Background

In line with the age and gender adjusted general population prevalence [1], approximately one third of pregnant women in Australia are overweight (BMI 25-29.9) or obese (BMI > 30) [2, 3]. These rates are similar to the rates in other developed and develo** countries [4]. Arguably, overweight and obesity are currently among the most common risk factors for adverse pregnancy outcomes [5]. Table 1 provides an overview of the quoted prevalence and odds ratios for a number of pregnancy and neonatal complications for obese women compared to women of normal weight, derived from a detailed literature review in this area. These complications include gestational diabetes, hypertensive disorders, caesarean section, thromboembolism, perinatal infections and in the neonate high birth weight or macrosomia, higher rates of intensive care nursery admission, congenital anomalies, preterm delivery, stillbirth and perinatal death [1, 6–8, 1, 9–13, 1, 14–16, 1, 15, 17–40]. Obesity in pregnancy is therefore associated with greater direct costs of $ 2387 (CI: $1799-$3109; P < 0.0001) per pregnancy [41].

Table 1 Prevalence and odds ratios for pregnancy and birth complications
Full size table

Increasing women's knowledge of the short and long-term risks of obesity to both their own and their offspring's health is likely to be an important first step in preventing obesity in pregnancy. Indeed recommendations to improve preconception care emphasize the need to ensure that women of childbearing age understand factors that increase the risks of childbearing, including obesity [42]. Our study was designed to ascertain whether or not women in the general pregnant population were aware of the increased risks associated with obesity in pregnancy. Furthermore, we investigated whether or not the pre pregnancy BMI was associated with differences in risk perception for complications in obese women.

Methods

We developed a questionnaire and surveyed 412 consecutive unselected women in early pregnancy as previously reported [43]. These women were either attending a public antenatal "first visit" clinic (n = 255), or undergoing a routine private ultrasound evaluation prior to 20 weeks gestation (n = 157)[44]. 61.9% of study participants were cared for in the public sector, similar to previously published proportions from Queensland [45]. Pre pregnancy BMI was available for 368 women. Women completed the survey independently while waiting for appointments. A trained research midwife was present at all times, to assist if participants required clarification regarding any component of the survey. The response rate for the questions varied between 96 and 100%. Permission for this study was obtained from the Royal Brisbane and Women's Hospital Health Research and Ethics Committee.

Data collection

Participants were asked to rate their perception of the risk of a pre-specified list of seven maternal and neonatal complications for women who were 'very underweight', 'normal weight' and 'very obese'. For each complication and with each weight status women were asked to indicate level of risk using a 5 point Likert scale (very low risk, low risk, average risk, high risk, very high risk, in addition to a "don't know" option). The specific questions used are shown in Additional file 1.

Participants were also asked "If a very obese woman was able to lose weight before pregnancy, how do you think this would affect her risk of pregnancy and birth complications?" The same seven factors were rated on a 5 point Likert scale using the following descriptors: She would be at much lower risk, She would be at lower risk, There would be no change in risk, She would be at higher risk, She would be at much higher risk (see Additional file 1).

Definition of knowledge about the risks of being obese prior to pregnancy

We assessed the way in which women rated risk for each complication for a normal and very obese woman. For the purposes of more detailed analysis, we evaluated women's broad knowledge about the risks of pregnancy and birth complications associated with being very obese. To be categorized as having broad knowledge about the risks of being very obese, women needed to rate the overall risk of complications as high or very high, and had to identify that weight loss prior to pregnancy is associated with a lower or much lower overall risk of complications.

Factors associated with knowledge about the risks of being obese

A number of demographic and obstetric history questions were included in the questionnaire. We explored the univariable and independent (of all other factors considered) associations of characteristics that we a priori thought were likely to be associated with knowledge and that might be useful in determining which groups of women should be specifically targeted to increase knowledge. The factors considered in these analyses were: maternal age (categorized as < 25 years, 25-35 years, > 35 years), parity (categorized as nulliparous or multiparous), smoking during current pregnancy (yes versus no), personal income (categorized as > 40 000 or ≤ 40 000 AUD per year), obstetric care (classified as public or private), pregnancy planning (categorized as planned or unplanned), highest educational status (classified as < Year 12, completed Year 12 or completed a tertiary qualification), body mass index (BMI) prior to pregnancy derived from self-report of pre pregnancy weight and height (categorized as < 25 kg/m2 or ≥25 kg/m2 ), periconceptual folate supplementation (yes versus no), attendance at a pre pregnancy planning visit with a doctor (yes versus no), weight loss attempts prior to current pregnancy (yes versus no), previous history of pregnancy-induced hypertension (yes versus no), of gestational diabetes (yes versus no) and of neonatal morbidity or mortality (including low birth weight baby, preterm baby, baby with a birth defect, death of baby within 1 month, baby requiring special or intensive care nursery).

