Characteristics of health behaviours and health status indicators among pregnant women in Slovenia 295 IzvIrnI znanStvenI članek Področje medicine: Mikrobiologija in imunologija Stomatologija Nevrobiologija Onkologija Reprodukcija človeka Srce in ožilje Metabolne in hormonske motnje Javno zdravstvo (varstvo pri delu) Psihiatrija Tip članka: Uvodnik Izvirni znanstveni članek Pregledni znanstveni članek Klinični primer Strokovni članek id 1 Department of Obstetrics and Gynecology, University Medical Centre Ljubljana, Ljubljana 2 Faculty of Pharmacy, University of Ljubljana, Ljubljana 3 Faculty of Medicine, University of Ljubljana, Ljubljana Korespondenca/ Correspondence: ksenija Geršak, e: ksenija. gersak@mf.uni-lj.si Ključne besede: nosečnost; zdravila; navade; zdravje nosečnic; socialno demografski dejavniki Key words: pregnancy; medication; health related behavior; socio-demographic characteristics Citirajte kot/Cite as: zdrav vestn. 2017; 86:295–317. Prispelo: 8. 2. 2017 Sprejeto: 5. 6. 2017 reprodukcija človekaIzvirni znanstveni članek zdrav vestn | julij – avgust 2017 | letnik 86 Characteristics of health behaviours and health status indicators among pregnant women in Slovenia Navade, povezane z zdravjem in socialno demografske značilnosti nosečnic v Sloveniji tina kek,1 nataša karas kuželički,2 Irena Mlinarič raščan,2 ksenija Geršak1,3 Izvleček Izhodišča: Nosečnost je obdobje, ko ženske za zdravje bodočega otroka kritično presojajo in ocenju- jejo svoje zdravje ter z zdravjem povezane navade. Skupaj z zdravstvenimi delavci se srečujejo s šte- vilnimi tveganji in odločitvami, ki postanejo veliko kompleksnejše, saj je v igri vpliv na dva organiz- ma. Namen članka je pregled ter iskanje morebitnih povezav med zdravjem, z zdravjem povezanimi navadami ter socialno demografskimi dejavniki v nosečnosti, ki so jih poročale sodelujoče ženske. Metode: Podatke študije smo pridobili v okviru raziskovalnega projekta Analiza bioloških označe- valcev presnove folatov pri ugotavljanju tveganja za nastanek napak nevralne cevi, orofacialnih shiz ter prirojenih srčnih napak, ki je potekal od maja 2013 do septembra 2015. Vprašalnik o maternalnem zdravju, z zdravjem povezanih navadah ter socialno demografskih karakteristikah je izpolnilo 450 sodelujočih žensk. Podatki zajemajo nosečnosti, ki datirajo od osemdesetih let prejšnjega stoletja pa do leta 2015. Rezultati: Mlajše in manj izobražene ženske so pogosteje poročale o kajenju med nosečnostjo, med- tem ko so starejše, bolj izobražene ter nulipare ženske v višjem odstotku jemale folatne/multivi- taminske dodatke. Opazili smo U-obliko porazdelitve uporabe zdravil (brez recepta in na recept) glede na stopnjo izobrazbe. Najvišji odstotek uporabe so imele ženske z najvišjo stopnjo izobrazbe (magisterij/doktorat) in ženske z doseženo najnižjo stopnjo izobrazbe (osnovna šola). Večjo uporabo zdravil med nosečnostjo so poročale tudi starejše ženske. Opazili smo trend višanja uporabe zdra- vil in folatnih/multivitaminskih pripravkov kot tudi incidence nosečniške sladkorne bolezni preko preiskovanega obdobja, z najvišjo incidenco v zadnjih desetletjih. Višjo pojavnost kroničnih bolezni smo opazili v skupini multiparih žensk. Zaključki: V navadah povezanih z zdravjem nosečnic obstajajo precejšnje socialno demografske razlike. Za izboljšanje zdravja so potrebni izboljšani javno zdravstveni ukrepi in individualno sveto- vanje, da lahko ustrezno naslovimo specifične potrebe socialno demografskih skupin z višjim tvega- njem za neželene izide nosečnosti. Abstract Background: Pregnancy is a period when women reconsider their own health and health related behaviour for the sake of their future child. Along with their health providers, they are faced with a number of risk assessments and decisions, which become far more complex as their effect on two or- ganisms rather than one is at play. This paper provides an overview of possible associations between self-reported health status, health behaviours and socio-demographics during pregnancy. 296 zdrav vestn | julij – avgust 2017 | letnik 86 reProdukCIja človeka Methods: Study data were obtained from the case-control research project “Analysis of folate me- tabolism biomarkers in the risk assessment for neural tube defects, orofacial clefts and congenital heart defects”, which recruited participants from May 2013 to September 2015. Questionnaires about maternal health, health related behaviour and socio-demographic characteristics were completed by 450 women. The data include pregnancies from the 1980s to 2015. Results: We observed that younger and less educated women more frequently reported positive smoking status during pregnancy, while higher prevalence of folate/multivitamin supplementation was found among more educated, older and nulliparous women. There was a U-shaped distribution of medication intake (over-the-counter and prescribed) with respect to educational level, with the highest intake in mothers with a masters/PhD degree and among those that completed elementary school. Higher medication usage was also reported among older women. With increasing maternal age there was an increase in medication intake, folate/multivitamin intake, as well as incidence of gestational diabetes over the studied time period, with the highest frequencies occurring in later de- cades. A higher incidence of chronic diseases was observed in a group of multiparous women than among monoparous women. Conclusions: Considerable socio-demographic disparities exist in health-related behaviour among pregnant women. Improved public health campaigns and individual health care counselling are needed to address specific requirements of socio-demographic groups at higher risks of adverse pregnancy outcomes. 1. Introduction Pregnancy is an important time in the lives of many women. It is also the time when they may consider or reconsider their own health, as well as the health of their future child. Some women may experience a conflict between the mana- gement of the two (1). Since the thalidomide disaster of the late 1950s and early 1960s women have been given strong messages about the importance of maintaining their health and avoiding toxins that can be transfer- red from mother to baby. Once pregnan- cy is confirmed, women are faced with a multitude of decisions and risk asses- sments  (2). They must decide what to eat, what to drink, what test they will un- dergo and nearly every pregnant woman faces a decision about whether or not to take a medication during pregnancy, ranging from over-the-counter (OTC) antacids for gastroesophageal reflux to prescription medications for life-thre- atening chronic conditions  (2,3). In the last three decades the use of medications during pregnancy has increased and it is now common during pregnancy  (3,4). Estimates of the usage prevalence of at least one prescription medication du- ring pregnancy range from less than 30 % to over 90 % of women (3). In mul- tinational study that collected data in years 2011–2012, 71.1 % of participating women reported use of medication du- ring pregnancy (OTC and/or prescribed medication) in Slovenia  (5). The safety of medication use during pregnancy is frequently unclear because, for the ma- jority of medications, there is a lack of sufficient data available to fully characte- rise the foetal risk. Thus, it is difficult to make informed clinical decisions about the best management of acute and chro- nic disorders during pregnancy  (3,4). Since 1979, health care providers used a standard five-letter nomenclature publi- shed by the Food and Drug Administra- tion (FDA) to assign a pregnancy risk category to medications  (6). The FDA recently revised its existing system and Characteristics of health behaviours and health status indicators among pregnant women in Slovenia 297 IzvIrnI znanStvenI članek introduced a new “Pregnancy and Lac- tation Labelling Rule”, which became ef- fective in June 2015. Women also make important decisi- ons during pregnancy in terms of other types of exposure, for example, alcohol, smoking and certain food items  (1). Many women continue to consume al- cohol in pregnancy despite an increasing body of evidence suggesting harm to the foetus  (7,8). Maternal smoking du- ring pregnancy also poses a significant risk to the unborn child. As there is no safe lower limit of cigarette use during pregnancy, the World Health Organiza- tion advises pregnant women to abstain from cigarettes altogether  (9). Despite the extensive information about dangers that smoking poses to their foetus, some women continue smoking. It has been estimated that 10–27 % of the pregnant women in the European Union continue smoking during pregnancy  (9,10). In 2014 published multi-national study, Slo- venian data revealed that out of 32.9 % of women who smoked before pregnancy, 6.7 % continued smoking during pre- gnancy (9). According to a study, Peter- sen et al. (2014), on women’s perception of risks of adverse foetal pregnancy out- comes that also included Slovenian data, alcohol and smoking were perceived to carry high risks in pregnancy  (1). Both substances are considered deleterious to the foetus although there are still deba- tes whether there is a safe threshold for drinking alcohol during pregnancy  (1). Raymond et al.  (2009) study showed that women found information and ad- vice about safe levels of drinking in pre- gnancy confusing and lacking in eviden- ce and detail (1,11). In contrast, the importance of nutriti- on during pregnancy with regard to the pregnancy outcome has long been reco- gnised  (12,13). For example, pregnancy causes a two- to three-fold increase in the requirement for iron, not only for haemoglobin synthesis but also for the foetus and the production of certain enzymes. There is also a ten- to twenty- -fold increase in folate requirements and a two-fold increase in the requirement for vitamin B12 (14). Over the last three decades, substantial epidemiological evidence has emerged to indicate that periconceptional folate and multivita- min supplementation reduces the risk of adverse pregnancy outcomes  (15,16). Women rely not only on medical in- formation from the expert knowledge of their health care providers but also on their own experience, education and cultural understanding  (2,17). Research related to a variety of perinatal and in- fant health outcomes suggested that ma- ternal socio-economic status (SES) that generally correlates with maternal level of education prominently influences perinatal and infant health  (17). Altho- ugh many studies have suggested that a lower SES is associated with an increa- sed risk of neural tube defects (NTDs), many other studies found no correlation with adverse pregnancy outcomes when adjusting for race-ethnicity, multivita- min intake, cigarette smoking and binge drinking (16-18). Public health discourse has increasin- gly framed personal health choices as so- cial and moral issues, and, as a consequen- ce, many women feel overly responsible for producing and maintaining a he- althy child  (2). This increasing pressure has resulted in hyper-vigilant pregnant women going out of their way to avoid any possible toxins, while also adopting unhealthy behaviours, such as smoking and excessive eating in response to the stress of “intensive motherhood”  (2,19). A recent Slovenian study reported that the same dilemmas continue during po- stpartum period, showing that breastfee- ding mothers used less medications and 298 zdrav vestn | julij – avgust 2017 | letnik 86 reProdukCIja človeka nutritional supplements compared to non-breastfeeding mothers (20). Precise and accurate information on medicati- ons, possible environmental toxins, diet and other health-related behaviours du- ring pregnancy is called for in order to help women and their health providers rationally solve their dilemmas. The ac- tion points of the European board and college of obstetrics and gynaecology (EBCOG, 2016) position statement asks the scientific public for an EU-wide as- sessment of current practice and future needs along with informational campa- ign to facilitate proactive information sharing and counselling about medica- tion use in pregnancy (21). Considering the latter, the aim of this paper is to ob- serve for possible associations between women’s self-reported health status, he- alth behaviours and socio-demographics during pregnancy in different time peri- ods. 2. Methods 2.1. Study design and participants This study was an observational, re- trospective cohort study of 450 mothers that delivered live-born, healthy new- -borns, with no major birth defects. It was part of the case-control research project “Analysis of folate metabolism bi- omarkers in the risk assessment for neu- ral tube defects, orofacial clefts and con- genital heart defects” (ARRS J3–5507). Women were recruited during routine post-partum 3-day hospitalization at the Division of Gynaecology and Obstetri- cs, University Medical Centre Ljublja- na (UMCLJ) and during routine follow up examinations of their children with orofacial clefts (OFC) or congenital he- art disease (CHD) at the Department of Maxillofacial and Oral Surgery at UM- CLJ and the Paediatric Clinic of UM- CLJ, respectively. In the latter instances pregnancy data of the healthy siblings of case children were included in the study. This is why the data do not cover just re- cent years but date back to 1980s because healthy siblings were born in a time in- terval from 1980 to 2015. This study was an observational, re- trospective cohort study of 450 mothers; 200 mothers who gave birth to healthy children without birth defects were recruited during routine 3-day post- -partum hospitalisation between 2013 and 2015. They provided information about their behaviour during their just- -concluded pregnancy. The most recent pregnancies included in this study were concluded in 2015. The remainder of the study parti- cipants were recruited during doctors’ appointments for their children with cleft lip/palate (OFC) or congenital heart disease (CHG). Mothers were recruited who delivered at least one live-born, he- althy new-born, with no major birth de- fects, before delivering at least one child with cleft lip/palate (OFC) and conge- nital heart disease (CHG). The mothers provided questionnaire answers about their behaviour during the pregnancy of their last child born with no birth defec- ts. The earliest pregnancies included in this study were concluded in 1980. Every participant gave information about only one pregnancy, i.e. their last pregnancy that was concluded with a li- ve-born child with no birth defects. The study gathered self-reported in- formation on the use of medication, folate and nutritional supplementation, health status, smoking status, and socio- -demographic characteristics via questi- onnaire. Before completing the questio- nnaire, all mothers were required to sign informed consent forms. The study was approved by the National Medical Ethics Committee (No. 57/02/13). Characteristics of health behaviours and health status indicators among pregnant women in Slovenia 299 IzvIrnI znanStvenI članek 2.2. Evaluation of maternal health-related and socio- demographic characteristics All participating mothers completed a questionnaire that included questions about their age at conception, height and weight, smoking status, education, num- ber of pregnancies, live births and mi- scarriages, child’s gender, history of ge- stational diabetes and chronic diseases, drug use during pregnancy, fever during pregnancy, folate and vitamin supple- mentation before conception and during pregnancy. 2.3. Drug classification Reported medications taken during pregnancy by the women in the studi- ed cohort were classified according to the ATC (Table 2) and FDA classificati- ons (Figure 3D). Statistical analysis was performed for drugs of all FDA classes (A,B,C,D,X) and for the three most repre- sented (> 5 %) groups of medications in the study cohort according to ATC classi- fication, i.e. A12C (other mineral supple- ments), B03A (iron preparations) and N02B (other analgesics and antipyretics) . 2.4. Statistical analysis Variables evaluated by the question- naire were classified as either continuo- us (maternal age at conception, height and weight, number of pregnancies, live births and miscarriages, year of delive- ry) or categorical (maternal smoking status, education, child’s gender, history of gestational diabetes and chronic di- seases, drug use during pregnancy, fever during pregnancy, folate and vitamin supplementation before conception and during pregnancy). The Kolmogorov- -Smirnov test was used to test the dis- tribution of continuous variables. For univariate statistical analysis the Mann- -Whitney U test was used for non-Ga- ussian continuous and ordinal variables and Fischer’s exact test for categorical va- riables. Logistic regression models were used to further explore characteristics of study cohort. Purposeful selection of candidate variables was done based on p value < 0.250, calculated by means of univariate statistical analysis. Only vari- ables that met these criteria were inclu- ded in the logistic regression models and adjusted for co-variables. Odds ratios (OR) were determined using logistic re- Table 1: Main characteristics of the study cohort. n designates the number of women included in the analysis. Characteristics N Number of cases (%) Smoking in pregnancy 449 52 (12 %) Fever in pregnancy 450 20 (4 %) Gestational diabetes 450 43 (10 %) Maternal chronic disease 450 54 (12 %) drug administration in pregnancy 450 179 (40 %) Folate and/or multivitamin supplements in pregnancy 450 351 (78 %) use of folate supplements : • Containing folic acid only • Containing metafolin and folic acid 450 317 317 317 (70 %) 291 (92 %) 26 (8 %) 300 zdrav vestn | julij – avgust 2017 | letnik 86 reProdukCIja človeka gression models, 95 % confidence inter- vals for OR and p values. A p value < 0.05 was considered statistically significant. All analyses were carried out using IBM SPSS Statistics 22. To calculate the cor- relation between the independent varia- bles of logistic regression models Spear- mann rho test was used. 3. Results A total of 450 women participated in the study. The main characteristics are shown in Table 1. All characteristics were investigated according to the delivery period, maternal education, parity and age. Figure 1: association of maternal demographic characteristics with the incidence of smoking in pregnancy, maternal chronic disease and gestational diabetes. Only characteristics exhibiting a statistically significant difference across the subgroups are shown. (A) Mothers with elementary school level of education had an increased prevalence of smoking in pregnancy compared to mothers with a college education (OR(95 %CI): 6 (1.3–26), p = 0.021) or university degree (OR(95 %CI): 10 (3–36), p < 0.001). (B) Mothers who smoked in pregnancy were younger at conception than non-smoking mothers (p = 0.019). The significance was not present after adjusting for maternal education in the logistic regression model. (C) Mothers who had more than 2 children had an increased prevalence of chronic diseases (OR(95 %CI): 3 (1.3–8), p = 0.027) compared with mono-parous mothers. (D) The data suggests a trend of increased incidence of gestational diabetes in the group of mothers who gave birth in more recent years (OR(95 %CI): 1.14 (1.04–1.26), p = 0.008). Characteristics of health behaviours and health status indicators among pregnant women in Slovenia 301 IzvIrnI znanStvenI članek 3.1. Maternal characteristics Of the 449 women who gave infor- mation on their smoking status, 52 (12 %) reported smoking during pregnancy (Table 1). According to the results, the percentage of smokers during pregnancy has decreased in the last 30 years, as 18 % of mothers, who delivered in 1980s, re- ported smoking during pregnancy com- pared to 10 % of woman who delivered in the period of 2011–2015. The decrease however was not statistically significant (p = 0.372). Difference in the percen- tage of smokers between study subsets with different levels of education was statistically significant (p = 9.6 × 10-5). A lower level of education was associated with smoking during pregnancy (32 % of women who finished elementary school smoked during pregnancy compared to just 5 % of those with masters/PhD de- grees). The significant association of ma- ternal education with smoking during pregnancy was still present after adju- sting for maternal age at conception in a logistic regression model (Figure 1A). There was no correlation between parity and smoking status (p = 0.259), with 11 %, 14 % and 8 % of smokers among women who had 1, 2 and > 2 children, respecti- vely. A statistically significant correlation between smoking in pregnancy and ma- ternal age at conception was observed (p = 0.019): smoking mothers conceived at a younger age. Adjusting for maternal education in a logistic regression mo- del (Figure 1B) statistical significance did not persist, possibly because of the correlation between maternal education and maternal age at conception (Spear- man rho = 0.420, p = 1.6 × 10-19). Out of 450 participating mothers, 20 (4 %) reported fever during pregnan- cy. There was no statistically significant association between reported fever du- ring pregnancy and year of delivery (p = 0.785), level of maternal education (p = 0.654), parity (p = 0.170) and mater- nal age at conception (p = 0.621)). Due to the small number of subjects that expe- rienced fever during pregnancy, logistic regression was not performed. Chronic disease was reported by 54 (12 %) participants and 396 (88 %) denied having a chronic health condition. There was a statistically significant association between parity and reported chronic di- sease. Mothers with 2 (10 %) and especi- ally more than 2 children (22 %) reported higher prevalence of chronic disease than monoparous women (8 %) (p = 0.004). The significance (for mothers with more than 2 children) was still present after adjusting for maternal age at conception and date of delivery in a logistic regres- sion model (Figure 1C). There was no si- gnificant association of chronic disease with any other maternal characteristic, i.e date of delivery (p = 0.192), maternal level of education (p = 0.560) and age at conception (p = 0.157). Among the 450 participants, 43 (10 %) reported occurrence of gestatio- nal diabetes (GD). There was suggested trend of higher GD incidence among women that delivered in more recent years, p = 0.008) (Figure 1D) and among older mothers (borderline statistical significance; p = 0.085). There was no significant association of GD with ma- ternal education (p = 0.501) and parity (p = 0.490). Because of the small num- ber of GD cases, it was not possible to construct a suitable logistic regression model. The majority of participating mothers reported supplementation with folic acid and multivitamins (N = 351, 78 %) during pregnancy (Table 1). Women who gave birth more recently were much more li- kely to take folate and multivitamin su- pplements than those who delivered in the earlier years (p < 0.001). The signi- 302 zdrav vestn | julij – avgust 2017 | letnik 86 reProdukCIja človeka Figure 2: association of maternal demographic characteristics with folate and multivitamin supplementation during pregnancy. Only characteristics exhibiting statistically significant difference across the subgroups are shown. (A) Mothers who gave birth in more recent years had greater folate and multivitamin consumption in pregnancy compared to those who delivered in earlier years, OR(95 %CI): 1.17 (1.11–1.24), p < 0.001. (B) Mothers with higher levels of education (i.e. vocational school, college, university and master/PhD degree) had higher folate and multivitamin consumption in pregnancy compared to those who completed only elementary school, OR(95 %CI): 4(1.3–11), p = 0.014, 9(2–36), p = 0.003, 8(2–26), p = 0.001, 10(1.4–64), p = 0.019, respectively. (C) Mono-parous mothers had higher consumption of folates and multivitamins during pregnancy compared to mothers having two (OR(95 %CI): 4(1.1–11), p = 0.028) or more than two children (OR(95 %CI): 4(1.2–15), p = 0.021). (D) Mothers who took folate and multivitamin supplements during pregnancy were older than those who did not (p < 0.001). The significance was not present after adjusting for year of delivery, maternal education and parity in the logistic regression model. ficance was still present after adjusting for education, parity and maternal age at conception in the logistic regression mo- del (Figure 2A). Furthermore, there was a significant statistical association betwe- en folate and multivitamin supplementa- tion and the level of maternal education. Mothers with lower level of education (i.e. elementary school) had a considera- bly lower percentage of supplementation (32 %) in contrast to mothers with finis- hed college (87 %), university (90 %) and master/PhD degree (91 %) (p < 1 × 10-6). The significance persisted after adjusting for delivery period, parity and maternal age at conception in the logistic regres- sion model (Figure 2B). There was also a statistically significant association be- Characteristics of health behaviours and health status indicators among pregnant women in Slovenia 303 IzvIrnI znanStvenI članek tween supplementation and parity, with highest percentage of supplement usage among monoparous women (96 %) in comparison to women who gave birth to 2 (73 %) and more than 2 children (67 %) (p < 1 × 10-6). The significance persisted after adjusting for maternal education, parity and maternal age at conception in the logistic regression model, (Figure 2C). Supplement usage was more preva- lent among older women, with higher age at conception, compared to younger mothers (p < 0.001), though significance was not present after adjusting for year of delivery, maternal education and pa- rity in logistic regression model (Figure 2D). Since only 8 % (N = 26) of mothers who took folate supplements used 5-me- thylfolate preparations, further analysis in this respect was not performed. Consumption of medications (pre- scribed or OTC) during pregnancy was reported by 60 % (N = 271) of partici- pating women (Table 1). The prevalence of medication usage during pregnan- cy has increased in the latest decades (p < 0.001). None of the women who gave birth in 1980s reported medication consumption during pregnancy, while in the following decades there was a signi- ficant trend of increased consumption (in 1990s 26 %, in 2001–2010 32 % and in 2011–2015 48 %). Significant association of higher drug intake with later year of delivery was still present after adjusting for maternal education and maternal age at conception in the logistic regressi- on model ((OR(95 %CI): 1.07(1.03–1.18), p = 0.001) (Figure 3A). We observed a U distribution of re- ported medication usage versus the level of maternal education (p = 0.012). The highest prevalence of medication usa- ge during pregnancy was noted in gro- ups with the lowest (elementary school, 44 %) and the highest levels of education (university, 46 % and master/PhD degree, 64 %) compared to groups with middle- -ranged levels of education (vocational school, 34 %, high school, 37 % and col- lege, 26 %). In accordance with this, after adjustment in the logistic regression mo- dels, mothers with completed elementa- ry school and masters/PhD degrees had higher prevalence of drug consumption than those with college degrees (Figure 3B). Parity showed no statistically signi- ficant correlation with medication usage during pregnancy (p = 0.339). There was a statistically significant trend of higher drug consumption during pregnancy among mothers that were older at con- ception (p = 0.022), however the signifi- cance did not persist after adjusting for the year of delivery and maternal edu- cation in the logistic regression model (Figure 3C). 