Midwifery 31 (2015) 138–146

Contents lists available at ScienceDirect

Midwifery journal homepage: www.elsevier.com/midw

Relationships between mental health symptoms and body mass index in women with and without excessive weight gain during pregnancy Skye McPhie, BAppSc(Psych)(Hons), DPsych(Health)(Lecturer)a, Helen Skouteris, BA(Psych) (Hons), PhD (Associate Professor)a,n, Matthew Fuller-Tyszkiewicz, BBSc, GD(Psychology), PhD(Senior Lecturer)a, Briony Hill, B App Sci (Ex Sports Sci) (Hons)(PhD candidate)a, Felice Jacka, BA, Pg Dip Sci, PhD(Associate Professor)b, Adrienne O'Neil, BA (Hons), PhD (Postdoctoral Research Fellow )b,c a

School of Psychology, Deakin University, 221 Burwood Highway, Burwood, Victoria 3125, Australia IMPACT Strategic Research Centre, School of Medicine, Deakin University, PO Box 281, Geelong 3220, Australia c School of Public Health and Preventive Medicine, Monash University, PO Box 281, Geelong 3220, Australia b

art ic l e i nf o

a b s t r a c t

Article history: Received 12 January 2014 Received in revised form 20 June 2014 Accepted 13 July 2014

Objective: This study investigated the prospective relationships between mental health symptoms (depressive and anxiety symptoms) and body mass index (BMI) in women with and without excessive weight gain during pregnancy. The secondary aim was to examine whether mental health symptoms and BMI were predictive of one another. Two models were tested: the first depicted depressive or anxiety symptoms predicting BMI, and the second model depicted BMI predicting depressive or anxiety symptoms. Design and participants: Women completed questionnaires at three time points throughout pregnancy, which comprised of the Depression, Anxiety and Stress Scale-21 and self-reported weight. Height and weight were also reported retrospectively at T1 to calculate pre-pregancy BMI category. To calculate total gestational weight gain (GWG), pre-pregnancy weight was substracted from weight at 36 weeks gestation. Methods: 183 women were tracked during pregnancy; Time (T)1 (mean ¼16.50 weeks of gestation, SD¼ .92), T2 (mean ¼24.40 weeks of gestation, SD ¼.92), and T3 (mean ¼32.61 weeks gestation, SD ¼.88). The sample was divided into those for whom weight gain exceeded the guidelines for GWG (excessive gestational weight gain; EGWG), and those who for whom it did not. Multigroup path analyses compared the longitudinal relationships between depressive or anxiety symptoms and BMI during pregnancy for women with and without EGWG. Findings: BMI did not predict depressive or anxiety symptoms. Depressive symptoms at T1, did however predict higher BMI at T2 for women without EGWG. Anxiety symptoms and BMI were not related, regardless of GWG status. Conclusion: These findings suggest that depressive symptoms may precede increased BMI during pregnancy in women who do not gain weight excessively. There may be longitudinal relationships between depressive symptoms and BMI during pregnancy; however, further research is required to identify the mechanisms that link these health outcomes and inform the focus of intervention design. & 2014 Elsevier Ltd. All rights reserved.

Keywords: Pregnancy Depression Anxiety Gestational weight gain

Introduction

n

Corresponding author. E-mail addresses: [email protected] (S. McPhie), [email protected] (H. Skouteris), [email protected] (M. Fuller-Tyszkiewicz), [email protected] (B. Hill), [email protected] (F. Jacka), [email protected] (A. O'Neil). http://dx.doi.org/10.1016/j.midw.2014.07.004 0266-6138/& 2014 Elsevier Ltd. All rights reserved.

Pregnancy is a period during which women experience physical and psychological change. Pregnant women are at elevated risk of rapid weight increase and hence maternal obesity (Gunderson and Abrams, 1999; Herring et al., 2008; Gould Rothberg et al., 2011). In fact, approximately 50% of women in Australia report gaining weight excessively during pregnancy (Hure et al., 2012; Hill et al., 2013), with similar rates of excessive gestational weight

