Evidence Review

Correlates of Physical Activity During

Pregnancy: A Systematic Review with
Implications for Evidence-based Practice
Meghan Garland, MSN, CNM ● JoEllen Wilbur, PhD, RN, FAAN ● Pamela Semanik,
PhD, APN ● Louis Fogg, PhD

Key words ABSTRACT

pregnancy, theory,
health behaviors,
health promotion/
health education,
quantitative,
research utilization
Background: Numerous attempts have been made to improve women’s physical activity par-
ticipation during pregnancy, but activity levels remain low.
Aim: To examine systematically the associations of physical activity participation during preg-
nancy with non-modifiable correlates (not subject to change) and modifiable theoretical cor-
relates of physical activity.
Methods: This systematic review followed PRISMA (preferred reporting items for systematic
review and meta-analysis) guidelines. It includes cross-sectional and longitudinal studies that
assessed non-modifiable correlates and modifiable theoretical correlates of physical activity.
Five electronic databases were searched for studies published in the English language be-
tween 2010 and 2017. An author-developed data collection tool was used to examine selected
variables; effect sizes were determined; and study bias was assessed.
Results: Ten studies were included in the final review. Overall, effect sizes for non-modifiable
correlates of physical activity were small to moderate, except for mental health (d = 1.35) and
prior physical activity (d = 0.63). By contrast, modifiable theoretical correlates of physical activ-
ity (e.g., self-efficacy [d = 0.96–1.42] & intention to be physically active [d = 1.62]) had moderate
to large effects in the expected direction with physical activity during pregnancy.
Discussion: The findings underscore the importance of developing physical activity interven-
tions for pregnant women that are guided by modifiable theoretical correlates, taking into
consideration non-modifiable correlates of physical activity.
Linking Evidence to Action: Clinicians should help pregnant women to increase self-confi-
dence in their ability to be physically active and provide anticipatory guidance to overcome
barriers to physical activity.

BACKGROUND AND INTRODUCTION
Just over half (50.4%) of women in childbearing years (18–
44 years of age) meet the 2008 Physical Activity Guidelines
for Americans of 150 min of moderate-intensity physical
activity weekly (Robbins et al., 2018). These guidelines in-
clude all accumulated physical activity in the domains of
household, leisure, occupational, and transportation, not
just planned exercise. Physical activity during pregnancy is
associated with numerous health benefits, including better
mental health, reduced physical discomforts of pregnancy,
and reduced risk of increasingly common complications of
pregnancy such as gestational diabetes and preeclampsia
(Aune, Saugstad, Henriksen, & Tonstad, 2014; Sanabria-
Martínez et al., 2015; Yeo & Davidge, 2008). Despite clear
benefits and 2008 physical activity guidelines to maintain
physical activity during pregnancy (health.gov, 2008), most
pregnant women reduce their physical activity (Currie
et al., 2013). Estimates suggest that fewer than 15% of preg-
nant women achieve recommended levels of physical activ-
ity (Pearce, Evenson, Downs, & Steckler, 2013).
Numerous attempts have been made to improve wom-
en’s physical activity participation during pregnancy. Two
reviews reported on 12 clinical trials of interventions to
increase physical activity during pregnancy (Currie et al.,
2013; Pearce et al., 2013). Only five showed improvements.
Interestingly, only five were guided by theory, and three of
those showed improvements (using transtheoretical model,
social cognitive theory, & protection motivation theory;
Pearce et al., 2013). Findings suggested that significant im-
provement in physical activity during pregnancy was twice
as likely (60% [3/5] vs. 29% [2/7]) if the intervention was
guided by theory.
Many interventions fail to consider the unique charac-
teristics that influence physical activity in pregnant women.