3.2. Medication usage in pregnancy according to the ATC and FDA classifications 3.2.1. ATC classification Intake of A12C group of drugs was reported by 27 (6 %) of participating women. The consumption of A12C me- dications varied across different delivery periods (p = 8.2 × 10-5). Ten percent of women who delivered after 2010 used A12C medications. Statistically signifi- cant increase in the incidence of A12C drug consumption was observed for mothers with higher levels of education (elementary school, 0 %, university, 8 % and master/PhD degree, 14 %; p = 0.009) and in monoparous mothers (15 % in contrast to mothers with 2 (4 %) or more than 2 children (0 %); p = 1 × 10-6). Due to the small number of women taking A12C medications a logistic regression model was inapplicable. Iron preparation intake (ATC code B03A) was reported by 78 (17 %) par- 304 zdrav vestn | julij – avgust 2017 | letnik 86 reProdukCIja človeka Figure 3: association of maternal demographic characteristics with drug administration in pregnancy. Only characteristics exhibiting statistically significant difference across the subgroups are shown. (A) Mothers who gave birth in more recent years had higher prevalence of drug consumption during pregnancy compared to those who delivered in the earlier years (OR(95 %CI): 1.07(1.03–1.18), p = 0.001). (B) The distribution of prevalence of drug consumption according to maternal education was U shaped, with higher prevalence during pregnancy in mothers with the lowest (elementary school) and highest (university and master/PhD degrees) level of education (p = 0.012). In accordance with this, after adjustment in the logistic regression models, mothers who completed elementary school (OR(95 %CI): 4(1.3–13), p = 0.018) and masters/PhD degrees (OR(95 %CI): 5(1.5–15), p = 0.007) had a higher prevalence of drug consumption than those with a college degree. (C) Mothers who took drugs during pregnancy were older than those who did not (p = 0.022). The significance was not present after adjusting for year of delivery and maternal education in the logistic regression model. (D) The distribution of incidence of drug usage during pregnancy according to FDA classification. ticipants. All of them gave birth after 1990, the highest incidence was after 2011 (1980s: 0 %, 1990s: 13 %, 2001–2010: 13 % and 2011–2015: 22 %). This trend was borderline statistically significant with a p value of 0.058. Correlations to all other tested maternal characteristics were statistically insignificant, i.e. level of maternal education (p = 0.694), parity (p = 0.178) and maternal age at concepti- on (p = 0.763). Due to the small number of women who took B03A medications, a logistic regression model was inappli- cable. Consumption of N02B drugs was re- ported by 30 (7 %) of participants. N02B intake according to maternal level of education gave U distribution, with the Characteristics of health behaviours and health status indicators among pregnant women in Slovenia 305 IzvIrnI znanStvenI članek Table 2: Prevalence of drug consumption during pregnancy in the study cohort according to the atC classification. ATC group Cases taking drugs from the specific ATC group Count (%) A ALIMENTARY TRACT AND METABOLISM a02a antacids 7 (1,6) a02B drugs for peptic ulcer and Gord 2 (0,4) a03B Belladona & derivatives 10 (2,2) a07B Intestinal adsorbents 1 (0,2) a07e Intestinal antiinflammatory agents 2 (0,4) a07F antidiarrheal microorganisms 1 (0,2) a11d vitamin B1 1 (0,2) a12a Calcium 2 (0,4) a12C other mineral supplements 27 (6,0)* B BLOOD AND BLOOD FORMING ORGANS B01a antithrombotic agents 2 (0,4) B03a Iron preparations 78 (17,3)* C CARDIOVASCULAR SYSTEM C01a Cardiac glycosides 1 (0,2) C01B antiarrhythmics, class I & III 1 (0,2) C01d vasodilatators for cardiac diseases 1 (0,2) C02a antiadrenergic agents, central acting 2 (0,4) G GENITO URINARY SYSTEM AND SEX HORMONES G01a antiinfectives & antiseptics 3 (0,7) G03C estrogens 1 (0,2) G03d Progestogens 6 (1,3) G03F Progesterons & estrogens in combination 1 (0,2) G04B urologicals 4 (0,9) H SYSTEMIC HORMONAL PREPARATIONS, EXCL. SEX HORMONES AND INSULINS H02a Corticosteroids for systemic use 1 (0.2) H03a thyroid preparations 8 (1.8) J ANTIINFECTIVES FOR SYSTEMIC USE j01C Penicillins 18 (4.0) j01F Macrolides, lincosamides & Streptogramines 4 (0.9) j01X other antibacterials 1 (0.2) 306 zdrav vestn | julij – avgust 2017 | letnik 86 reProdukCIja človeka highest percentage of users in groups with the lowest and highest levels of education (elementary school 12 %, ma- ster/PhD degree 23 %). Also, N02B users tended to be older. These associations had borderline statistical significance with p values of 0.063 for level of edu- cation and 0.052 for age at conception. No statistically significant association was found versus parity (p = 0.202) or date of delivery (p = 0.825). Due to the small number of women who took N02B medications, a logistic regression model was inapplicable. 3.2.2. FDA classification Out of 450 participants, 87 (19 %) reported taking FDA class A medica- tions. There was a trend of increasing consumption over the time period of the study. This was especially significant for mothers who gave birth after year 2011 (1990s: 13 %, 2001–2010: 16 % and 2011–2015: 23 %). The association had a borderline statistical significance with a p value of 0.087. All other tested mater- nal characteristics [maternal education (p = 0.584), parity (p = 0.313) and mater- nal age at conception (p = 0.406)] did not show statistically significant associ- ations. Since there was only one variable that reached a threshold p value, a logi- stic regression model was not applicable. Consumption of FDA class B me- dication was reported by 59 (13 %) of participants. The highest percentage of usage was reported by mothers with the highest levels of education (master/PhD ATC group Cases taking drugs from the specific ATC group Count (%) N NERVOUS SYSTEM n02B other analgesics & antipyretics 30 (6.7)* n02C antimigraine preparations 1 (0.2) n03a antiepileptics 4 (0.9) n05a antipsyhotics 1 (0.2) n05B anxiolytics 1 (0.2) n06a antidepressants 3 (0.7) R RESPIRATORY SYSTEM r01a decongestants and other nasal preparations for topical use 1 (0.2) r02a throat preparations 3 (0.7) r03a adrenergics inhalants 2 (0.4) r03B other drugs for obstructive airway diseases–inhalants 1 (0.2) r03d other systemic drugs for obstructive airway diseases 1 (0.2) r06a antihistamines for systemic use 1 (0.2) S SENSORY ORGANS S01a antiinfectives 1 (0.2) *Only ATC classes of drugs with prevalence of consumption above 5 % were subjected to further statistical analysis. Characteristics of health behaviours and health status indicators among pregnant women in Slovenia 307 IzvIrnI znanStvenI članek degree, 41 %; Fisher exact test, p = 0.010) and among mothers who were older at conception (p = 0.015). There was no si- gnificant correlation with the year of de- livery (p = 0.218) or parity (p = 0.511). Intake of FDA class C medication was reported by 57 (13 %) participating mothers. Higher consumption of class C drugs was statistically significant in monoparous women (p = 0.001) and in women who delivered after year 2010 (p = 0.042). When adjusted for date of delivery, statistical significance per- sisted for parity (having 2 children vs. 1: OR(95 %CI): 2.2(1.1–4.6), p = 0.028), but significance did not persist for the year of delivery in the logistic regression model. This might be the result of the correlati- on between date of delivery and parity (Spearman rho = -0.456, p < 0.001). No significant association was found betwe- en class C usage and maternal education (p = 0.260) or maternal age at concepti- on (p = 0.431). Out of 450 participants 13 (3 %) re- ported intake of FDA class D (N = 12) or X (N = 1) drugs. None of the investigated maternal characteristics showed statisti- cally significant correlation with class D or X consumption during pregnancy, i.e. date of birth (p = 0.810), maternal edu- cation (p = 0.288), parity (p = 0.599) and maternal age at conception (p = 0.187). 3.3. Statistically significant correlations between dependant variables Mothers, who reported medication consumption during pregnancy, also had a higher percentage of folate and multi- vitamin supplementation (p = 0.005). As expected, mothers with chronic diseases reported higher percentage of drug usa- ge during pregnancy (p = 3.14 × 10-4). A higher percentage of A12C medication usage was found in mothers who repor- ted fever during pregnancy (p = 0.026) and gestational diabetes (p = 0.034). Mothers with gestational diabetes also reported higher folate and multivitamin supplementation (p = 0.034). 4. Discussion The purpose of this study was to as- sess the health-related behaviours and health status of pregnant women and to discover associations among them. This study contributes a comparative descrip- tive data analysis of self-reported health status, medication usage, health behavio- urs and socio-demographics of a cohort of women from Slovenia. It provides an insights into potential determinants for suboptimal health behaviours, such as smoking, and yields several findings that may be important for clinical practice. Mothers with a lower level of edu- cation and a lower age at conception were significantly more likely to smoke during pregnancy. After adjustment in the logistic regression model, the asso- ciation remained statistically significant only for education but not for the age at conception, which is most likely because of correlation present between mater- nal educational level and maternal age at conception (Spearman rho = 0.420, p = 1.6 × 10-19). The correlation between lower maternal education and lower ma- ternal age is expected because women increasingly delay childbirth in order to receive more education and take advan- tage of increased career opportunities created thereby  (22). Our findings are in line with previous studies, which re- ported that smoking, including passive smoking, was the health behaviour most strongly associated with educational di- sparities, finding a higher percentage of smokers in groups of pregnant women with lower levels of education (19,23,24). Study Lu Y et al.  (2001) also reported 308 zdrav vestn | julij – avgust 2017 | letnik 86 reProdukCIja človeka that smokers with higher levels of edu- cation were more likely to stop smoking when they discovered they were pre- gnant, leading to even greater disparities in smoking (19,25). According to a cross- -sectional study of pregnant women and new mothers from 15 European co- untries, which also included data from Slovenia, prevalence of smoking during pregnancy ranged from 18.9 % in Croa- tia to 4.2 % in Iceland (9). Slovenia had 6.7 % of prevalence of smoking during pregnancy, which is almost two-fold less than reported in our study (12 % (95 %CI: 9–15 %)). Since our data covers pregnan- cy characteristics from the 1980s to 2015, while the previously cited comparative study only covers the period 2011–2012, the difference is in line with a recent decline of prevalence rates of smoking during pregnancy in western countries from 20–30 % in the 1980s to 10–20 % in the early 2000s  (9,26,27). Indeed, when looking at the part of our study cohort, mothers that gave birth in the same time period as the cohort of Smedberg J et al. (9), we found that the percent of smo- king pregnant women in this subgroup (9 % (95 %CI: 3–20 %)) was comparable to the one reported by the above-men- tioned study (i.e. 6.7 %). A younger age at conception, in addition to being cor- related with lower education, may also act as individual risk factor. The afore- mentioned multi-national study  (9) fo- und that pregnant women who were 20 years of age or younger were more likely to smoke before and during pregnancy. This may be explained by the menacing shift in tobacco use in women, driven by marketing directed specifically at youn- ger women (9). Our findings are a rem- inder that smoking is still prevalent du- ring pregnancy. Studies strongly suggest association of smoking during pregnan- cy with increased risk of spontaneous miscarriage, preterm delivery, low birth weight, premature rupture of membra- nes, placental abruption, placenta previa, stillbirth, higher percentage of repeated unintentional injuries in children and possible increase in pain perception le- vels of the newborns  (9,28,29). In view of these deleterious effects of smoking during pregnancy, precautions to pre- vent exposure ought to be an important public health goal. There was no statistically significant association between the maternal cha- racteristics, which were investigated and the incidence of fever during pregnan- cy, except for a higher consumption of A12C medications. An increase in ma- ternal body temperature is a well-reco- gnized cause of birth defects (30). A12C medications are a group of mineral su- pplements; in the study data the majo- rity were Mg citrate supplements. Ma- gnesium is known antipyretic; since the work of Schutz (1916), it has been known that magnesium injected directly into the brain substance (tuberal area) may lower body temperature  (31). This fact is somewhat contradicted by the hig- her percentage of Mg supplementation correlated to the group of women with fever during pregnancy. However, with respect to the health outcomes of these pregnancies, Mg supplementation might have a protective effect, lowering fever and diminishing the detrimental effect of fever on the foetus. This effect could be due to the role of Magnesium in pro- tein synthesis and in the stability of the cell membrane and genome (32). Mothers with more than two children had a higher incidence of chronic dise- ases; the statistical significance of this association persisted after adjustment for maternal age at conception. During pregnancy, the maternal body undergoes significant anatomical and physiological changes in order to nurture and accom- modate the developing foetus (14). These Characteristics of health behaviours and health status indicators among pregnant women in Slovenia 309 IzvIrnI znanStvenI članek changes represent a physiological chal- lenge as they impose stress to the orga- nism as a whole. Increase in oxygen de- mands, even during a normal pregnancy, is associated with an increased genera- tion of reactive oxygen species, which have been linked to various diseases (33). Furthermore, some pre-existing chronic diseases are known to worsen during pregnancy. Studies found that the rate of hospitalizations for asthma is higher among pregnant women than non-pre- gnant women, after adjusting for age and exacerbation history  (34,35). It is also recognised that even women with mild asthma are at increased risk of severe exacerbation during pregnancy (34). Our findings suggest that pregnancy somew- hat increases pre-existing susceptibility to diseases. Additionally, higher frequen- cy of contacts with health providers le- ads to the diagnosis rate of unrecognized health conditions. Up to 70 % of late-on- set adult chronic diseases share their risk factors with those in pregnancy (36). For example, occurrence of gestational di- abetes is linked to an increased risk of chronic diabetes mellitus type 2 in later life (37). Since over 80 % of women have at least one child, pregnancy provides a single common, widespread opportunity to profit from an increased and sustain- ed contact with health care providers to affect behaviour change and thus benefit in terms of both pregnancy outcomes and long-term chronic disease risks for mother and child (36). Gestational diabetes is diabetes with first onset or first recognition during pregnancy. We observed a trend of hig- her GD frequency among older women and women who delivered after 2010. According to Slovenian guidelines adop- ted in 2011 by the national health system, every pregnant woman’s blood glucose must be checked at least twice during pregnancy  (38). The first routine blood glucose test is at the first examination, when pregnancy is confirmed. The se- cond test performed in a normally pro- gressing pregnancy takes place between the 24th and 28th week in the form of a 75 g OGTT test (oral glucose tolerance test). As a result of the new guidelines, the population of pregnant women is tested far more thoroughly than in the past. Because of the time span of our study data acquisition (1980–2015) some earlier GD cases might have been undi- agnosed, particularly those occurring in 1980s. Even though epidemiological stu- dies have reported a rising trend of GD, these studies must be interpreted with caution: the rising trends of GD may be due to better diagnostics (39-42). Among women who reported having GD during pregnancy we observed higher percenta- ge of A12C drugs (in majority these were magnesium preparations) and folate/ multivitamin supplement consumption, most likely in pursuit of minimising de- trimental effects of GD with health pro- moting behaviour. Higher incidence of folate and mul- tivitamin supplementation was obser- ved in women who gave birth after 2000, mothers with higher levels of education, nulliparous women and those who were older at conception. Significance per- sisted after adjustment in the logistic re- gression model for all former instances except for age at conception, presuma- bly because of correlation with date of birth (Spearman rho = 0.370, p < 0.001) and maternal education (Spearman rho = 0.420, p < 0.001). Since 1993 daily supplementation with 400 μg folic acid at least 3 months prior and after concep- tion is recommended in Slovenia by the public health service  (43). We observed a steady increase in the consumption of multivitamin and folate supplementati- on with time: in the 1980s (18 %(95 %CI: 4–47 %)), in the 1990s (26 %(95 %CI: 14– 310 zdrav vestn | julij – avgust 2017 | letnik 86 reProdukCIja človeka 41 %)), with highest increase after 2000 (73 %(95 %CI: 64–80 %)) that further amplified after 2010 (91 %(95 %CI: 87– 94 %)). This might be explained by hig- her supplementation literacy among he- alth practitioners and pregnant women that gradually increased in the decade after recommendations were formula- ted because of the abundance of medical and commercial attention that folic acid garnered. Our findings are in line with numerous studies that reported higher prevalence of folate and multivitamin supplementation among women with higher levels of education  (5,44,45,46), nulliparous women  (47) and among women who were older at concepti- on  (47). Higher maternal education is associated with higher health literacy, health control beliefs and risk adverse behaviour, which corresponds to peri- conceptional supplementation (9,19,48). The study by Baron et al. (2015) reported sizable disparities in unbeneficial health status and behaviours according to edu- cational level, in that women with lower education have low health control beli- efs, higher incidences of both obesity and being underweight, higher incidence of depression/anxiety, higher incidence of smoking and passive smoking exposure, lower antenatal attendance, lower folic acid supplementation, higher incidence of skipping breakfast daily and unplan- ned pregnancies and lower incidence of daily fruit consumption  (19). Nulli- parous women are perhaps more anxi- ous about pregnancy outcome and try to minimise possible risks by following recommendations on health promoting behaviour. In contrast, women’s personal experience of delivering a healthy child despite not taking folate/multivitamin supplementation might create distrust in the scientific recommendations, which results in the refusal of folate/multivita- min supplementation in succeeding pre- gnancies (9). Only 8 % (95 %CI: 6–12 %) of women reported the consumption of nutritional supplements with active form of folic acid – 5-methylfolate. Interestingly, higher incidence of me- dicine (OTC and prescribed) consump- tion during pregnancy was reported by women with the highest and lowest levels of education, by women who de- livered in recent years (1980s 0 % vs. 2011–2015 48 %(95 %CI:42–54 %)) and by those with