S. McPhie et al. / Midwifery 31 (2015) 138–146

gain being reported in the US (67% of women: Wrotniak et al. (2008); 64.1% of women who are overweight pre-pregnancy: Chu et al. (2009)). Excessive gestational weight gain (EGWG) has been defined as an increase in weight during pregnancy that exceeds the recommendations of the US Institute of Medicine (Rasmussen and Yaktine, 2009) (defined below). Women who gain excessive weight during pregnancy have an increased risk of post partum obesity and retention of gestational weight gain (GWG) post birth. The latter has been shown to be a strong predictor of maternal overweight and obesity in the ensuing decade (Olson, 2008; Siega-Riz et al., 2009). Pregnant women are at risk of elevated depressive and anxiety symptoms with prevalence estimates ranging from 11.4% at 18 weeks gestation to 13.1% at 32 weeks gestation for antenatal depression and 14.6% at 18 weeks gestation to 15.6% at 32 weeks gestation for anxiety (Heron et al., 2004). Similarly, rates of depression and anxiety have been found to increase over the pregnancy and into the post-birth period (Skouteris et al., 2009a, 2009b). Depression during pregnancy also appears to be associated with an increased the risk of depression post partum (Milgrom et al., 2008; Sexton et al., 2012).’ Obesity and mental health symptoms, both common gestational health conditions, share a bidirectional relationship, with each health condition being a risk factor for the other (Desai et al., 2009; Williams et al., 2009; Luppino et al., 2010; Lykouras and Michopoulos, 2011; Lin et al., 2013). Both EGWG and maternal mental health symptoms are associated with negative health outcomes for both mother and child, including preeclampsia, complications with labour/childbirth, lower rates of breast feeding for the former, birth defects, late fetal death, reduced cognitive development in early childhood, and emotional and behavioural issues in mid to late childhood for the latter (Cnattingius et al., 1998; Baeten et al., 2001; Sebire et al., 2001; Watkins et al., 2003; Cedergren, 2006; Pedersen et al., 2011; Grigoriadis et al., 2013; Koutra et al., 2013; Leis et al., 2013). Yet, to date, there remains a paucity of research evaluating the longitudinal relationships between these depressive and anxiety symptoms and weight during pregnancy. Carter et al. (2000) examined the relationships between pre-pregnancy BMI and BMI at four months post-birth with anxiety and depressive symptoms during pregnancy and these two post partum time points. These authors found that BMI at both time points was positively associated the depression and anxiety during the post-birth period, and not during pregnancy. However, this study did not measure BMI during pregnancy. To our knowledge, Webb et al. (2008) have conducted the only study examining the longitudinal relationship between antenatal depressive symptoms and weight changes over the course of pregnancy. The overall aim of their study was to identify the psychosocial risk factors of adequate GWG. The authors found that antenatal depressive symptoms in early pregnancy (o20 weeks gestation) and mid-pregnancy (27–29 weeks gestation) were risk factors for EGWG. However, this study did not examine the bidirectional relationship between depressive symptoms and BMI, nor did they include anxiety symptoms as a measure of maternal mental health during pregnancy. Therefore, the purpose of this study was to examine the extent to which maternal mental health symptoms, as measured by depressive and anxiety symptoms, and weight (measured via BMI to account for height) are related during pregnancy for those with and without EGWG. There were three specific aims: (1) To explore the prospective relationships between mental health symptoms (depressive and anxiety symptoms) and BMI in women with and without EGWG (as determined by recommended guidelines). (2) To explore two models to determine whether antenatal depressive and anxiety symptoms precede weight gain, or

139

whether BMI is a predictor of mental health symptoms during pregnancy. (3) To test a stability model of depressive and anxiety symptoms and BMI through pregnancy.

Methods Study design A longitudinal study was conducted comprising 183 pregnant women over the age of 18 years who were tracked across three time points in pregnancy; Time 1 (T1: 16 weeks gestation; mean of 16.50 weeks of gestation, SD ¼.92), T2 (24 weeks gestation; mean of 24.40 weeks of gestation, SD ¼.92), and T3 (32 weeks gestation; mean of 32.61 weeks gestation, SD¼ .88) for depressive and anxiety symptoms, as well as BMI. Procedure Participants were recruited via advertising on online mother, child and baby forums, in parenting magazines, at baby and children's markets, and at obstetrician clinic waiting rooms in Geelong/Melbourne in the state of Victoria, Australia. Women between 10 and 16 weeks pregnant were targeted. After expressing interest in the study, prospective participants were provided a copy of the Plain Language Statement and given the opportunity to ask any questions before informed consent was provided. Participants were then mailed a series of questionnaires at each time point and were provided reply paid envelopes in order to complete and return the questionnaires within 1–2 weeks of receipt. Ethics approval was obtained from the Deakin University Human Research Ethics Committee. Study measurements Body mass index Pre-pregnancy BMI (kg/m2) was calculated using self-reported retrospective height and weight data collected at T1. Participants were classified as underweight (BMI o18.5 kg/m2), normal weight (BMI 18.5–24.9 kg/m2), overweight (BMI 25.0–29.9 kg/m2) or obese (BMI Z30.0 kg/m2) at pre-pregnancy (World Health Organization, 1995). According to the US Institute of Medicine's (IOM) guidelines (Rasmussen and Yaktine, 2009), recommended weight gain for women during pregnancy is as follows: those who are underweight should gain between 13 and 18 kg; those who are normal weight should gain between 11 and 16 kg; those who are overweight should gain 7–11 kg; and women who are obese should gain 5–9 kg. Total gestational weight gain was calculated at the end stages of pregnancy. EGWG was defined as weight gain beyond these parameters for pregnancy weight gain at 36 weeks of gestation, calculated as pre-pregnancy weight subtracted from weight at 36 weeks. All weight measures were assessed via maternal self-report and were used to calculate BMI at each time point.

Depression, Anxiety, and Stress Scale-21 (Lovibond and Lovibond, 1995; DASS-21) Participants completed both the Depression and Anxiety scales of the DASS-21 at T1, T2, and T3. The Depression scale consists of seven questions relating to feelings of helplessness and hopelessness, while the seven questions in the Anxiety scale ask participants to report worry, panic, and physical symptoms of anxiety. Participants rate the extent to which each statement

