Review
State–trait anxiety inventory (STAI) scores during pregnancy following intervention with complementary therapies

https://doi.org/10.1016/j.jad.2012.04.027Get rights and content

Abstract

Objective

We review intervention trials that have used the State–Trait Anxiety inventory (STAI) as a measure of maternal anxiety in pregnancy in order to provide ranges in scores before and after participation in complementary therapy-based interventions to highlight the expected ranges of scores in pregnancy and determine whether anxiety in pregnancy is amenable to change when measured by the STAI.

Methods

Combinations of the key words “STAI”, “state anxiety”, “pregnancy”, “anxiety”, “maternal”, “stress”, “outcome” and “intervention” were used to search publications between January 1970 and January 2011. Studies eligible for inclusion recruited low risk, adult women to a non-pharmacological intervention or a comparator group, and measured anxiety at baseline and post-intervention.

Results

Ten studies were eligible. Scores were routinely high compared to expected ranges in non-pregnant populations. Studies examining the immediate effects of an intervention consistently reported significantly lowered STAI scores after a single session. Likewise, studies examining the effect of interventions consisting of multiple sessions over the course of pregnancy found that those in the intervention group were more likely to show an improvement in STAI scores.

Limitations

Heterogeneity in type and duration of intervention prevent drawing conclusions on which were most effective in reducing anxiety.

Conclusion

Scores on the STAI appear amenable to change during pregnancy, both after a single session and multiple sessions of interventions designed to reduce maternal anxiety. This review offers a guideline for the expected range of scores for future studies examining the efficacy of interventions in pregnancy when using the STAI.

Introduction

Pregnancy inevitably causes upheavals to many aspects of a woman’s life. Anxieties and fears concerning alterations to her daily life, and her own and her baby’s health are normal (Clark et al., 2009, Cantwell and Cox, 2003), but some mothers may become distressed if they feel unable to cope with this adjustment to impending motherhood (Furber et al., 2009). Furthermore, although there are conflicting data, antenatal anxiety, independent of depression, has been associated with adverse effects to both the mother and infant. Antenatal anxiety may be a risk factor for maternal mental health problems, such as an increased likelihood of postnatal depression (Coelho et al., 2011, Austin et al., 2007) and impaired bonding (Lindgren, 2001, Condon and Corkindale, 1997), and obstetric complications, such as length of labour (Lederman et al., 1978) and premature delivery (Hobel et al., 1999). Furthermore, longitudinal studies have shown the infants of mothers with elevated anxiety while pregnant to be at greater risk of behavioural problems as neonates and toddlers (Austin et al., 2005, Van den Bergh and Marcoen, 2004, O'Connor et al., 2002). The mechanism by which increased anxiety may precipitate these adverse outcomes is unclear, although over-stimulation of the hypothalamic–pituitary–adrenal (HPA) axis, with elevated secretion of glucocorticoids such as cortisol, has been cited as a key component (Sarkar et al., 2008, Talge et al., 2007). This increased physiological stress response may interfere with fetal development with studies showing fetuses of mothers reporting high anxiety exhibiting differential blood flow, heart rate variability and motor activity compared to fetuses of mothers experiencing low anxiety (Sjostrom et al., 1997, Sjostrom et al., 2002, Groome et al., 1995; Van den Bergh, 1990, Field et al., 2003).

Although a growing body of evidence highlights the potential adverse outcomes of increased maternal anxiety, pharmacological interventions to counter such effects are limited by side effects, pharmacokinetic alterations in pregnancy, and risk of fetal toxicity. Consequently, the National Institute of Clinical Health and Excellence (NICE) in the UK has called (2007) for randomised controlled trials to identify non-pharmacological interventions that may prevent the escalation of sub-threshold anxiety symptoms into a clinical diagnosis, and improve a woman’s ability to cope with day-to-day life in both the antenatal and postnatal periods.