higher age at conception. The latter was not significant after ad- justment with date of delivery and ma- ternal education. A 2011 study using U.S. data from 1976–2008 reported that most women (88.8 %) take at least one medi- cation during pregnancy and 70 % take at least one prescribed medication (49). A 2014 multi-national European study on data from 2011–2012 reported that 81.2 % of women take at least one medi- cation (OTC or prescribed) during pre- gnancy and 68.4 % and 17 % of women reported the use of prescribed medica- tion for acute and chronic conditions, respectively  (5). Data of our study co- hort revealed a slightly lower prevalence of medication usage (40 %(95 %CI:35– 44 %)) but the trend of increasing con- sumption is in line with an American study which observed that over the last three decades, use of prescription drugs during pregnancy has increased by over 60 % (49). There are several studies that observed an association between higher maternal education and more preva- lent use of medication during pregnan- cy (49,50,51,52). Contrary to these, others reported more prevalent use of medicati- ons among pregnant women with lower education (5,53,54). It appears as though our study cohort with an U-shaped di- stribution fits with the recent findings. The aforementioned European study lin- ked the higher prevalence of drug usage among lower educated women to a hig- Characteristics of health behaviours and health status indicators among pregnant women in Slovenia 311 IzvIrnI znanStvenI članek her percentage of medications for chro- nic/long term conditions  (5). Since we did not find such an association, higher consumption of medication observed in groups of women with the lowest and highest levels of education might reflect higher confidence in self-treatment and use of medications in general  (5). Risk perceptions are lower in both groups for different reasons. Educated women have the highest health literacy while the least educated have the weakest health control beliefs (1). Furthermore, among women who took medication during pregnancy, there was a higher prevalence of folate and multivitamin supplementation. This could be explained with the correlation observed between multivitamin/fola- te supplementation and maternal cha- racteristics that influence drug intake, namely the year of delivery (Spearman rho = 0.458, p < 0.001), level of education (Spearman rho = 0.325, p < 0.001) and maternal age at conception (Spearman rho = 0.247, p < 0.001). Unsurprisingly, there was higher medication intake re- ported by women with chronic health conditions. Reported medications were classified according to ATC and FDA classes. We analysed data for possible associations with maternal characteristics and for those medication groups that were taken most frequently (> 5 % abundance). A12C–mineral supplement medica- tion intake (6 %(95 %CI:4–9 %)) was as- sociated with a higher level of education and nulliparity. Women who reported A12C supplementation gave birth after 2010. In our study, A12C medications were mostly magnesium supplements (Mg citrate). Magnesium is an essential mineral required for maintaining nerve and muscle cell electrical potential, regu- lation of body temperature and synthe- sis of nucleic acids and proteins  (32). Magnesium supplementation during pregnancy may be able to reduce foetal growth restriction, low birthweight and pre-eclampsia  (55). Similar distribution of consumption was found for folate/ multivitamin supplementation, and hen- ce we could explain it in a similar frame; better health literacy and more vigilant health promoting behaviour among the- se groups. In addition, there might be some recall bias considering time dis- tribution. Education was also associated with better recall accuracy of medicati- ons in most studies investigating this re- lation (57-61). The most recent Cochrane review (2014) of Mg supplementation in pregnancy concluded that there is not enough high-quality evidence to show that dietary magnesium supplementa- tion during pregnancy is beneficial for improving maternal and neonatal/in- fant health outcomes. It is further sta- ted that until additional evidence from large, well designed, randomised trials becomes available, current evidence is insufficient to make recommendations for routine clinical practice on the use of magnesium supplementation during pregnancy (55). B03A – Iron preparations were the most prevalent group of medications (17 %(95 %CI:14–21 %)) in our study co- hort. The only significant association was with increasing consumption over the studied three decades (1980s 0 %, 2011–2015 22 %(95 %CI:17–27 %)). Ac- cording to WHO recommendations, daily oral iron (and folic acid supple- mentation) is recommended as part of the antenatal care to reduce the risk of low birth weight, maternal anaemia and iron deficiency. The latest Cochrane re- view  (2015) on daily iron supplementa- tion concluded that supplementation re- duces risk of maternal anaemia and iron deficiency in pregnancy, but the positive effect on maternal and infant outcomes is less clear. The implementation of iron 312 zdrav vestn | julij – avgust 2017 | letnik 86 reProdukCIja človeka supplementation recommendations may produce heterogeneous results depen- ding on the populations’ background risk for low birthweight and anaemia, as well as the level of adherence to the intervention  (56). According to the UK guidelines on management of iron de- ficiency during pregnancy women are encouraged to maximise iron dietary in- take and absorption, though routine iron supplementation for all women in pre- gnancy is not recommended. There are clinical hazards of routine supplementa- tion, such as risk of elevated Hb and oxi- dative stress (62,63,64). Elevated Hb was associated with higher rates of perinatal deaths, low birth weight and preterm delivery  (63), and markers of oxidative stress were found to be elevated in the placenta of women with routine iron supplementation  (64). In Slovenian he- alth practice, there is an individualised approach to iron supplementation based on concentration of blood Hb, with Hb levels less than 110 g/L considered too low in pregnant women and warranting supplementation. A possible explanation for the increase in iron supplementation in the last three decades lies in wider so- cietal dietary trends. The increase in iron supplementation runs parallel with an increase in the consumption of starchy, processed foods with low nutritional content. This might explain the greater incidence of iron supplementation in pregnancy. Nonetheless, higher recall certainty for recent pregnancies might play a role. Intake of N02B (mostly paracetamol and acetylsalicylic acid) was reported by 7 %(95 %CI:5–9 %) of the participants. Distribution of consumption was U- -shaped according to maternal educati- on. The highest usage was reported by women with the highest and lowest le- vels of education and by women with a higher age at conception. The same cha- racteristics were observed in relation to reported intake of all medications. Sin- ce we did not observe any difference in reported health status among groups of different levels of education and age at conception, the most likely explanation lies in the risk perceptions of the sub- -groups. In a recent study on women’s perception of risks of adverse foetal pre- gnancy outcomes, risk perceptions were lower with increasing age and with hig- her level of education, which may reflect increased access to information, perso- nal experience and thus higher confiden- ce in medication intake (1). In the same study individuals with lower education perceived thalidomide (a well-known te- ratogen) to be less risky than did women with higher education, which could indi- cate an increased confidence in self-drug administration among those with lower education (1). Medications were also classified ac- cording to the FDA classes. The hig- hest prevalence of consumption was reported for FDA class A medicati- ons (19 %(95 %CI:16–23 %)) for which adequate and well-controlled studies fai- led to demonstrate a risk to the foetus. In our study cohort these were mostly iron supplements (17.3 %). The only signifi- cant association was with date of delive- ry: the highest prevalence of intake was reported by women who delivered after 2011. This correlates with results for iron preparations. Intake of FDA class B medications was reported by 13 %(95 %CI:10–17 %) of par- ticipants. Class B medications are those medications for which animal reproduc- tion studies have failed to demonstrate a risk to the foetus, however there are no adequate and well-controlled studies in pregnant women available to fully define the effects of the medication. The highest consumption was found among higher educated mothers and those who were Characteristics of health behaviours and health status indicators among pregnant women in Slovenia 313 IzvIrnI znanStvenI članek older at conception. Prevalent class B drugs in our study cohort were paraceta- mol and antibiotics (mostly amoxicillin). The results unsurprisingly in part corre- late with characteristics of N02B con- sumption, since paracetamol is a major representative of both groups. The only differences are seen in the distribution according to education, in which the as- sociation between intake and education is not U-shaped as in the case of N02B. This difference may be attributed to the variability in antibiotic consumption, which would be in-line with results of a Finnish study that found higher preva- lence of antibiotic consumption among women in older maternal