140

S. McPhie et al. / Midwifery 31 (2015) 138–146

related to them in the past month on a 4-point Likert scale ranging from 0 (Did not apply to me at all) to 3 (Applied to me very much, or most of the time). For the Depression scale, scores of 0–4 are rated as normal, scores of 5–6 as mild, scores of 7–10 as moderate, scores of 11–13 as severe, and scores of 14 or greater as extremely severe. In contrast, the Anxiety scale rates scores of 0–3 as normal, 4–5 as mild, 6–7 as moderate, 8–9 as severe, and 10 or greater as extremely severe (Lovibond and Lovibond, 1995). The DASS-21 has been found to have good construct validity and scale reliability (Henry and Crawford, 2005) and has been used with pregnant samples (Milgrom et al., 2011; Mohammad et al., 2011). For the present study, the Depression scale had Cronbach's α of .75, .75, and .74 for T1, T2, and T3, respectively. Similarly, the Cronbach's α for the Anxiety scale was .62 at T1, .74 at T2, and .72 at T3. Multidimensional Scale of Perceived Social Support (Zimet et al., 1988; MSPSS) The MSPSS was completed at T1 and assessed perceived social support across three scales: Family, Friends, and Significant Others. There are 12 questions that enquire how support is experienced from these three sources, which are answered on a 7-point Likert scale ranging from 1 (Very Strongly Agree) to 7 (Very Strongly Disagree). This scale has been found to be reliable and valid (Zimet et al., 1988), and the Cronbach's α for the present study is .90. Pittsburgh Sleep Quality Index (Buysse et al., 1988; PSQI) The PSQI is a measure of sleep quality. In this study, a single item from the PSQI (During the past month, how would you rate your sleep quality overall?) was used to assess sleep quality at T1. This question is answered on a 4-point Likert scale of Very Good, Fairly Good, Fairly Bad, or Very Bad. The PSQI has good test–retest reliability as well as sensitivity and specificity (Buysse et al., 1988) and maintains sound psychometrics with pregnant women (Jomeen and Martin, 2007). This single item inquiring about overall sleep quality has been found to highly correlate with global scores on the PSQI (r ¼.83) (Buysse et al., 1988). Demographic and other clinical information Participants provided information on their age, family income, educational attainment, parity and history of mental health concerns at T1. Data analysis To examine the relationships between depressive symptoms, anxiety symptoms and BMI during pregnancy, path analysis was undertaken using AMOS version 21. Data were screened for normality, outliers, and multicollinearity to ensure they met the assumptions of path analysis (Tabachnick and Fidell, 2007). Expectation maximisation was used for data imputation to address missing data. Where data were not normally distributed, data were transformed (i.e., square-root transformation for social support, depression and anxiety and log transformation for BMI). Descriptive statistics were used to summarise the data (see Table 1) and bivariate correlations between the covariate, mental health symptoms and BMI variables were calculated to assess the strength of the associations between variables for women with and without EGWG separately (see Tables 2 and 3). A multigroup path analysis was conducted to assess whether the relationships between depression/anxiety and BMI during pregnancy differed for those reporting EGWG compared to those without. We first explored these relationships without the covariate variables and subsequently included sleep quality, social support, and pre-pregnancy BMI as covariates for each model. Sleep quality, pre-pregnancy BMI and perceived social support were selected as covariates in the models

as they have been found to be linked to depressive symptoms during pregnancy (Skouteris et al., 2009a, 2009b; Kamysheva et al., 2010; Boothe et al., 2011; Dotlic et al., 2014; Nylen et al., 2013). Each of these models was tested twice. In the first test, model parameters were unconstrained, allowing for the relationships between the variables to be freely estimated for women who gain weight within or over the IOM GWG recommendations (referred to as the baseline model). The second test imposed restraints of the covariance parameters to assess whether the relationships between the variables are comparable between the two groups of women (referred to as the constrained model). To compare the significance of difference in fit between the baseline and constrained models, the χ2 difference test was used with degrees of freedom equal to the difference in df between the two models (Vandenberg and Lance, 2000). In the presence of non-invariance, modification indices were inspected to help identify the model parameter(s) that differed across groups.

Findings Sample characteristics The key characteristics of the sample are displayed in Table 1. The average age of the sample at T1 was 30.66 years (SD ¼4.31, Range¼ 19–44), and the majority were tertiary-educated (61.8% had a university degree) and had a household income in excess of AUD$105,000 (59.9%). The mean household income for Australians was $87,776 in 2009–2010 (Australian Bureau of Statistics, 2012). Mothers were primarily born in Australia (84.2%). Approximately half of the participants were first time mothers (51.9%) and most self-reported they were normal weight before pregnancy (70.5%). Self-reported weight and height may not reflect participants' true BMI, however the accuracy of self-reported weight during pregnancy has been found to be stronger than during other life stages (Herring et al., 2008). Fifty-six (30.8%) reported a history of mental health concerns, predominantly minor depression (17.5%), and/or anxiety disorder (12.0%). Seventy-one participants (38.8%) fulfilled criteria for EGWG, whereas 112 women did not exceed the GWG guidelines at 36 weeks gestation. Women who completed questionnaires at T1 only, T1 and T2, and at all three time-points were examined for between-group differences in relation to age, weight at 16 weeks gestation, family income, educational achievement, employment status, and parity. Results of a one-way ANOVA demonstrated that there were no significant differences between the groups based on these variables. However, there was a significant difference in gravidity across the groups, with 63.04% of those completing T1 only, 79.17% of those completing T1 and T2, and 48.50% of those completing all three time points having been pregnant more than once (p¼ .01, Fisher's Exact Test). Path analyses Multigroup path analysis – longitudinal relationship between antenatal depression and BMI The baseline and constrained models demonstrated a good fit (baseline model: χ2 ¼ 42.62(28), po.05, comparative fit index¼.99, standardised root mean error of approximation¼.05; constrained model: χ2 ¼69.72(44), p4.05, comparative fit index¼.99, standardised root mean error of approximation¼.06). The results from the χ2 difference test suggest that there was a significant worsening of model fit (Δχ2 ¼27.10, po.05) from the baseline to the constrained model. The modification indices for the constrained model suggest that there is a positive relationship between overall sleep quality at 16 weeks gestation and depressive symptoms at 32 weeks gestation for those who gained weight within the GWG recommended guidelines

S. McPhie et al. / Midwifery 31 (2015) 138–146

141

Table 1 Key characteristics of study sample. Variable

Mean (SD) or % (n) Total sample (n¼ 183)