In determining whether an intervention is effective, it is important that the outcome measures chosen to assess anxiety are robust and valid. In a recent meta-analysis, the Spielberger State–Trait Inventory (STAI) was found to be by far the most widely used questionnaire to measure anxiety in pregnancy (Spielberger et al., 1970, Spielberger, 1983, Littleton et al., 2007). The STAI is a 40-item scale, using a 4-point Likert scale for each item. The scale can be used to measure both trait anxiety (how dispositionally anxious a person is across time and situations) and state anxiety (how anxious a person is feeling at a particular moment) as it consists of two separate sub-scales (STAI-T and STAI-S, respectively) which each have 20 items. The STAI has been repeatedly used in intervention trials with non-pregnant individuals, with reliability coefficients (Cronbach’s alphas) of 0.93 for the trait scale and 0.95 for the state scale (van der Ploeg et al., 1980, Spielberger et al., 1970). However the STAI was validated in non-pregnant individuals (Spielberger et al., 1970) and STAI-S scores reported by pregnant women in the third trimester were significantly higher than in non-pregnant women, matched for age, level of education and socio-economic status (Fatoye et al., 2004). This disparity may be due to pregnancy being a significant life event, which may provoke changes in trait and state anxiety (Hundley et al., 1998). Conversely, due to the physiological changes that occur during pregnancy, some items on the questionnaire may lose their face validity in a pregnant population and reflect somatic features of pregnancy rather than maternal mood, and consequently provide an unrepresentative score. Ambiguity in interpreting certain items when pregnant is a problem that has been highlighted through factorial analysis of the Beck Depression Inventory (Salamero et al., 1994). Thus, non-anxious pregnant women experiencing pregnancy-specific physical strain and somatic symptoms may show similar scores to someone with high anxiety but no somatic symptoms due to different interpretations of the same word (e.g., ‘tense’, ‘strained’, ‘comfortable’). Consequently when non-pregnancy-specific tools of assessment are used in a pregnant population, the somatic symptoms experienced by many pregnant women may lead to such women achieving high scores and the mistaken interpretation that anxiety disorders are more prevalent in pregnant women compared to the general population (Faisal-Cury and Menezes, 2007). With regard to intervention studies implemented in pregnancy, it is important to determine the extent that scores on measures such as the STAI-S can be changed or whether scores will remain ‘fixed’ or increase due to the escalation of somatic symptoms as gestation progresses.

Maternal anxiety, when measured by both STAI subscales, is observed to fluctuate during pregnancy. According to Haddad et al., 1985, state anxiety decreased in the second trimester but increased again in the third trimester. The concept of a U-shaped curve in the pattern of state anxiety was further validated by both Teixeira et al. (2009) and Bhagwanani et al. (1997), who showed state anxiety levels peaked in the first and third trimesters, while depression decreased throughout pregnancy. Several studies have shown that STAI-S scores are significantly increased in the third trimester compared to the first and second trimesters (Gunning et al., 2010, Da Costa et al., 1999). Trait anxiety has shown similar changes across the course of pregnancy, with Haddad et al., 1985 reporting a decrease in STAI-T scores between 16 and 28 weeks gestation, and STAI-T scores also being shown to positively correlate with gestational age (Albrecht and Rankin, 1989). The course of anxiety in pregnancy as measured by the STAI may be of clinical importance; after controlling for obstetric risk, pregnancy specific anxiety, ethnicity, parity and life events, Glynn et al. (2008) linked an increase in STAI-S score between the second and third trimester to an elevated risk of premature delivery. Interventions to curb increased third trimester anxiety may reduce the potentially adverse effects of antenatal anxiety upon mother and infant.

In light of the NICE guidelines’ (2007) call for more randomised controlled trials of non-pharmacological interventions, the aim of this review was to catalogue the range of scores reported in published studies that tested complementary therapies and utilised the STAI-S to monitor the effect on anxiety in pregnancy. This information will aid future studies by providing the expected range and, importantly, the likely change in STAI-S score following different types of non-pharmacological intervention in STAI-measured anxiety of pregnant women. This review does not aim at providing a comprehensive overview of STAI scores at specific time points in pregnancy but solely in intervention studies to give an indication of possible change.

Section snippets

Search strategy

Studies were retrieved from several databases [EMBASE, MEDLINE and PsychINFO (as the STAI was created in 1970, studies published between 1970 and April 2012 were examined)] using combinations of the key words “STAI”, “state anxiety”, “pregnancy”, “anxiety” “maternal” “stress”, “outcome” and “intervention”. The citation lists of relevant studies and reviews were checked for additional studies.