age groups and among women belonging to a hig- her socio-economic group – having hig- her level of education (65). FDA class C medication intake was reported by 13 %(95 %CI:10–16 %) of participating women. Class C consists of medications for which animal reproduc- tion studies have shown an adverse effect on the foetus and there are no adequate and well-controlled studies in humans, but potential benefits may permit use of the drug in pregnant women despite potential risks. Nulliparous women and those who gave birth after 2010 had the highest percentage of class C drug inta- ke. After adjusting in logistic regression, the association between lower parity and higher intake persisted, but we obser- ved a correlation between date of deli- very and parity (Spearman rho = -0.456, p < 0.001). Since a major representative of class C and A12C medications in our study data was magnesium citrate, gro- ups share similar characteristics, i.e. an increasing trend of consumption and null parity. We found no associations between the researched maternal characteristics and consumption of class D and X me- dications. For class D medications the- re is positive evidence of human foetal risk based on adverse reaction data from investigational or marketing experience or studies in humans, but potential be- nefits may warrant the use of the drug in pregnant women despite potential risks. Class X medications have demon- strated foetal abnormalities in animal or human studies, and/or there is positive evidence of human foetal risk based on adverse reaction data from investigati- onal or marketing experience, and the risks associated with the use of the drug in pregnant women clearly outweigh po- tential benefits. Most likely no associati- ons were found due to the small number of women who reported intake of these drug classes (3 %(95 %CI:2–5 %)). Study results must be interpreted in the view of the following limitations. The number of participants was relati- vely small. The study was retrospective in design and there might be some in- complete data because all researched va- riables are dependent on women’s recall, accuracy and perception. The main limi- tation of the study is that the reliability of self-reports is lower in the mothers of healthy siblings of OFC/CHD cases, who were interviewed at different time-points after the corresponding pregnancy, com- pared to the mothers of healthy new- borns who were interviewed immedia- tely after the delivery. The comparison of the two subgroups by Fisher`s exact test revealed that they differ in almost all of the outcome characteristics, except for smoking in pregnancy and fever du- ring pregnancy. However, it is important to note that all of the healthy newborns were born between 2011 and 2015, while only 27 % of the healthy siblings of OFC/ CHD cases were born in the same pe- 314 zdrav vestn | julij – avgust 2017 | letnik 86 reProdukCIja človeka riod (p < 0.001). When comparing the healthy newborns and healthy siblin- gs of OFC/CHD cases that were born in the same period  (2011–2015), none of the outcome characteristics differed among the two groups, except for mater- nal chronic disease which was still more prevalent in mothers of healthy siblings of OFC/CHD cases than in the mothers of healthy newborns (p = 0.029). This indicates that differences in outcome characteristics between the two subgro- ups are due to the different delivery pe- riods and not due to the selection bias related to the family anamnesis of OFC or CHD. The only analysis that might be biased due to the cohort heterogeneity is the chronic disease of the mother, which was indicated in some studies as one of the risk factors for OFC and CHD. Self- -report bias/underreporting cannot be ruled out in the case of smoking and possible drug consumption due to the social stigma that is attached to these exposures. On the other hand, health- -promoting behaviours might have been over-reported. Some selection bias ca- nnot be excluded, as some of participa- ting mothers had previous adverse pre- gnancy outcome (CHD or OFC), and thus their behaviour during subsequent pregnancies might have been more alert to adverse birth effects, which may not be representative for general population of childbearing women. However, this bias was absent in the majority of study cases, as 73 % of healthy siblings of OFC and CHD cases were born prior to their sick sibling. Study did not include pos- sibly important factors such as exposure to passive smoking, income and emplo- yment in order to fully assess smoking and socio-economic status. 5. Conclusions The study found considerable dispari- ties in health conducts and socio-demo- graphics of women during pregnancy. It revealed that level of education, parity and age have an impact on women’s he- alth behaviour, namely medications and micronutrient supplementation intake and smoking status. Younger and less educated women more frequently reported positive smo- king status during pregnancy while hi- gher prevalence of folate/multivitamin supplementation was found among more educated, older and nulliparous women. There was a U-shaped distribu- tion of higher prevalence of medicati- on intake (OTC and prescribed) in the groups of women with the highest and lowest levels of education. Higher usage was also reported among older women. We observed an increasing trend of me- dication, folate/multivitamin intake and incidence of GD over the studied peri- od. Higher incidence of chronic diseases was observed in a group of multiparous women. Identification of disadvantaged women in terms of health-related risky behaviour during pregnancy provides an opportunity to assist women in reducing major health hazards and to identify the key factors of a healthy pregnancy. Public health campaigns are needed that would educate and provide healthy lifestyle choices with advice on diet, exercise, al- cohol, smoking, medications and folic/ multivitamin supplementation. Further qualitative research is needed to explore in-depth the dilemmas that women and their health care providers face during pregnancy (2,7). Characteristics of health behaviours and health status indicators among pregnant women in Slovenia 315 IzvIrnI znanStvenI članek References 1. Petersen I, McCrea RL, Lupattelli A, Nordeng H. Women’s perception of risk of adverse fetal pre- gnancy outcomes: a large-scale multinational stu- dy. BML Open. 2015;5(6):e007390. 2. McDonald K, Amir LH, Davey MA. Maternal bo- dies and medicines: a commentary on risk and decision-making of pregnant and breastfeeding women and health professionals. MNC Public He- alth. 2011;11 Suppl 5:S5. 3. Honein MA, Gilboa SM, Broussard CS. The need for safer medication use in pregnancy. Expert Rev Clin Pharmacol. 2013;6(5):453–5. 4. Thorpe PG, Gilboa SM, Hernandez-Diaz S, Lind J, Cragan JD, Briggs G, et al. Medications in the first trimester of pregnancy: most common exposures and critical gaps in understanding fetal risk. Phar- macoepidemiol Drug Saf. 2013;22(9):1013–8. 5. Lupattelli A, Spigset O, Twigg MJ, Zagorodnikova K, Mårdby AC, Moretti ME, et al. Medicaion use in pregnancy: a cross-sectional, multinational web- -based study. BMJ Open. 2014;4(2):e004365. 6. Servey J, Chang J. Over-the-counter medications in pregnancy. Am Fam Physician. 2014;90(8):548–55. 7. Mullally A, Cleary BJ, Barry J, Fahey TP, Murphy DJ. Prevalence, predictors and perinatal outcomes of peri-conceptonal alcohol exposure-retrospecti- ve cohort study in an urban obstetric population in Ireland. BMC Pregnancy Childbirth. 2011;11:27. 8. Lassi ZS, Imam AM, Dean SV, Bhutta ZA. Peri- conceptional care:caffeine, smoking, alcohol, dru- gs and other environmental chemical/radiation exposure. Reprod Health. 2014;11 Suppl 3:S6. 9. Smedberg J, Lupattelli A, Mårdby AC, Nordeng H. Characteristics of women who continue smoking during pregnancy: a cross-sectional study of pre- gnant women and new mothers in 15 European countries. BMC Pregnancy Childbirth. 2014;14:213. 10. European Medicines Agency. Guideline on the Development of Medicinal Products for the Tre- atment of Smoking. London: European Medicines Agency; 2008. 11. Raymond N, Beer C, Glazebrook C, Sayal K. Pre- gnant women’s attitudes towards alcohol con- sumption. BMC Public Health. 2009;9:175. 12. Khoushabi F, Shadan MR, Miri A, Sharifi-Rad J. Determination of maternal serum zinc, iron, cal- cium and magnesium during pregnancy in pre- gnant women and umbilical cord blood and their association with outcome of pregnancy. Mater So- ciomed. 2016;28(2):104–7. 13. Masoumi SZ, Parsa P, Kazemi F, Soltanian AR, Ha- bib DGS. Investigation of Nutritional Behaviors in the First and Second Trimesters in Pregnant Women Referring to Clinics in Hamadan, Iran, in 2013. Glob J Health Sci. 2016;8(9):261–70. 14. Soma-Pillay P, Nelson-Piercy, Tolppanen H, Me- bazaa A. Physiological changes in pregnancy. Car- diovasc J Afr. 2016;27(2):89–94. 15. Murthy GVS, Kolli SR, Neogi SB, Singh S, Allagh KP, John N, et al. Mixed-Method Study to Deter- mine the Benefits of Periconceptional Folic Acid Supplementation And Effects of Folic Acid DefI- ciency in Mothers on Birth Outcomes. JMIR Res Protoc 2016;5(2):e129. 16. Shaw GM, Nelson V, Carmichael S, Lammer E, Finnell R, Rosenquist T. Maternal Periconcep- tional Vitamins: Interactions with Selected Fac- tors and Congenital Anomalies? Epidemiology. 2002;13(6):625-30. 17. Carmichael SL, Nelson V, Shaw GM, Wasserman CR, Croen LA. Socio-economic status and risk of conotruncal heart defects and orofacial clefts. Pae- diatr Perinat Epidemiol. 2003;17(3):264-71. 