Statistical analysis

Differences between women with a pre pregnancy BMI < 25.0 or ≥25.0 were analysed by two-sided Χ2 tests. P < 0.05 was considered statistically significant. Logistic regression was used to assess the relationship between each explanatory variable and "knowledge" of the risks of being very obese prior to pregnancy. Continuous variables (maternal age, BMI) were explored both as continuous and categorical variables, to ensure that this did not have an important effect on any of the multiple logistic regression models. Variables with several categories (parity, personal income) were explored using the original multiple categories and the dichotomized variable presented here in the results, to ensure that this did not substantially alter any of the odds ratios presented here. Multivariable logistic regression was used to further investigate some of the positive associations that we found. All analyses were performed with the statistical software package STATA v11.0.

Results

The baseline characteristics of the participating women are presented in Table 2. There was no difference in the baseline characteristics between women with a pre pregnancy BMI of < 25.0 and those with ≥ 25.0, except for BMI itself. Participants were asked to rate risk for a normal weight and a very obese woman for a variety of pregnancy and birth outcomes. These results are stratified by BMI < 25.0 or ≥ 25.0 and presented in Figure 1 and Additional file 2. There were no statistically significant differences in the responses from the women in the two BMI categories; all rated the risk for adverse pregnancy and birth outcomes higher for a very obese woman. In general, women were more confident of the effect of obesity on maternal than neonatal outcomes with 9.0-16.8% and 20.7-22.3% responding "Don't Know" in the BMI < 25.0 and 4.5-13.6% and 14.5-19.2% in the BMI ≥ 25.0 group for the different maternal and neonatal outcomes respectively. A majority of women rated the risks for a very obese woman of overall complications (74.6% vs. 71.6%), gestational diabetes (87.8% vs. 86.5%), blood pressure problems (88.2% vs. 88.3%) or caesarean section (53.6% vs. 50.7%) as high to very high, whereas the risk of preterm delivery (62.8% vs. 60.9%), admission to special nursery care (63.9% vs. 59.1%) and congenital anomalies (58.0% vs. 62.7%) were rated as average to high in the BMI < 25.0 vs. BMI ≥ 25.0 groups respectively.

Table 2 Participant demographic characteristics.
Full size table
Figure 1
figure 1

Risks for maternal and infant complications for a normal weight woman or a very obese woman respectively as assessed by pregnant women with a pre pregnancy BMI < 25.0 (white box and light grey box respectively) or BMI ≥ 25.0 (black box and dark grey box respectively) on a five point Likert scale. Results are expressed as mean ± SD. N = 354 for women with BMI < 25.0 and 111 for women with BMI ≥ 25.0. ***, P < 0.001 between the risks for a normal weight woman and an obese woman. There were no statistically significant differences between the assessments of women with a pre pregnancy BMI < or ≥ 25.0 kg/m2.

Full size image

In Table 3 data is presented regarding how participants rated the risk of a very obese woman in comparison to a woman of normal weight, again stratified according to pre pregnancy BMI. The majority of the respondents were aware of the increased risk of overall complications, gestational diabetes and hypertensive disorders in obese women whereas a smaller proportion identified higher risks for caesarean section, adverse neonatal outcomes and especially congenital anomalies. There were no significant differences in the responses of women with a pre pregnancy BMI < 25.0 compared to those with a BMI ≥ 25.0.

Table 3 Participant rated risk of pregnancy and childbirth complications for women with a BMI < 25.0 or ≥ 25.0
Full size table

The majority of respondents thought that weight loss prior to pregnancy would lower the risk of all pregnancy and birth complications independent of their own pre pregnancy BMI (Table 4).

Table 4 Responses regarding change in risk if an obese woman were to lose weight prior to pregnancy
Full size table

Two hundred and thirty-five women (57% of the total cohort independent of pre pregnancy BMI) were categorized as knowing about the risks of being obese on pregnancy, birth and neonatal outcomes. Table 5 provides information about a number of variables that we hypothesized might be related to knowledge about the risks of overweight and obesity. Educational status was consistently associated with knowledge of overweight and obesity prior to pregnancy. Women who were cared for in the private sector were more likely to be categorized as having a broad knowledge of the risks of overweight and obesity. These women were also more likely to have attended a preconception visit (98 of 157 women with private care (62.4%) vs. 122 of 255 women with public care (47.8%), P < 0.01). We adjusted this analysis for maternal educational status, and found that increased maternal educational status fully explained the difference in knowledge between women cared for in the private and public sector.

