, the assessment of the status of health-promoting behaviours in Taiwanese adolescents showed a low score in terms of social support, accountability and physical activity (Chen et al., 2007).
This inconsistency in the results might be related to the diversity of the subjects regarding the age, gender and lack of pregnancy since the specific pregnancy conditions of women in the present study can affect the adoption of behaviours. In general, pregnant women tend to change their behaviours to achieve the desired outcomes and they are more likely to display hygienic behaviours. Regarding the results, one of the important problems in overweight women was the low scores in the dimensions of spirituality and self-actualization. According to the findings of other studies, increasing the conception of spirituality and self-actualization is associated with the reduction of high-risk behaviours in pregnant women and addressing spirituality by decreasing stress improves the health condition during pregnancy. Proper interventions in this domain can be helpful. Another important issue was obtaining a low score in the physical activity dimension in both groups.
However, the factors, such as constraints for women in society and spent time for family-related tasks, can be effective; in this regard, pregnancy and belief in more rest during this period can be an important factor in getting the lowest score, compared to other dimensions of health-promoting behaviours. Given the fact that physical activity during pregnancy is associated with the improvement of maternal and neonatal outcomes and control of weight gain, planning is required for proper interventions in this respect.
The comparison of the nutritional-behavioural patterns indicated a significant difference between the two groups in this regard.
Furthermore, this comparison showed a correlation between eating pattern and BMI, in a way that overweight women had improper nutritional patterns, such as the consumption of fast food and sweets, and emotional eating with no planning and consumed less healthy and low-fat foods, compared to the subjects in the control group.
These findings are in line with the results of the present study (Cardon et al., 2016;Chitsaz, Javadi, Lin, & Pakpour, 2017). Other studies have also reported the association between the overall score of nutritional behaviours and BMI (Bashirian, Jalily, & Barati, 2016).
A review of the related literature revealed that overweight people usually have an improper diet and deal with more nutritional deficiencies, compared to normal-weight individuals. In addition, the unbalanced reception of macronutrients and micronutrients is higher in these subjects (Groth & Morrison-Beedy, 2013;Hui et al., 2012).
As it was reported by Kolko, Emery, Marcus, and Levine (2017), a significant percentage of overweight pregnant women lose their eating control in pregnancy, compared to pre-pregnancy period and emotional eating increased in these cases. Shloim, Rudolf, Feltbower, Blundell-Birtill, and Hetherington (2018) marked a significant relationship between BMI and emotional eating and increased fast food consumption.
On the other hand, in the aforementioned study, no association was observed between dietary restrictions and BMI during pregnancy.
Similarly, no significant relationship was noticed between BMI and skipping a meal in the present study. According to the literature, it was TA B L E 3 Results of logistic regression analysis for evaluation of effect of health-promoting behaviours on body mass index According to the obtained results of the present study, however, the overall score and various dimensions of perceived social support were lower in overweight individuals, compared to those of normal-weight subjects, and this difference was not statistically significant. In addition, low scores were obtained in both groups in terms of understanding friends' support. These findings are comparable to the results of other studies. Based on the results of a study carried out by Susan et al. (2017a) in Australia, no difference was observed between the two groups of overweight and normal-weight pregnant women in terms of perceived social support. Similarly, Johnson et al. (2006) (Stark & Brinkley, 2007;Sui, Turnbull, & Dodd, 2013;Walker, Cooney, & Riggs, 1999), it seems that further studies are required to take measures in this regard.

| LIMITATI ON S OF THE S TUDY
There are some limitations despite the fact that the present study is the first one to examine the differences in health-promoting lifestyle in overweight and normal-weight pregnant women. Self-report was one of the drawbacks of this study. In addition, the lack of reliable and valid questionnaires for the pregnant women was another study limitation. Therefore, the obtained results of the present study TA B L E 5 Results of logistic regression analysis for evaluation of effect of nutritional-behavioural patterns on body mass index can only be generalized to overweight and normal-weight cases and not to obese and low-weight individuals. Although the sampling did not match, we used random sampling. Furthermore, since different changes occur in various pregnancy trimesters, it is suggested to perform and compare reviews every 3 months between matched groups.

| CON CLUS ION
The identification and better perception of the status of health-related behaviours, such as how to adopt health behaviours, especially nutrition and social support perception, as a moderator of behaviours for positive support and health improvement are significantly crucial. The obtained results of the present study showed that while overweight pregnant women achieved lower scores in adopting health-promoting behaviours, both groups obtained moderate scores. They especially gained lower scores in the dimensions of stress management and physical activity. In addition, overweight women had unfavourable conditions in most domains, and despite being pregnant, they had an inappropriate food pattern. Both groups had a moderate condition in terms of social support, and the support from friends was reported at the lowest level. Therefore, according to the obtained results of this study, it is recommended to integrate healthy plans in future healthpromoting interventions to achieve optimal results.

ACK N OWLED G EM ENTS
We thank the women who participated in the study. We also appreciate the funding support from Shahid beheshti University of Medical Sciences (SBMU).

CO N FLI C T O F I NTE R E S T
The authors declare that they have no competing interests.

AUTH O R CO NTR I B UTI O N S
AF-K contributed to development of the concept, collected data, analysed data and wrote the draft and final article. SH contributed to development of concept and reviewed the draft and final article.
All authors read and approved the final manuscript.