Age, years Educational achievement (%) (n) Did not finish school Year 12 or equivalent Diploma or certificate University degree Family income (%) (n) $25,0001–$65,000 $65,001–$105,000 $105,001 and over Missing Parity (%) (n) First pregnancy Not first pregnancy Country of birth (%) (n) Australia New Zealand UK Europe North America Asia Africa History of mental health concerns (%) (n) Yes No Missing Most common responses to Have you ever been diagnosed with [insert mental health diagnosis]? (%) (n) Minor depression Anxiety disorder Major depression Postnatal depression Depression severity at T1 (%) (n) Normal Mild Moderate Severe Extremely severe Anxiety severity at T1 (%) (n) Normal Mild Moderate Severe Extremely severe Pre-pregnancy BMI (kg/m2) (%) (n) Underweight Normal weight Overweight Obese Gestational weight gain Pre-pregnancy BMI Kilograms gained over pregnancy n

30.66 (4.31)

Without EGWG group (n¼ 112)

With EGWG group (n¼ 71)

30.3 (4.39)

31.2 (4.16)

2.2 8.2 27.9 61.8

(4) (15) (51) (133)

3.6 5.4 26.8 55.4

(4) (6) (30) (72)

.0 12.8 25.6 57.7

(0) (9) (21) (41)

11.4 28.4 59.9 .5

(21) (52) (109) (1)

10.7 26.8 61.6 .9

(12) (30) (69) (1)

12.7 31.0 56.3 .0

(9) (22) (40) (0)

51.9 (95) 48.1 (88)

58.9 (66) 41.1 (46)

40.8 (29) 59.2 (42)

84.2 4.4 4.2 2.7 2.2 1.1 .5

87.5 4.5 2.7 1.8 2.7 .9 .0

78.9 4.2 8.5 4.2 1.4 1.4 1.4

(154) (8) (9) (5) (4) (2) (1)

(98) (5) (3) (2) (3) (1) (0)

(56) (3) (6) (3) (1) (1) (1)

30.6 (56) 68.9 (126) .5 (1)

32.1 (36) 67.0 (75) .9 (1)

28.2 (20) 71.8 (51) .0 (0)

17.5 12.0 4.4 2.2

19.6 11.6 4.5 1.8

(22) (13) (5) (2)

14.1 12.7 4.2 2.8

(10) (9) (3) (2)

87.5 8.9 .9 .9 1.9

(98) (10) (1) (1) (2)

88.7 8.5 2.8 .0 .0

(63) (6) (2) (0) (0)

(32) (22) (8) (4)

88% (161) 8.7% (16) 1.6% (3) .5%(1) 1.1% (2) 82% 10.9% 3.3% 3.8% .0%

(150) (20) (6) (7) (0)

73.2 11.6 .9 4.5 .0

(93) (13) (1) (5) (0)

80.3 9.9 7.0 2.8 .0

(57) (7) (5) (2) (0)

2.2 55.7 26.2 15.8

(4) (102) (48) (29)

2.0 70.5 10.7 16.1

(3) (79) (12) (18)

1.4 32.4 50.7 15.5

(1) (23) (36) (11)

25.41 (5.23) 13.10 (5.15)

24.88 (5.56) 10.57 (4.35)

26.26 (4.76) 17.07 (3.55)n

po .001.

only (β ¼.21, MI¼4.82). These results suggest that only one minor additional pathway (a covariate as a positive predictor of depressive symptoms in late pregnancy) that would improve the fit of the model for women without EGWG, and the models depicting the longitudinal relationships between depressive symptoms and BMI during pregnancy differ for women with and without GWG. Figs. 1 and 2 depict the standardised coefficients for the longitudinal path model of antenatal depressive symptoms and BMI for women who gained over or within the GWG recommendations over the course of their pregnancy, respectively. For both groups of pregnant women, both depressive symptoms and BMI appeared to be stable across pregnancy regardless of whether weight was gained excessively over pregnancy. Antenatal depression at 16 weeks gestation positively predicted BMI at 24 weeks of

gestation for women without EGWG. Depression and BMI were not related amongst women who reported EGWG.

Multigroup path analysis – longitudinal relationship between antenatal anxiety and BMI Both the baseline and constrained models were well fitting models (baseline model: χ2 ¼36.66(28), p 4.05, comparative fit index¼ 1.00, standardised root mean error of approximation ¼.04; constrained model: χ2 ¼63.80(44), p o.05, comparative fit index¼ .99, standardised root mean error of approximation ¼.05). The χ2 difference test suggests that was a significant worsening of model fit (Δχ2 ¼27.14, p o.05) from the baseline to the constrained model. There were no modification indices above 3.84

142

S. McPhie et al. / Midwifery 31 (2015) 138–146

Table 2 Means, standard deviations and correlations of sleep quality, social support, depression, anxiety and BMI for women with EGWG (n¼ 71).

(1) Sleep Quality at T1† (2) Social Support at T1‡ (3) Prepregnancy BMI (4) Depression at T1§ (5) Depression at T2§ (6) Depression at T3§ (7) Anxiety at T1§ (8) Anxiety at T2§ (9) Anxiety at T3§ (10) BMI at T1 (11) BMI at T2 (12) BMI at T3 Mean SD