Table 5 Association between demographic variables and broad knowledge of obesity-related risk for pregnancy complications and outcomes.
Full size table

Discussion

57% of the women in this study knew that being very obese prior to pregnancy increased the overall risk of pregnancy and birth complications, and that weight loss prior to pregnancy in an obese woman would reduce the overall risk of complications. The responses did not differ between normal weight and overweight or obese women.

The majority of women correctly identified the impact of overweight and obesity on maternal complications including diabetes and hypertensive disorders develo** in pregnancy. The impact of pre pregnancy weight on caesarean section rates and neonatal outcomes was less well known (Figure 1 and Additional file 1). This is perhaps not surprising, given that relative risks are lower than for maternal adverse outcomes (Table 1). In addition, the increased risk of preterm delivery and congenital abnormalities is not consistently reported in the literature, until women are extremely obese (Class II and III obesity) although recent meta-analyses have indicated increased risks for both overweight and obese women [38, 40]. Given that a healthy baby is a highly valued outcome of pregnancy [46], increasing women's knowledge about the impact on overweight and obesity on neonatal problems such as birth defects might encourage women to actively attempt to lose weight prior to pregnancy. A meta-analysis of Leventhal's common-sense models as a theoretical basis for intervention programs identified moderate to strong relationships between knowledge of disease, co** behaviors and outcomes [47]. Tailored diet and exercise interventions for at-risk individuals have been shown to be effective in improving outcomes in type 2 diabetes in a number of studies [48–50]. Therefore a program that will encompass an increase in knowledge of the risks of obesity for maternal and neonatal pregnancy outcomes with tailored easily implementable lifestyle interventions may improve pregnancy outcome for obese women.

Tertiary degree qualification was associated with knowledge about the risks of overweight and obesity. Maternal educational status also fully explained the difference we found in knowledge of the risks of being obese between women cared for in the private and public sectors and between women who did or did not smoke during pregnancy. Educational status is an important predictor of birth outcomes [51], and is associated with better knowledge of other preconception health issues such as periconceptual folate supplementation [5]. Our data would suggest that to improve knowledge regarding the risks of obesity, targeting public health messages at those with lower levels of education would be important.

This study identifies the pre pregnancy health check as an excellent opportunity for improving education of women regarding the risks of obesity prior to pregnancy. Slightly more than half of all women attended a doctor for a pre pregnancy health check. It is important that women have their BMI determined at their pre pregnancy health check, are advised about the risks associated with pre pregnancy overweight and obesity, and where appropriate are provided with support to lose weight [52]. However, this study also showed that education levels are associated with the level of knowledge and preconception visits to health care professionals, and efforts to increase knowledge about the risks associated with obesity during pregnancy in women with lower education levels should include additional measures besides information during preconception visits.

Strengths and Limitations

This study provides information on risk perception relating to the influence of being overweight and obese on pregnancy and birth complications in a relatively large unselected cohort of pregnant women cared for in the private and public sectors. Given the dearth of previous information in this area, we believe that our data will provide useful information to help develop public health interventions for reducing optimizing preconception weight mas well as providing a baseline against which to measure changes in knowledge after future interventions.

We were concerned that this cohort might have been particularly skewed towards well educated women. Women in our cohort had only slightly higher rates of tertiary education (34.9% vs. 28.8%), and similar rates of secondary school non completion (26.2% vs. 27.4%) in comparison to national Australian data [1, 17, 20, 53, 54], hence this should not be a major source of bias in this study. It is possible that the responses in this survey might have been positively influenced by local media coverage regarding the problems of overweight and obesity which occurred at around the time of questionnaire administration and it would be useful to repeat this survey again in this population and also in other populations.

It would be worthwhile to conduct a similar survey in health professionals, to assess their understanding of the risks associated with being overweight and obese prior to conception. A detailed knowledge in this group, of the adverse health consequences associated with elevated BMI on pregnancy would be associated with opportunities to address weight loss preconception. This would be especially amongst general practitioners, who generally would provide preconception check ups and could target women requiring weight loss prior to conception.

All the outcomes that we examined are associated with obesity. One of the limitations of this study is that we did not include a false outcome to test whether participants simply assumed all adverse outcomes would be more common in obese women (reflecting the relatively widespread portrayal as obesity as a major contributor to general ill-health). However, the relative risk of each outcome comparing obese to non-obese women does vary in the published literature (Table 1) and knowledge of the effect of obesity on outcomes with a lower relative risk was lower in this survey, indicating that the results may reflect real knowledge.

Conclusions

This study provides evidence that many women correctly identify that overweight and obesity increases the overall risk of complications of pregnancy and childbirth and that this was independent of the woman's own BMI. There remains scope for improvement in women's knowledge about obesity as a risk factor for pregnancy, birth and neonatal complications. Less well educated women are less likely to know about the risks of overweight and obesity in pregnancy, and so future public health campaigns need to ensure that these women are specifically considered.