Inclusion/exclusion criteria

Studies eligible for inclusion had administered the STAI-S at two separate time points during pregnancy

Results

Ten studies of the 146 abstracts identified by the search strategy met the inclusion criteria. Eight of the studies were randomised controlled trials. The interventions tested by the studies meeting the inclusion criteria are shown in Table 1.

Discussion

The data summarised in this review shows the change in STAI-S scores reported in women taking part in interventions designed to reduce maternal anxiety and stress in pregnancy. Encouragingly, STAI-S scores were shown to be amenable to change in single session intervention studies and studies that examined the effects of multiple sessions of an intervention. Moreover, differences in mean change between intervention and control groups suggest that changes in score are not solely due to feelings

Summary

Scores for the state version of the STAI are amenable to change after participation in both single and multiple sessions of an intervention. A differential change in scores between the experimental groups and women who receive treatment as usual suggests these changes are not due simply to changes in pregnancy-specific symptoms. Furthermore, comparison with the direction of scores reported for other measures suggests that the STAI-S is not simply measuring emotional states such as depression or

Role of funding source

This review had no received no direct funding. This review was conducted while the first author was working on a PhD studentship funded by Tommy’s: the baby charity. Tommy’s: the baby charity had no involvement in the design, conduct or writing or the review, nor in the decision to submit this paper for publication.

Conflict of interest

No author for this paper has a conflict of interest.

Acknowledgements

We would like to thank Mrs Karen Clarke for proof reading of this review article.

References (53)

  • C.J. Hobel et al.

    Maternal plasma corticotropin-releasing hormone associated with stress at 20 weeks' gestation in pregnancies ending in preterm delivery

    American Journal of Obstetrics & Gynecology

    (1999)
  • V. Hundley et al.

    Can anxiety in pregnant women be measured using the state–trait anxiety inventory

    Midwifery

    (1998)
  • E.S. Ji et al.

    The effects of Qi exercise on maternal/fetal interaction and maternal well-being during pregnancy

    Journal of Obstetric, Gynaecologic and Neonatal Nursing

    (2010)
  • R.P. Lederman et al.

    The relationship of maternal anxiety, plasma, catecholamines, and plasma cortisol to progress in labor

    American Journal of Obstetrics & Gynecology

    (1978)
  • H.L. Littleton et al.

    Correlates of anxiety symptoms during pregnancy and association with perinatal outcomes: a meta-analysis

    American Journal of Obstetrics & Gynecology

    (2007)
  • T.G. O'Connor et al.

    Antenatal anxiety predicts child behavioral/emotional problems independently of postnatal depression.[see comment]

    Journal of the American Academy of Child & Adolescent Psychiatry

    (2002)
  • K. Sjostrom et al.

    Maternal anxiety in late pregnancy and fetal hemodynamics

    European Journal of Obstetrics, Gynecology, & Reproductive Biology

    (1997)
  • K. Sjostrom et al.

    Maternal anxiety in late pregnancy: effect on fetal movements and fetal heart rate

    Early Human Development

    (2002)
  • C. Teixeira et al.

    Anxiety and depression during pregnancy in women and men

    Journal of Affective Disorders

    (2009)
  • C. Urech et al.

    Effects of relaxation on psychobiological wellbeing during pregnancy: a randomized controlled trial

    Psychoneuroendocrinology

    (2010)
  • S.A. Albrecht et al.

    Anxiety levels, health behaviors, and support systems of pregnant women

    Maternal-Child Nursing Journal

    (1989)
  • S.G. Bhagwanani et al.

    Relationship between prenatal anxiety and perinatal outcome in nulliparous women: a prospective study

    Journal of the National Medical Association

    (1997)
  • M.Y. Chang et al.

    Effects of music therapy on psychological health of women during pregnancy

    Journal of Clinical Nursing

    (2008)
  • A. Clark et al.

    My baby body: a qualitative insight into women's body-related experiences and mood during pregnancy and the postpartum

    Journal of Reproductive and Infant Psychology

    (2009)
  • S. Cohen et al.

    A global measure of perceived stress

    Journal of Health and Social Behaviour

    (1983)
  • Cohen, S., Williamson, G., (1988). Perceived stress in a probability sample of the United States. In: The Social...
  • Cited by (0)

    View full text