Worldviews on Evidence-Based Nursing, 2019; 1–9. 1

© 2019 Sigma Theta Tau International
Correlates of Pregnancy Physical Activity
Successful change in physical activity behavior requires
understanding the non-modifiable (cannot be changed)
correlates of physical activity (e.g., demographic, personal
health, & pregnancy characteristics), and modifiable theo-
retical correlates (can be changed; Conn, Hafdahl, & Mehr,
2011). Gaston and Cramp’s (2011) review of 25 studies that
examined physical activity correlates found non-modifiable
correlates associated with greater physical activity during
pregnancy, including demographic (e.g., higher education
and income, White, & not having children in the home)
and personal health characteristics (e.g., being more physi-
cally active prior to pregnancy). There is also evidence that
high prepregnancy BMI (another personal health character-
istic) is associated with lower physical activity levels during
pregnancy (Seneviratne, McCowan, Cutfield, Derraik, &
Hofman, 2015). Other non-modifiable pregnancy charac-
teristics associated with lower physical activity levels during
pregnancy include increased gestational age (Evenson &
Wen, 2011) and experiencing pregnancy-related discom-
forts, such as nausea (Doran & Davis, 2011).
As of Gaston and Cramp’s review (2011), only 8 of their
25 studies investigated modifiable theoretical correlates of
physical activity. They noted that self-efficacy (confidence
in one’s ability to be physically active) was associated with
performing physical activity during pregnancy. Two recent
reviews examined physical activity during pregnancy ex-
clusively in relation to modifiable theoretical correlates of
physical activity. A meta-analysis of studies that used the
theory of planned behavior found that intention (readi-
ness to be more physical active) and perceived behavioral
control (perceived ease or difficulty performing physical
activity) both had positive relationships with physical ac-
tivity during pregnancy (De Vivo, Hulbert, Mills, & Uphill,
2016). A systematic review by Thompson, Vamos, and
Daley (2017) focused on studies guided by health behav-
ior theory. Most of the modifiable theoretical correlates of
physical activity that had negative relationships with phys-
ical activity performance were barrier-focused, including
normal discomforts of pregnancy, as well as non-health-
related issues, such as lack of motivation or self-efficacy.
The reviews by Gaston and Cramp (2011) and Thompson
et al. (2017) did not estimate the effect sizes across studies.
Thus, it is difficult to draw conclusions about the superior
fit of one theory or modifiable theoretical correlate. Further,
prior reviews offered limited evaluation of studies’ meth-
odological quality or internal validity. A systematic review
of original research of the correlates of physical activity that
includes measures of effect sizes for both non-modifiable
correlates and modifiable theoretical correlates and assesses
methodical quality is needed.
AIMS
The aim of this study was to examine systematically the
associations of physical activity during pregnancy with
2 Worldviews on Evidence-Based Nursing, 2019; 1–9.
non-modifiable correlates and modifiable theoretical cor-
relates of physical activity. The focus of this review was on
nonexperimental studies published since the 2008 Physical
Activity Guidelines for Americans (health.gov., 2008). We
report on the following: (a) the strength of association of
physical activity during pregnancy with non-modifiable
and modifiable theoretical correlates of physical activity and
(b) the methodological quality of the studies or risk of bias
that may have influenced results across and within studies.
PICOT QUESTION
Consistent with nonexperimental studies, only P (popula-
tion), I (indicator), and O (outcome of interest), but not C
(comparison/control) or T (types of studies) of PICOT were
addressed (Liberati, 2009). In pregnant women (P) what
is the association between non-modifiable and theoretical
correlates of physical activity (I) and physical activity be-
havior (O)?
METHODS
Design
This systematic review identified and appraised nonex-
perimental studies published between January 1, 2009, and
December 31, 2017 (i.e., since the 2008 Physical Activity
Guidelines for Americans (health.gov., 2008)) that examined
associations of physical activity during pregnancy with non-
modifiable and modifiable theoretical correlates of physical
activity. This systematic review incorporates observational
studies intended to identify association or correlation be-
tween variables (Joanna Briggs Institute; Joannabriggs.org,
2017). PRISMA (preferred reporting items for systematic
review & meta-analysis) guidelines for reporting were fol-
lowed (Liberati, 2009). Joanna Briggs (Joannabriggs.org,
2017) critical appraisal tools were used to appraise the re-
search evidence and assess methodological quality.
Study Selection
Inclusion criteria for this review were that the publica-
tion: (a) be a peer-reviewed research study published in
English, (b) use a cross-sectional or longitudinal design,
(c) include a sample of healthy pregnant women, (d) in-
clude at least one measure of recent physical activity during