(1) Sleep Quality at T1

(2) Social Support at T1

(3) Prepregnancy BMI

(4) Depression at T1

(5) Depression at T2

(6) Depression at T3

(7) Anxiety at T1

(8) Anxiety at T2

(9) Anxiety at T3

(10) BMI at T1

(11) BMI at T2

(12) BMI at T3

–

 .40nn

.00

.27n

.21

.07

.34nn

.35nn

.09

.04

.04

.04

–

 .06

 .19

 .20

 .16

 .13

 .13

 .12

 .10

 .11

 .10

–

.04

.05

.22

.02

.03

.21

.98nn

.97nn

.97nn

–

.63nn

.51nn

.54nn

.49nn

.35nn

.03

.03

.04



.72nn

.41nn

.67nn

.44nn

.03

.03

.06

nn

nn

nn

.21

.20

.22

nn

.03

 .03

 .02

nn

.05

 .06

 .04

.21

.21

.22

–

.98nn –

–

.28

–

.46

nn

.70

–

.66 .32 .51

–

2.35 .63

73.30 10.39

26.26 4.76

1.99 1.81

1.70 1.64

1.72 1.79

1.87 2.23

1.62 2.11

1.78 2.00

27.97 4.70

.97nn .98nn – 29.91 31.56 4.41 4.15

Abbreviations: BMI ¼ Body Mass Index; SD ¼standard deviation. n

po .05. p o.01. Pittsburgh Sleep Quality Scale. Multidimensional Scale of Perceived Social Support. Depression, Anxiety and Stress Scale-21.

nn

† ‡ §

(the threshold for significant improvement in model χ2) to suggest that freeing one or more parameter estimates across groups would improve model comparison. Therefore, from a practical perspective, the model was equivalent across the two groups, and the difference in model fit may be attributable to the accumulation of several or more minor differences in parameter size across models. Figs. 3 and 4 illustrate standardised coefficients for the longitudinal path model of antenatal anxiety and BMI for women who gained over or within the recommended GWG guidelines over the course of their pregnancy, respectively. For women who gained weight within the GWG guidelines, anxiety symptoms at 16 weeks gestation positively predicted anxiety at 32 weeks gestation. A common element between both groups of women was that anxiety symptoms at 16 weeks of gestation predicted anxiety symptomology at 24 weeks of gestation. Additionally, BMI had stable relationships across the time points throughout pregnancy for both groups of women. However, there were no relationships observed between anxiety symptoms and BMI in women with or without EGWG.

Discussion In a sample of pregnant women monitored for 20 weeks (16–36 weeks gestation), antenatal depression and anxiety symptoms remained relatively stable over the observation period, with mental health symptom scores recorded early in the gestational period predicting subsequent mental health symptoms. These findings are consistent with previous research (Heron et al., 2004; Skouteris et al., 2009a, 2009b; Sexton et al., 2012). The hypothesis that BMI would predict depressive and/or anxiety symptoms in pregnancy was not supported by the data.

However, there was partial support for the model that proposed depressive symptoms would precede BMI during pregnancy. That is, we observed that depressive symptoms at 16 weeks gestation positively predicted BMI at 24 weeks gestation in those without EGWG. This finding is consistent with Webb et al. (2008), which to the authors' knowledge is the only other study to have examined the longitudinal relationships between antenatal depressive symptoms and weight gain during pregnancy. Additionally, this finding also supports the literature that identifies depression as a risk factor (1.7-fold increase) for female overweight and obesity (Williams et al., 2009), and weight retention post-birth (Pedersen et al., 2011). Therefore, symptoms of depression in early pregnancy may be a significant risk factor for increased BMI in midpregnancy for women without EGWG. The multigroup path analysis also suggested that sleep quality at 16 weeks gestation was a positive predictor of depressive symptoms at 32 weeks gestation for women without EGWG. This result is inconsistent with the research that finds poor sleep quality to be associated with higher depressive symptoms during pregnancy (Skouteris et al., 2009a, 2009b; Kamysheva et al., 2010). However, as this positive relationship between early sleep quality and later antenatal depressive symptoms is not supported by previous findings, it may reflect a type 1 error. BMI scores at all time points were shown to be related positively to one another, independent of EGWG status. This suggests that higher BMI at the end of pregnancy and EGWG are not interchangeable measures for predicting antenatal depressive symptoms. This has been illustrated previously (Dotlic et al., 2014), whereby BMI at the end of pregnancy, not weight changes during pregnancy, were related to depression symptoms in the early-postpartum period. The absence of statistically significant relationships between antenatal mental health symptoms (depressive and anxiety

S. McPhie et al. / Midwifery 31 (2015) 138–146

143

Table 3 Means, standard deviations and correlations of sleep quality, social support, depression, anxiety and BMI for women without EGWG (n¼ 112).

(1) Sleep Quality at T1† (2) Social Support at T1‡ (3) Pre-pregnancy BMI (4) Depression at T1§ (5) Depression at T2§ (6) Depression at T3§ (7) Anxiety at T1§ (8) Anxiety at T2§ (9) Anxiety at T3§ (10) BMI at T1 (11) BMI at T2 (12) BMI at T3 Mean SD

(1) Sleep Quality at T1

(2) Social Support at T1

(3) Prepregnancy BMI

(4) Depression at T1

(5) Depression at T2

(6) Depression at T3

(7) Anxiety at T1

(8) Anxiety at T2

(9) Anxiety at T3

(10) BMI at T1

(11) BMI at T2

(12) BMI at T3

–

 .13

.05

.16

.05

.26nn

.15

.11

.14

.03

.01

.02

 .03

.03

.10

 .10

 .09

 .10

–

2.12 .64

74.14 8.88

n

 .13

 .08

 .19

–

.09

 .05

.12

.15

.03

.03

.97nn

.96nn

.96nn

–

.65nn –

.50nn .56nn –

.30nn .19nn .23nn –

.28nn .43nn .38nn .57nn –

.32nn .37nn .31nn .62nn .81nn –

.09  .05 .15 .16 .02 .04 –

.15  .03 .16 .16 .01 .04 .98nn –

2.20 2.52

1.78 1.93

2.01 2.08

1.82 2.03

1.33 2.01

1.59 2.01

25.55 4.94

26.87 4.79

.12  .04 .15 .15 .02 .05 .98nn .99nn – 28.11 4.62

24.88 5.46

 .22

n

Abbreviations: BMI¼ Body Mass Index; SD ¼ standard deviation. n

po .05. p o.01. Pittsburgh Sleep Quality Scale. Multidimensional Scale of Perceived Social Support. Depression, Anxiety and Stress Scale-21.