18. Charugundla P. The role of maternal nutrition, risk factor avoidance and gene-environment inte- ractions in orofacial clefting. A master of science in Environmental health Sciences thesis. Los An- geles: University of California; 2013. 19. Baron R, Manniën J, te Velde SJ, Klomp T, Hutton EK, Brug J. Socio-demographic inequalities across a range of health status indicators and health be- haviours among pregnant women in prenatal pri- mary care: a cross-sectional study. BMC Pregnan- cy Childbirth. 2015;15:261. 20. Krajec M, Natek N, Geršak K. Medication use in the postpartum period in Slovenia with regard to breastfeeding. Zdrav Vestn. 2016;85(9):483-90. 21. Van Calsteren K, Geršak K, Sundseth H, Klin- gmann I, Dewulf L, Van Assche A, Mahmood T. Position statement from the European Board and College of Obstetrics & Gynaecology (EB- COG): The use of medicines during pregnancy– call for action. Eur J Obstet Gynecol Reprod Biol. 2016;201:189–91. 22. Khalil A, Syngelaki A, Maiz N, Zinevich Y, Nicolai- des KH. Maternal age and adverse pregnancy out- come: a cohort study. Ultrasound Obstet Gynecol. 2013;42(6):634–43. 23. Lanting CI, van Wouwe JP, van den Burg I, Segaar D, van der Pal-de Bruin KM. Smoking during pre- gnancy: trends between 2001 and 2010. Ned Tijd- schr Geneeskd. 2012;156(46):A5092. 24. Aurrekoetxea JJ, Murcia M, Rebagliato M, Fer- nandez-Somoano A, Castilla AM, Guxens M, et al. Factors associated with second-hand smoke expo- sure in non-smoking pregnant women in Spain: self-reported exposure and urinary cotinine levels. Sci Total Environ. 2014;470–471:1189–96. 25. Lu Y, Tong S, Oldenburg B. Determinants of smo- king and cessation during and after pregnancy. Health Promot Int. 2001;16(4):355–65. 26. Cnattingius S. The epidemiology of smoking du- ring pregnancy: smoking prevalence, maternal characteristics, and pregnancy outcomes. Nicotine Tob Res. 2004;6 Suppl 2:S125-S40. 27. Al-Sahab B, Saqib M, Hauser G, Tamim H. Preva- lence of smoking during pregnancy and associated risk factors among Canadian women: a national survey. BMC Pregnancy Childbirth. 2010;10:24. 28. Junger M, Japel C, Coté S, Xu Q, Boivin M, & Tremblay R. Smoking and Medication During Pregnancy Predict Repeated Unintentional Injuri- es in Early Childhood But Not Single Unintentio- nal Injuries. Prev Sci. 2013;14;13–24. 316 zdrav vestn | julij – avgust 2017 | letnik 86 reProdukCIja človeka 29. Tekin M, Yıldırım S, Aylanç H, Kaymaz N, Battal F, Topaloğlu N, et al. Does intrauterine tobacco exposure increase the pain perception of new- borns? J Pain Res. 2016;9:319–23. 30. Vieira AR, Orioli IM. Birth order and oral clefts: a meta analysis. Teratology. 2002;66:209–16. 31. Myers RD, Veale WLJ. The role of sodium an cal- cium ions in the hypothalamus in the control of body temperature of the unanaesthetized cat. Physiol. 1971;212(2):411-30. 32. Krapels IP, van Rooij IA, Ocke MC, West CE, van der Horst CM, Steegers-Theunissen RP. Maternal nutritional status and the risk for orofacial cleft offspring in humans. J Nutr. 2004;134(11):3106–13. 33. De Lucca L, Rodrigues F, Jantsch LB, Neme WS, Gallarreta FMP, Gonçalves TL. Oxidative Pro- file and δ-Aminolevulinate Dehydratase Acti- vity in Healthy Pregnant Women with Iron Su- pplementation. Int J Environ Res Public Health 2016;13(5):463. 34. Murphy VE. Managing asthma in pregnancy. Bre- athe. 2015;11:258–67. 35. Kim S, Kim J, Park SY, Um HY, Kim K, Kim Y, et al. Effect of pregnancy in asthma on health care use and perinatal outcomes. J Allergy Clin Immunol. 2015;136(5):1215–23. 36. O’Keeffe LM, Dahly D L, Murphy M, Greene RA, Harrington JM, Corcoran P, et al. Positive lifestyle changes around the time of pregnancy: a cross- -sectional study. BMJ Open. 2016;6(5):e010233. 37. Kim SY, Kotelchuck M, Wilson HG, Diop H, Sha- piro-Mendoza CK,England LJ. Prevalence of ad- verse pregnancy outcomes, by maternal diabetes status at first and second deliveries, Massachusetts, 1998–2007. Prev Chronic Dis. 2015;12:E218 38. Tercelj M, Sketelj M. Recommendations for gesta- tional diabetes management in Slovenian diabe- tes management guidelines[Internet]. Ljubljana: University Medical Centre Ljubljana; 2011 [cited 2016 Nov 13]. Available from: http//endodiab.si/ priprocila/smernice-za-vodenje-sladkorne-bole- zni/ 39. Ferrara A, Kahn H, Quesenberry C, Riley C, Hed- derson M. An Increase in the Incidence of Ge- stational Diabetes Mellitus: Northern California, 1991–2000. Obstet Gynecol. 2004;103(3):526-33. 40. Hunt KJ, Schuller KL. The Increasing Prevalence of Diabetes in Pregnancy. Obstet Gynecol Clin North Am. 2007;34(2):173–8. 41. Paulin S, Kelšin N, Korošec A, Zaletel J, Nadrag P, Zaletel M. Ekonomsko breme sladkorne bolezni v Sloveniji 2012. Ljubljana: Nacionalni inštitut za javno zdravje; 2014. 42. Berglund SK, García-Valdés L, Torres-Espinola FJ, Segura T, Martínez-Zaldívar C, Aguilar MJ, et al. Maternal, fetal and perinatal alterations associated with obesity, overweight and gestational diabetes: an observational cohort study (PREOBE). BMC Public Health. 2016;16:207. 43. Kokalj TS, Rejc B, Geršak K. Incidence and pre- vention of neural tube defects in Slovenia. Eur J Obstet Gynecol Reprod Biol. 2011;156(1):119–20. 44. Baron R, Manniën J, de Jonge A, Heymans MW, Klomp T, Hutton EK, Brug J. Socio-Demographic and Lifestyle-Related Characteristics Associa- ted with Self-Reported Any, Daily and Occasi- onal Smoking during Pregnancy. PLoS ONE. 2013;8(9):e74197. 45. Dean SV, Lassi ZS, Imam AM, Bhutta ZA. Precon- ception care: nutritional risks and interventions. Reprod Health. 2014;26;11 Suppl 3:S3. 46. Zeitlin J, Mortensen L, Prunet C, Macfarlane A, Hindori-Mohangoo AD, Gissler M, et al. Socioe- conomic inequalities in stillbirth rates in Europe: measuring the gap using routine data from the Euro-Peristat Project. BMC Pregnancy Childbirth. 2016;16:15. 47. Verstappen GMPJ, Smolders LJ, Munster JM, Aar- noudse JG, Hak E. Prevalence and predictors of over-the-counter medication use among pregnant women: a cross-sectional study in the Netherlan- ds. BMC Public Health. 2013;13:185. 48. Onah MN, Field S, van Heyningen T, Honikman S. Predictors of alcohol and other drug use among pregnant women in a peri-urban South African setting. Int J Ment Health Syst. 2016;10:38. 49. Mitchell AA, Gilboa SM, Werler MM, Kelley KE, Louik C, Hernández-Díaz S. National Birth De- fects Prevention Study. Medication use during pregnancy, with particular focus on prescrip- tion drugs: 1976–2008. Am J Obstet Gynecol. 2011;205(1):51.e1–8. 50. Odalovic M, Vezmar Kovacevic S, Nordeng H, Ilic K, Sabo A, Tasic L. Predictors of the use of me- dications before and during pregnancy. Int J Clin Pharm. 2013;35(3):408–16. 51. Refuerzo JS, Blackwell SC, Sokol RJ, Lajeunesse L, Firchau K, Kruger M, Sorokin Y. Use of over-the- -counter medications and herbal remedies in pre- gnancy. Am J Perinatol. 2005;22(6):321–324. 52. Donati S, Baglio G, Spinelli A, Grandolfo ME. Drug use in pregnancy among Italian women. Eur J Clin Pharmacol. 2000;56(4):323–8. 53. Olesen C, Thrane N, Henriksen TB, Ehrenstein V, Olsen J. Associations between socio-economic factors and the use of prescription medication du- ring pregnancy: A population-based study among 19,874 Danish women. Eur J Clin Pharmacol. 2006;62(7):547-53. 54. Stokholm J, Schjørring S, Pedersen L, Bischoff AL, Følsgaard N, Carson CG, et al. Prevalence and Pre- dictors of Antibiotic Administration during Pre- gnancy and Birth. PLoS ONE. 2013;8(12):e82932. 55. Makrides M, Crosby DD, Bain E, Crowther CA. Magnesium supplementation in pregnancy. Co- chrane Database of Systematic Reviews 2014; CD000937. 56. Peña-Rosas JP, De-Regil LM, Garcia-Casal MN, Dowswell T. Daily oral iron supplementation du- ring pregnancy. Cochrane Database of Systematic Reviews 2015; CD004736. 57. Radin RG, Mitchell AA, Werler MM. Predictors of recall certainty of dates of analgesic medication use in pregnancy. Pharmacoepidemiol Drug Saf. 2013;22(1):25–32. 58. deJong PCMP, Berns MPH, Duynhoven YTHPv, Nijdam WS, Eskes TKAB, Zielhuis GA. Recall of medication during pregnancy: Validity and accu- racy of an adjusted questionnaire. Pharmacoepi- demiol Drug Saf. 1995;4(1):23–30. Characteristics of health behaviours and health status indicators among pregnant women in Slovenia 317 IzvIrnI znanStvenI članek 59. Goodman MT, Nomura AMY, Wilkens LR, Kolo- nel LN. Agreement between interview information and physician records on history of menopausal estrogen use. Am J Epidemiol. 1990;131(5):815–25. 60. West SL, Savitz DA, Koch G, Sheff KL, Strom BL, Guess HA, Hartzema AG. Demographics, health behaviors, and past drug use as predictors of re- call accuracy for previous prescription medication use. J Clin Epidemiol. 1997;50(8):975–80. 61. Cotterchio M, Kreiger N, Darlington G, Steingart A. Comparison of self-reported and physicianre- ported antidepressant medication use. Annals Epidemiol. 1999;9(5):283–9. 62. Pavord S, Myers B, Robinson S, Allard S, Strong J, Oppenheimer C. UK guidelines on the manage- ment of iron deficiency in pregnancy. Br J Haema- tol. 2012;156(5):588-600. 63. Murphy JF, O’Riordan J, Newcombe RG, Coles EC, Pearson JF. Relation of haemoglobin levels in first and second trimesters to outcome of pregnancy. The Lancet. 1986;1(8488):992–5. 64. Devrim E, Tarhan I, Ergüder IB, Durak I. Oxi- dant/Antioxidant Status of Placenta, Blood, and Cord Blood Samples From Pregnant Women Su- pplemented With Iron. J Soc Gynecol Investig. 2006;13(7):502–5. 65. Artama M, Gissler M, Malm H, Ritvanen A. Natio- nwide register-based surveillance system on drugs and pregnancy in Finland 1996–2006. Pharma- coepidemiol Drug Saf. 2011;20(7):729–38