pregnancy, (e) include modifiable theoretical correlates
of physical activity, and (f) examine the associations of
physical activity during pregnancy with non-modifiable
and modifiable theoretical correlates of physical activity.
Studies were excluded if they (a) had an intervention to
promote physical activity or (b) included pregnant women
with health problems or obstetrical risks.
Search Method
The literature search was conducted in PubMed, PsycINFO,
Cumulative Index of Nursing and Allied Health Literature
© 2019 Sigma Theta Tau International
 Evidence Review
(CINAHL), Scopus, and Cochrane. With assistance from Table 1.  Demographic Determinants and
a reference librarian, all searches used (MeSH terms) Physical Activity Behavior: Interpretation/
“Physical activity” OR “Exercise” AND (keywords) Significance and Effects Size
“Pregnant” AND “Pregnancy” AND “Behavior.” The fol-
lowing keywords were added and removed one at a time: Interpretation/ Effect
“Correlates”; “Barriers”; “Facilitators”; “Determinants”; Author significance size d
“Social support”; “Self-efficacy”; “Attitude”; “Motivation”; Age
“Environment;” and “Socio-economic.”  Haakstad ↑ Age ↓ physical −0.65
et al. (2013) activity*
Search Outcome  Redmond ↑ Age ↑ physical 0.20
The initial search resulted in 984 titles, of which 383 were et al. (2015) activity (ns)
duplicates (Figure S1). Titles of the remaining 601 were
 Ribeiro and Age > 25 ↓ physical −0.07
read to determine whether they met inclusion criteria; 535 Milanez activity (ns)
did not. The most common reason was that the paper did (2011)
not measure potentially modifiable theoretical correlates of Education
physical activity. Next, abstracts of the remaining 66 papers
 Haakstad University (yes) ↑ 0.12
were reviewed, and 22 were excluded. The most common et al. (2013) physical activity (ns)
reason was that the study did not measure modifiable theo-
 Redmond College graduate (yes) 0.23
retical correlates of physical activity (n = 16). The full texts et al. (2015) ↑ physical activity (ns)
of the remaining 44 papers were reviewed independently
 Ribeiro and University (yes) ↑ 0.26
by two authors, who discussed papers where there was a Milanez physical activity**
lack of agreement until consensus was reached. Thirty-four (2011)
papers were excluded. Most did not examine modifiable Employment
theoretical correlates of physical activity (n = 29). Ten pa-
 Redmond Employed ≥ 20 hrs (yes) 0.45
pers met the inclusion criteria. et al. (2015) ↑ physical activity (ns)
 Ribeiro and Paid employment (yes) 0.02
Data Extraction and Synthesis Milanez ↑ physical activity (ns)
Study variables were abstracted using data collection tools (2011)
developed by the investigators and presented in narrative Income
tabular format. Initially, the studies were abstracted for
 Ribeiro and ↑ Income ↑ (physical 0.29
design, number subjects, theory, demographics (Table S1), Milanez activity ns)
then pregnancy characteristics, health behaviors, physical (2011)
activity measure (Table S2), then interpretation and signifi-
Note d = Cohen’s d; ns = not significant.
cance of non-modifiable and modifiable theoretical cor- *p < .01. **p ≤ .001.
relates, and calculation of effect size (Tables 1‒4). The 10
studies were also assessed for risk of bias across and within
studies. Two authors reviewed the 10 studies, coded results, This checklist assesses whether the study meets eight factors
cross-checked for coding inconsistencies, and obtained (Table S3): (a) criteria for inclusion in the sample clearly de-
consensus. fined, (b) study subjects and the setting described in detail,
(c) objective, standard criteria used for measurement of the
Analytical Strategies condition, (d) confounding factors identified, (e) strategies
Effect sizes were obtained for the correlates of physical to deal with confounding factors stated, (f) outcomes meas-
activity and physical activity during pregnancy using two ured in a standard, reliable way for all participants, finally,
open-source effect size calculators (Daniel & Kostic, 2017; and (g) was there appropriate statistical analysis. Factor 3
DeCoster, 2012). Reported correlations and associations were was not applicable to a study of healthy pregnant women;
converted to Cohen’s d (Rosenthal, Rosnow, & Rubin, 2000). therefore, only seven factors were assessed.
Effect sizes using Cohen’s d can be described as small (0.20),
medium (0.50), or large (0.80; Cohen, 1988). The following
statistical tests were performed: Pearson correlation, t tests, RESULTS
mean of two correlations, and chi-square test converted to Country, Study Design, Theory, and Demographic
Cramer’s V. Finally, risk of bias was assessed using the Joanna Physical Activity Correlates
Briggs Institute Critical Appraisal Checklist for Analytical Of the 10 studies, two were conducted in the United States,
Cross-Sectional Studies (Joannabriggs.org, 2017). This ap- three each from Canada and Europe, one each from Australia
praisal checklist was used to inform synthesis and interpre- and Brazil; six were cross-sectional; and four were longitu-
tation of the results of the study (Joannabriggs.org., 2017). dinal (Table S1). Sample sizes ranged from 38 (Hausenblas,