nn

† ‡ §

.17 .17

Depression at T1 .00 .00

.69** .69**

.00 .00

.00

.00 .00

.00 -.32 .87** .87**

.59** .59** .00

1.66

.00

1.66

-.32 BMI at T1

Depression at T3

Depression at T2

BMI at T2 .17*

.69** .69**

BMI at T3

.17*

Fig. 1. Model of the longitudinal relationship between antenatal depression and BMI for women with EGWG (n¼ 71). Beta weights between Depression and BMI at T1, T2 and T3 after controlling sleep quality and social support at T1 and pre-pregnancy BMI (in bold). *po .05, **p o .001.

symptoms) and EGWG during pregnancy may be explained by the small proportion of participants with elevated depressive and/or anxiety symptoms in this sample. Additionally, the majority of the sample did not report a history of mental illness. Given that having a history of depression has been found to be related to post-natal depression amongst women who are obese when entering pregnancy (Cline and Decker, 2012), the low rates of previous and current mental health concerns in the current sample may not have been high enough to detect relationships between antenatal mental health symptoms and BMI.

(Herring et al., 2008). Finally, this was a selective sample of welleducated women who had a normal pre-pregnancy BMI. Therefore, the generalisability of the results is somewhat limited. Notwithstanding these limitations, the sample size, multiple assessment points, and longitudinal nature of the data are strengths of this study, allowing investigation of the complex prospective and bidirectional relationship of psychological status and BMI in those with and without EGWG.

Strengths and limitations

In conclusion, the results of this study corroborate those from others in finding that mental health symptoms are relatively stable across pregnancy, independent of GWG. Given the chronicity of the antenatal mental health symptoms, these results support existing calls for mandatory depression screening in antenatal clinics to allow for early intervention to prevent postnatal depression (Marcus et al., 2003; Setse et al., 2009;Muzik and Borovska, 2010). Additionally, depressive symptoms in early pregnancy may influence BMI in mid-pregnancy for women without excessive GWG. However, the mechanisms that link mental health and

This study had several limitations that should be considered in interpreting the findings. Although the DASS-21 is a validated and widely used self-report assessment tool, psychiatric diagnostic interviews are the gold standard measurement for assessing mental disorders. Additionally, the use of self-report anthropometric measurements may have resulted in inaccurate recording of participants' BMI status. However, self-reported weight during pregnancy is considered more accurate than at other life stages

Recommendations for future research

144

S. McPhie et al. / Midwifery 31 (2015) 138–146

.14 .14 Depression at T1 .01** .01** .00 .00

Depression at T3

Depression at T2 .51** .51**

.55** .56**

.00 .00

.00 .00

-.70

.00 .00

.55 .55

-.70 .91** .91** BMI at T2

BMI at T1

.80** .80** BMI at T3

.12*

.12*

Fig. 2. Model of the longitudinal relationship between antenatal depression and BMI for women without EGWG (n¼ 112). Beta weights between Depression and BMI at T1, T2 and T3 after controlling sleep quality and social support at T1 and pre-pregnancy BMI (in bold). *po .05, **p o .001.

.01

.01 Anxiety at T1 .65** .65**

.00 .00

.00 .00

.55** .55** .00 .00

.00 .00

.00 -.01

Anxiety at T3

Anxiety at T2

-1.23 -1.23

BMI at T1

2.27 2.27

.87** .87**

BMI at T2 .17*

.69**.69**

BMI at T3

.17*

Fig. 3. Model of the longitudinal relationship between antenatal anxiety and BMI for women with EGWG (n¼ 71). Beta weights between Anxiety and BMI at T1, T2 and T3 after controlling sleep quality and social support at T1 and pre-pregnancy BMI (in bold). *p o.05, **p o .001.

.30** .30** Anxiety at T1

Anxiety at T3 Anxiety at T2

.00 .00

.53**

.55** .55**

.00 .00

.00

.00

.0 1

.0 0 .12 .12 .92** .92**

BMI at T2

BMI at T1

.53**

.37 .37 .79** .79**

.00 .00

BMI at T3

.12* .12* Fig. 4. Model of the longitudinal relationship between antenatal anxiety and BMI for women without EGWG (n ¼112). Beta weights between Anxiety and BMI at T1, T2 and T3 after controlling sleep quality and social support at T1 and pre-pregnancy BMI (in bold). *po .05, **p o .001.

S. McPhie et al. / Midwifery 31 (2015) 138–146

weight gain are unclear; it may be that depressive symptoms such as, fatigue, feelings of worthlessness/helplessness, as well as sleep and appetite disturbances, lead to increased weight gain through poor self-care regarding diet and physical activity. For example, Leung and Kaplan (2009) advocate that perinatal nutrition and antenatal mood are likely to be associated with one another, and this may be one possible mechanism by which antenatal mental health and weight are connected. Likewise, weight gain may contribute to symptoms of depression by triggering eating as a coping mechanism to combat poor self-esteem, body image concerns or other negative feelings. Therefore, further research to better understand the how and why mental health and weight may be linked during pregnancy is warranted. In particular, by identifying the mechanisms that might mediate the bidirectional relationships between antenatal mental health symptoms and increases in BMI during pregnancy (Milgrom et al., 2012), the focus of interventions can be tailored to meet these needs.

Conflict of interest The authors have no conflicts of interest to disclose.

Acknowledgements Nil. The authors alone are responsible for the content and writing of the paper. References Australian Bureau of Statistics, 2012. Household income, expenditure and wealth. 〈http://www.abs.gov.au/ausstats/[email protected]/Lookup/by%20Subject/1301. 0  2012  Main%20Features  Household%20income,%20expenditure%20and% 20wealth 193〉 (last accessed 27 September 2013). Baeten, J.M., Bukusi, E.A., Lambe, M., 2001. Pregnancy complications and outcomes among overweight and obese nulliparous women. Am. J. Public Health 91, 436–440. Boothe, A.S., Brouwer, R.J.N., Carter-Edwards, L., Østbye, T., 2011. Unmet social support for healthy behaviors among overweight and obese postpartum women: results from the active mothers postpartum study. J. Womens Health 20, 1677–1685. Buysse, D.J., Reynolds III, C.F., Monk, T.H., Berman, S.R., Kupfer, D.J., 1988. The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research. Psychiatry Res. 28, 193–213. Cedergren, M., 2006. Effects of gestational weight gain and body mass index on obstetric outcome in Sweden. Int. J. Gynecol. Obstet. 93, 269–274. Carter, A.S., Baker, C.W., Brownwll, K.D., 2000. Body mass index, eating attitudes, and symptoms of depression and anxiety in pregnancy and the postpartum period. Psychosom. Med. 62, 264–270. Chu, S.Y., Callaghan, W.M., Bish, C.L., D'Angelo, D., 2009. Gestational weight gain by body mass index among US women delivering live births, 2004–2005: fueling future obesity. Am. J. Obstet. Gynecol. 200 (e1-271–e1-277). Cline, K.M.C., Decker, J., 2012. Does weight gain during pregnancy influence postpartum depression? J. Health Psychol. 17, 333–342. Cnattingius, S., Bergström, R., Lipworth, L., Kramer, M.S., 1998. Prepregnancy weight and the risk of adverse pregnancy outcomes. N. Engl. J. Med. 338, 147–152. Desai, R.A., Manley, M., Desai, M.M., Potenza, M.N., 2009. Gender differences in the association between body mass index and psychopathology. CNS Spectr. 14, 372–383. Dotlic, J., Terzic, M., Babic, D., et al., 2014. The influence of body mass index on the perceived quality of life during pregnancy. Appl. Res. Qual. Life 9, 387–399. http://dx.doi.org/10.1007/s11482-013-9224-z. Gould Rothberg, B.E., Magriples, U., Kershaw, T.S., Rising, S.S., Ickovics, J.R., 2011. Gestational weight gain and subsequent postpartum weight loss among young, low-income, ethnic minority women. Am. J. Obstet. Genecol. 204, e1–e11. Grigoriadis, S., Vonder Porten, E.H., Mamisashvili, L., et al., 2013. The impact of maternal depression during pregnancy on perinatal outcomes: a systematic review and meta-analysis. J. Clin. Psychiatry 74, e321–e341. Gunderson, E.P., Abrams, B., 1999. Epidemiology of gestational weight gain and body weight changes after pregnancy. Epidemiol. Rev. 21, 261–275. Henry, J.D., Crawford, J.R., 2005. The short-form version of the Depression Anxiety Stress Scales (DASS-21): construct validity and normative data in a large nonclinical sample. Br. J. Clin. Psychol. 44, 227–239. Heron, J., O'Connor, T.G., Evans, J., Golding, J., Glover, V., 2004. The course of anxiety and depression through pregnancy and the postpartum in a community sample. J. Affect. Disord. 80, 65–73.