Worldviews on Evidence-Based Nursing, 2019; 1–9. 3

© 2019 Sigma Theta Tau International
Correlates of Pregnancy Physical Activity

Table 2.  Pregnancy Determinants and Physical Table 3.  Health Determinants and Physical
Activity Behavior: Interpretation/Significance Activity Behavior: Interpretation/Significance
and Effects Size and Effects Size

Interpretation/ Effect Interpretation/ Effect

Authors significance size d Author significance size d
Parity Mental health
 Haakstad Multiparity (yes) ↓ −0.15  De Wit et al. Good well-being ↑ physical 1.35
et al. physical activity* (2015) activity*
(2013)
Subsequent pregnancy ↓ −0.003 Prepregnancy exercise
physical activity**
 Da Costa Prepregnancy exercise ↑ 0.44
Gestational age and Ireland physical activity*
(2013)
 Ribeiro and ↑ Gestational age ↓ −0.41
Milanez physical activity***  Redmond Prepregnancy exercise (yes)↑ 0.62
(2011) et al. (2015) physical activity*
Discomforts Substance use
 Haakstad Pelvic girdle pain (yes) ↓ −0.16  Haakstad Daily smoking (yes) ↓ 0.00
et al. physical activity*** et al. (2013) physical activity (ns)
(2013)
Incontinence (yes) ↓ −0.18 Prepregnancy BMI
physical activity*
 Downs et al. Normal weight ↑ 2nd 0.25
 De Wit Few or no pregnancy 0.18 (2015) trimester physical activity (ns)
et al. worries ↑ physical
Normal weight ↑ 3rd trimester 0.19
(2015) activity (ns)
physical activity (ns)
Note d = Cohen’s d; ns = not significant.  Haakstad ↑ BMI ↓ physical activity** 0.82
*p < .01. **p < .05. ***p ≤ .001. et al. (2013)
BMI ≥ 25 ↓ physical 0.18
activity**
Giacobbi, Cook, Rhodes, & Cruz, 2011) to 467 (Haakstad,  Redmond Prepregnancy BMI ≥ 30 ↑ 0.62
Voldner, & Bø, 2013). Half of the studies had fewer than et al. (2015) physical activity (ns)
100 subjects. The studies were guided by five theories. The Note d = Cohen’s d; ns = not significant.
theory of planned behavior (four studies) and social cogni- *p < 0.05. **p ≤ 0.001.
tive theory (four studies) were used most often. Other theo-
ries included self-determination theory, the transtheoretical
model, and the extended parallel process model. 2009; Downs et al., 2015; Hausenblas et al., 2011). Four
The mean age of study participants ranged from 25.0 to studies examined discomforts of pregnancy (Cramp &
33.3 years (Table S1). In nine studies, 70% or more of par- Bray, 2009; Da Costa & Ireland, 2013; De Wit et al., 2015;
ticipants were Caucasian; ethnicity was not identified in one Haakstad et al., 2013), and two examined mental health
study. In the five studies that reported marital status, most (Da Costa & Ireland, 2013; De Wit et al., 2015).
subjects (≥70%) were married. In six of seven studies report- Three studies assessed prepregnancy self-reported phys-
ing employment status, 55% or more were employed. Six ical activity (Da Costa & Ireland, 2013; Gaston et al., 2013;
studies reported incomes representing three different mon- Redmond et al., 2015); two assessed substance use (De Wit
etary denominations. Most (Cramp & Bray, 2009; Da Costa et al., 2015; Haakstad et al., 2013); and three reported pre-
& Ireland, 2013; Downs, Devlin, & Rhodes, 2015; Ribeiro & pregnancy BMI (De Wit et al., 2015; Downs et al., 2015;
Milanez, 2011) fell between 40,000 and 100,000 USD which Haakstad et al., 2013).
is considered the middle-income range (Geiger, 2018). Nine of 10 studies used self-report measures of physical
activity. Only one self-report measure captured lifestyle phys-
Pregnancy Correlates, Health Correlates, and ical activity specific to pregnancy (Hausenblas et al., 2011);
Physical Activity Measures one other used a general self-report measure of lifestyle
The number of subjects who had never given birth varied physical activity (Mullan et al., 2016). The remaining seven
widely, from 34% to 67% (Table S2). Three of the six cross- studies had self-report measures of purposeful physical ac-
sectional studies had participants from more than one tivity, including sport, recreation, and leisure time (Cramp &
pregnancy trimester (De Wit et al., 2015; Gaston, Wilson, Bray, 2009; Downs et al., 2015; Gaston et al., 2013; Haakstad
Mack, Elliot, & Prapavessis, 2013; Redmond, Dong, & et al., 2013; Redmond et al., 2015; Ribeiro & Milanez, 2011).
Frazier, 2015). Three of the four longitudinal studies fol- Only one of the 10 used accelerometry to measure lifestyle
lowed participants through all trimesters (Cramp & Bray, physical activity objectively (De Wit et al., 2015).