145

Herring, S.J., Oken, E., Haines, J., et al., 2008. Misperceived pre-pregnancy body weight status predicts excessive gestational weight gain: findings from a US cohort study. BMC Pregnancy Childbirth 8, 54. Hill, B., Skouteris, H., McCabe, M., Fuller-Tyszkiewicz, M., 2013. Body image and gestational weight gain: a prospective study. J. Midwifery Womens Health 58, 189–194. Hure, A.J., Collins, C.E., Giles, W.B., Paul, J.W, Smith, R., 2012. Greater maternal weight gain during pregnancy predicts a large buy lean fetal phenotype: a prospective cohort study. Matern. Child Health J. 16, 1374–1384. Jomeen, J., Martin, C.R., 2007. Assessment and relationship of sleep quality to depression in early pregnancy. J. Reprod. Infant Psychol. 25, 87–99. Koutra, K., Chatzi, L., Bagkeris, M., Vassilaki, M., Bitsios, P., Kogevinas, M., 2013. Antenatal and postnatal maternal mental health as determinants of infant neurodevelopment at 18 months of age in a mother–child cohort (Rhea Study) in Crete, Greece. Soc. Psychiatry Psychiatr. Epidemiol. 48, 1335–1345. Kamysheva, E., Skouteris, H., Wertheim, E.H., Paxton, S.J., Milgrom, J., 2010. A prospective investigation of the relationships among sleep quality, physical symptoms, and depressive symptoms during pregnancy. J. Affect. Disord. 123, 317–320. Leis, J.A., Heron, J., Stuart, E.A., Mendelson, T., 2014. Associations between maternal mental health and child emotional and behavioural problems: does prenatal mental health matter? J. Abnorm. Child Psychol. 42, 161–171. http://dx.doi.org/ 10.1007/s10802-013-9766-4. Leung, B.M., Kaplan, B.J., 2009. Perinatal depression: prevalence, risks, and the nutrition link: a review of the literature. J Am Diet Assoc. 109, 1566–1575. Lin, H.-Y., Huang, C.-K., Tai, C.-M., et al., 2013. Psychiatric disorders of patients seeking obesity treatment. BMC Psychiatry 13, 1. Lovibond, S.H., Lovibond, P.F., 1995. Manual for the Depression Anxiety Stress Scales, 2nd edn. Psychology Foundation, Sydney, NSW. Luppino, F.S., de Wit, L.M., Bouvy, P.F., et al., 2010. Overweight, obesity, and depression. Arch. Gen. Psychiatry 67, 220–229. Lykouras, L., Michopoulos, J., 2011. Anxiety disorders and obesity. Psychiatriki 22, 307–313. Marcus, S.M., Flynn, H.A., Blow, F.C., Barry, K.L., 2003. Depressive symptoms among pregnant women screened in obstetrics settings. J. Womens Health 12, 373–380. Milgrom, J., Gemmill, A.W., Bilszta, J.L., et al., 2008. Antenatal risk factors for postnatal depression: a large prospective study. J. Affect. Disord. 108, 147–157. Milgrom, J., Schembri, C., Ericksen, J., Ross, J., Gemmill, A.W., 2011. Towards parenthood: an antenatal intervention to reduce depression, anxiety and parenting difficulties. J. Affect. Disord. 130, 385–394. Milgrom, J., Skouteris, H., Worotniuk, T., Henwood, A., Bruce, L., 2012. The association between ante- and postnatal depressive symptoms and obesity in both mother and child: a systematic review of the literature. Womens Health Issues 22, e319–e328. Mohammad, K.I., Gamble, J., Creedy, D.K., 2011. Prevalence and factors associated with the development of antenatal and postnatal depression among Jordanian women. Midwifery 27, e238–e245. Muzik, M.M., Borovska, S., 2010. Perinatal depression: implications for child mental health. Ment. Health Fam. Med. 7, 239–247. Nylen, K.J., O'Hara, M.W., Engeldinger, J., 2013. Perceived social support interacts with prenatal depression to predict birth outcomes. J. Behav. Med. 36, 427–440. Olson, C.M., 2008. Achieving a healthy weight gain during pregnancy. Annu. Rev. Nutr. 28, 411–423. Pedersen, P., Baker, J.L., Henriksen, T.B., et al., 2011. Influence of psychosocial factors on postpartum weight retention. Obesity 19, 639–646. Rasmussen, K.M., Yaktine, A.L. (Eds.), 2009. In Weight Gain During Pregnancy: Reexamining the Guidelines. National Academy Press, Washington, DC. Sebire, N.J., Jolly, M., Harris, J.P., et al., 2001. Maternal obesity and pregnancy outcome: a study of 287,213 pregnancies in London. Int. J. Obes. Relat. Metab. Disord. 25, 1175–1182. Setse, R., Grogan, R., Pham, L., et al., 2009. Longitudinal study of depressive symptoms and health-related quality of life during pregnancy and after delivery: the health status in pregnancy (HIP) study. Matern. Child Health J. 13, 577–587. Sexton, M.B., Flynn, H.A., Lancaster, C., et al., 2012. Predictor of recovery from prenatal depressive symptoms from pregnancy through postpartum. Int. J. Womens Health 21, 43–49. Siega-Riz, A.M., Viswanathan, M., Moos, M.K., et al., 2009. A systematic review of outcomes of maternal weight gain according to the Institute of Medicine recommendations: birthweight, fetal growth, and postpartum weight retention. Am. J. Obstet. Genecol. 201, 399.e1–399.e14. Skouteris, H., Wertheim, E.H., Germano, C., Paxton, S.J., Milgrom, J., 2009a. Assessing sleep during pregnancy: a study across two time points examining the Pittsburgh Sleep Quality Index and associations with depressive symptoms. Womens Health Issues 19, 45–51. Skouteris, H., Wertheim, E.H., Rallis, S., Milgrom, J., Paxton, S., 2009b. Depression and anxiety through pregnancy and the early postpartum: an examination of prospective relationships. J. Affect. Disord. 113, 303–308. Tabachnick, B.G., Fidell, L.S., 2007. Using Multivariate Statistics. Pearson/Allyn & Bacon, Boston. Vandenberg, R.J., Lance, C.E., 2000. A review and synthesis of measurement invariance literature: suggestions, practices, and recommendations for organizational research. Organ. Res. Methods 3, 4–70. Watkins, M.L., Rasmussen, S.A., Honein, M.A., Botto, L.D., Moore, C.A., 2003. Maternal obesity and risk for birth defects. Pediatrics 111, 1152–1158. Webb, J.B., Siega-Riz, A.M., Dole, N., 2008. Psychosocial determinants of adequacy of gestational weight gain. Obesity 17, 300–309.

146

S. McPhie et al. / Midwifery 31 (2015) 138–146

Williams, L.J., Pasco, J.A., Henry, M.J., et al., 2009. Lifetime psychiatric disorders and body composition: a population-based study. J. Affect. Disord. 118, 173–179. World Health Organization, 1995. Physical Status: The Use and Interpretation of Anthropometry. Report of a WHO Expert Committee. WHO Technical Report Series 854. Author, Geneva.

Wrotniak, B.H., Shults, J., Butts, S., Stettler, N., 2008. Gestational weight gain and risk of overweight in the offspring at age 7 y in a multicenter, multiethnic cohort study. Am. J. Clin. Nutr. 87, 1818–1824. Zimet, G.D., Dahlem, N.W., Zimet, S.G., Farley, G.K., 1988. The multidimensional scale of perceived social support. J. Personal. Assess. 51, 30–41.