4 Worldviews on Evidence-Based Nursing, 2019; 1–9.

© 2019 Sigma Theta Tau International
 Evidence Review
Table 4.  Theoretical Determinants and Physical Activity: Interpretation/Significance and Effects Size

Effect size d (single

Author Interpretation/significance effect or range)
Social cognitive theory
 Cramp and Bray (2009) ↑ Barrier self-efficacy (confidence overcoming barriers) ↑ 0.96–1.27
physical activity *
↑ Exercise self-efficacy (confidence ability to exercise) ↑ 1.04–1.42
physical activity*
Theory of planned behavior
 Downs et al. (2015) ↑ Behavioral beliefs (enjoyable, health benefits)2nd trimester 0.18–0.57
↑ physical activity **
↑ Behavioral beliefs 3rd trimester ↑ physical activity ** 0.25–0.40
↑ Control beliefs (ease adopting physical activity) 2nd 0.23–0.85
trimester ↑ physical activity *
↑ Control beliefs 3rd trimester ↑ physical activity * 0.08–0.56
↑ Normative beliefs (social pressure to be physically active) 0.18–.27
2nd trimester ↑ physical activity (ns)
↑ Normative beliefs 3rd trimester ↑ physical activity ** 0.17–0.42
Self-determination theory
 Gaston et al. (2013) ↑ Identified regulation (valuing benefits of exercise) ↑ 1.09
physical activity *
↑ Intrinsic regulation (exercise fun) ↑ physical activity * 0.92
↑ External regulation (exercise because told to) ↑ physical 0.10
activity (ns)
↑ Introjected regulation (exercise because feel guilty) ↑ 0.38
physical activity (ns)
↑ Barriers (e.g., weight, time) ↓ physical activity * −1.00
Extended parallel process model
 Redmond et al. (2015) ↑ Response efficacy (control harm to fetus) ↓ physical 0.08
activity (ns)
↑ Self-efficacy ↑ physical activity *** 0.69
↑ Susceptibility to threat (perceived amount of risk to fetus)↓ −0.39
physical activity **
↑ Severity of threat (perceived amount of harm to fetus)↓ −0.67
physical activity ***
Perceived behavioral control, intention, habit
 Mullan et al. (2016) ↑ Perceived behavioral control (confidence ability to engage 1.31
in physical activity) ↑ physical activity **
↑Intention ↑ physical activity * 1.62
↑ Habit (prior physical activity) ↑ physical activity * 1.53
Note d = Cohen’s d; ns = not significant.
*p < 0.01. **p < 0.05. ***p ≤ 0.001.

Relationship Between Physical Activity During
Pregnancy and Physical Activity Correlates
Three of the 10 studies examined the relationship of physi-
cal activity during pregnancy with demographic char-
acteristics (Table 1). The three studies that examined age
had mixed findings (two negative & one positive effect;
Haakstad et al., 2013; Redmond et al., 2015; Ribeiro &
Milanez, 2011). The same three studies found higher edu-
cational achievement had small positive effects. Being em-
ployed and having higher income were associated with
small positive effects (Redmond et al., 2015; Ribeiro &
Milanez, 2011).

Worldviews on Evidence-Based Nursing, 2019; 1–9. 5

© 2019 Sigma Theta Tau International
Correlates of Pregnancy Physical Activity
Only three of the 10 studies examined the relation-
ship of physical activity during pregnancy with pregnancy
characteristics (De Wit et al., 2015; Haakstad et al., 2013;
Hausenblas et al., 2011; Table 2). Multiparity, unpleasant
pregnancy symptoms, and greater gestational age were all
associated with small negative effects on physical activity
during pregnancy.
Five of the 10 studies examined the relationship of
physical activity during pregnancy with health character-
istics (Da Costa & Ireland, 2013; De Wit et al., 2015; Downs
et al., 2015; Haakstad et al., 2013; Redmond et al., 2015;
Table 3). The one study that examined mental health and
physical activity during pregnancy used the World Health
Organization measure of mood (five uni-directional state-
ments of positive well-being rated on a 1–6 Likert scale of
agreement) and found a large association (De Wit et al.,
2015). Better mental health was associated with higher
physical activity. In addition, large associations were found
with higher prepregnancy physical activity (Redmond
et al., 2015), lower prepregnancy BMI (Haakstad et al.,
2013; Redmond et al., 2015), and higher physical activity
during pregnancy.
Five of the 10 studies examined the relationships of
physical activity during pregnancy with modifiable the-
oretical correlates (Cramp & Bray, 2009; Downs et al.,
2015; Gaston et al., 2013; Mullan et al., 2016; Redmond
et al., 2015 Table 4). Although theories differed across
studies, some modifiable theoretical correlates exam-
ined were similar. Three studies looked at self-efficacy
or confidence (Cramp & Bray, 2009; Mullan et al., 2016;
Redmond et al., 2015). They showed moderate to large
positive effects between physical activity during preg-
nancy and confidence in overcoming barriers to physical
activity (n = 1), confidence in ability to exercise (n = 2),
confidence in ability to control harm to the fetus (n = 1),
and confidence in ability to safely exercise (n = 1). Large
positive effects were also seen between physical activ-
ity during pregnancy and high levels of both identified
regulation (valuing the benefits of exercise) and intrin-
sic regulation (e.g., exercise is fun; Gaston et al., 2013).
Likewise, another study showed moderate positive effects
between physical activity during pregnancy and high lev-
els of both “behavior beliefs that physical activity was en-
joyable” and “beliefs about health benefits from physical
activity” (Downs et al., 2015).
Higher perceived barriers (Gaston et al., 2013) to
exercising during pregnancy and higher perceived risk
and severity of harm to the fetus had moderate to strong
negative effects with physical activity during pregnancy
(Redmond et al., 2015). Further, moderate positive ef-
fects with physical activity during pregnancy were seen
with high levels of intention to be physically active, high
participation in physical activity prior to pregnancy, and
high levels of control beliefs (Downs et al., 2015; Mullan
6 Worldviews on Evidence-Based Nursing, 2019; 1–9.
et al., 2016). All of these effects were in the expected
directions.
Methodological Quality
Four studies received a positive score on two of the seven
Briggs criteria (Table S3): confounding factors identified,
and strategies for dealing with confounders stated (Da
Costa & Ireland, 2013; de Wit et al., 2015; Gaston et al.,
2013; Redmond et al., 2015). No study obtained a perfect
score on all seven criteria of quality assessment. Five of the
10 studies had scores below 50% (Cramp & Bray, 2009;
Downs et al., 2015; Hausenblas et al., 2011; Mullan et al.,
2016; Ribeiro & Milanez, 2011).
DISCUSSION
Overall, findings indicated that non-modifiable correlates
had predominantly weak effects, while modifiable theo-
retical correlates had larger effects with physical activity
during pregnancy. Consistent with Gaston and Cramp
(2011), our findings indicated that the demographic cor-
relates of higher education and higher income were posi-
tively associated with physical activity during pregnancy.
Also consistent with Gaston and Cramp (2011), the preg-
nancy characteristics most associated with physical activity
during pregnancy were fewer previous pregnancies, earlier
gestational age, and fewer discomforts of pregnancy.
The health characteristics most associated with physical
activity during pregnancy were better mental health and
being more physically active prior to pregnancy. Better men-
tal health was associated with higher levels of physical activ-
ity during pregnancy. Although additional longitudinal study
is needed, this finding suggests that increasing physical activ-
ity during pregnancy may well have a positive impact on the
woman’s health. Further, according to Ohara (2009), women
with better mental health during pregnancy are less likely to
have postpartum depression. Higher levels of physical activ-
ity prior to pregnancy are also associated with higher levels
of physical activity during pregnancy. This suggests the im-
portance of encouraging women to become more physically
active prior to pregnancy rather than waiting until they are
pregnant to start a new behavior which is more challenging.
Although all studies reviewed used a theoretical frame-
work, only half examined the relationships of physical
activity during pregnancy with modifiable theoretical cor-
relates. Consistent with the review by Gaston and Cramp
(2011), we found that physical activity during pregnancy
was associated with increased self-efficacy (confidence)
and higher belief in the health benefits and enjoyment of
physical activity. Greater perceived susceptibility to threat
of risk and perceived threat of harm to the fetus caused by
physical activity were associated with lower physical activ-
ity during pregnancy. In addition, consistent with earlier
studies, greater perceived barriers (e.g., pregnancy weight
© 2019 Sigma Theta Tau International

gain or time to exercise) were associated with lower phys-
ical activity during pregnancy (Gaston & Cramp, 2011;
Thompson et al., 2017). The large effects on physical ac-
tivity during pregnancy of these modifiable theoreti-
cal correlates of physical activity suggest the importance
of addressing them in interventions. Interventions are
needed that increase a pregnant woman’s (a) confidence
in her ability to be physically active, (b) understanding
of the substantial benefits and minimal risks of physical
activity to herself and her fetus, and (c) ability to manage
discomforts that are barriers to physical activity during
pregnancy. Importantly, more research is needed on deter-
mining what physical activities may be most enjoyable for
pregnant women.
There were several issues related to physical activity mea-
surement in the studies reviewed. First, most studies focused
on purposeful physical activity, including sport, recreation,
and leisure time exercise (Cramp & Bray, 2009; Da Costa
& Ireland, 2013; Downs et al., 2015; Haakstad et al., 2013;
Redmond et al., 2015; Ribeiro & Milanez, 2011). Thus, find-
ings may have underestimated total lifestyle physical activ-
ity. Only one study used a tool validated to measure lifestyle
physical activity specific to pregnancy (Hausenblas et al.,
2011). To develop interventions, a better understanding is
needed of how the pregnancy experience can influence life-
style physical activity and change through trimesters.
The Briggs scale analyses revealed issues related to lack
of objective measurement of physical activity. Self-report
measures of lifestyle physical activity are often overesti-
mated (Bell et al., 2013), and most studies reviewed used
self-report measures. Only one study used an accelerome-
ter to measure lifestyle physical activity objectively (De Wit
et al., 2015). Future studies that examine the correlates of
physical activity during pregnancy should use both self-re-
port and objective measures to generate a more accurate
picture of total activity.
The most notable limitation of these studies was that
samples consisted of predominantly Caucasian, middle-in-
come participants. Little is known about physical activity
during pregnancy in minority women or women with
lower socioeconomic status. In addition, findings indicate
that the methodological quality of these studies, based on
the Briggs scale, was somewhat limited.
Additional preliminary research is needed to understand
better the unique needs of pregnant women from diverse
ethnic groups and socioeconomic status. Physical activity in-
terventions may need to be adjusted throughout the course
of pregnancy to accommodate changes in pregnancy-re-
lated discomforts. Interventions must be developed to target
young women during their childbearing years to boost the
number of women who enters pregnancy with good expe-
riences and motivation to remain physically active.
Like Gaston and Cramp (2011), we identified a paucity
of longitudinal studies. Cross-sectional studies are unable
to detect changes over the course of pregnancy that may
Worldviews on Evidence-Based Nursing, 2019; 1–9.
© 2019 Sigma Theta Tau International
Evidence Review
influence physical activity (Thompson et al., 2017). In ad-
dition, identification of and controlling for potential con-
founders will strengthen methodological quality. Future
studies should examine the relationships among non-mod-
ifiable correlates, modifiable theoretical correlates, and
physical activity behavior within the same study.
Limitations
This review has several limitations. We excluded qualitative
studies, which may explore different factors that influence
physical activity during pregnancy and differ from those
factors captured in quantitative studies. Additionally, we
excluded women with pregnancy complications; thus, no
conclusions can be drawn regarding correlates of physical
activity in that population.
CONCLUSIONS
Physical activity is a complex behavior, and the dynamic
physical and psychological changes during pregnancy
create additional challenges for developing effective in-
terventions. Findings of this review underscore the im-
portance of developing physical activity interventions for
pregnant women that are guided by a theoretical frame-
work and consider non-modifiable correlates of physical
activity. Future interventions should focus on increasing
total energy expenditure in all physical activity domains.
WVN
LINKING EVIDENCE TO ACTION
• Prior to conception, clinicians should encourage
women to establish physical activity behaviors.
• Clinicians should provide anticipatory guidance to
pregnant women about symptoms and changes in
body shape as pregnancy advances and suggest activity
modification, support garments, or alternative physi-
cal activities so they do not present barriers.
• Clinicians should help women to increase self-confi-
dence in their ability to be physically = active and
identify barriers to physical activity, and strategies to
overcome them.
• Clinicians should emphasize physical activity safety,
health benefits, and enjoyment of physical activity
during pregnancy. Physical activity fosters mental
well-being and may reduce rates of antenatal and post-
partum depression.
• Future research should include objective measures of
physical activity and subjective measures that capture
lifestyle physical activity.
7
Correlates of Pregnancy Physical Activity
Author information
Meghan Garland, Student, Rush University College of
Nursing, Chicago, IL, USA; JoEllen Wilbur, Professor, Rush
University College of Nursing, Chicago, IL, USA; Pamela
Semanik, Assistant Professor, Rush University College of
Nursing, Chicago, IL, USA; Louis Fogg, Associate Professor,
Rush University College of Nursing, Chicago, IL, USA
Address correspondence to Meghan Garland, Rush
University College of Nursing, 600 S Paulina St., Chicago,
IL 60612, USA; Meghan_Garland@rush.edu
Accepted 24 March 2019
© 2019 Sigma Theta Tau International
References
Aune, D., Saugstad, O., Henriksen, T., & Tonstad, S. (2014).
Physical activity and the risk of preeclampsia. Epidemiology,
25(3), 331–343. https​://doi.org/10.1097/EDE.00000​00000​
000036
Bell, R., Tennant, P., McPharlin, C., Pearce, M., Adamson, A.,
Rankin, J., & Robson, S. (2013). Measuring physical activity
in pregnancy: A comparison of accelerometry and self-com-
pletion questionnaires in overweight and obese women.
European Journal of Obstetrics & Gynecology and Reproductive Biology,
170(1), 90–95. https​://doi.org/10.1016/j.ejogrb.2013.05.018
Cohen, J. (1988). Statistical power analysis for the behavioral sciences.
Hillsdale, NJ: Lawrence Erlbaum Associates.
Conn, V., Hafdahl, A., & Mehr, D. (2011). Interventions to
increase physical activity among healthy adults: Meta-
Analysis of outcomes. American Journal of Public Health, 101(4),
751–758. https​://doi.org/10.2105/AJPH.2010.194381
Cramp, A. G., & Bray, S. R. (2009). A prospective examina-
tion of exercise and barrier self-efficacy to engage in lei-
sure-time physical activity during pregnancy. Annals of
Behavioral Medicine, 37(3), 325–334. https​://doi.org/10.1007/
s12160-009-9102-y
Currie, S., Sinclair, M., Murphy, M., Madden, E., Dunwoody,
L., & Liddle, D. (2013). Reducing the decline in physical ac-
tivity during pregnancy: A systematic review of behaviour
change interventions. PLoS ONE, 8(6), e66385. https​://doi.
org/10.1371/journ​al.pone.0066385
Da Costa, D., & Ireland, K. (2013). Perceived benefits and
barriers to leisure-time physical activity during preg-
nancy in previously inactive and active women. Women
and Health, 53(2), 185–202. https​://doi.org/10.1080/03630​
242.2012.758219
Daniel, T., & Kostic, B. (2017). RStats effect size calculator for t tests.
Retrieved from http://www.misso​urist​ate.edu/rstat​s/
Tables-and-Calcu​lators.htm
De Vivo, M., Hulbert, S., Mills, H., & Uphill, M. (2016).
Examining exercise intention and behaviour during
pregnancy using the Theory of Planned Behaviour: A
8 Worldviews on Evidence-Based Nursing, 2019; 1–9.
meta-analysis. Journal of Reproductive and Infant Psychology, 34,
122–138. https​://doi.org/10.1080/02646​838.2015.1118022
De Wit, L., Jelsma, J., van Poppel, M., Bogaerts, A., Simmons,
D., Desoye, G., … Snoek, F. (2015). Physical activity, de-
pressed mood and pregnancy worries in European obese
pregnant women: Results from the DALI study. BMC
Pregnancy and Childbirth, 15(1), 158. https​://doi.org/10.1186/
s12884-015-0595-z
DeCoster, J. (2012). Converting effect sizes. Retrieved from http://
www.stat-help.com/sprea​ d shee​ t s/Conve​ r ting​ %20eff​
ect%20siz​es%202012-06-19.xls.
Doran, F., & Davis, K. (2011). Factors that influence physical
activity for pregnant and postpartum women and impli-
cations for primary care. Australian Journal of Primary Health,
17(1), 79. https​://doi.org/10.1071/py10036
Downs, D., Devlin, C., & Rhodes, R. (2015). The power of
believing: Salient belief predictors of exercise behavior in
normal weight, overweight, and obese pregnant women.
Journal of Physical Activity and Health, 12(8), 1168–1176. https​://
doi.org/10.1123/jpah.2014-0262
Evenson, K. R., & Wen, F. (2011). Prevalence and correlates
of objectively measured physical activity and seden-
tary behavior among US pregnant women. Preventive
Medicine, 53(1–2), 39–43. https​ ://doi.org/10.1016/j.
ypmed.2011.04.014
Gaston, A., & Cramp, A. (2011). Exercise during pregnancy:
A review of patterns and determinants. Journal of Science and
Medicine in Sport, 14(4), 299–305. https​://doi.org/10.1016/j.
jsams.2011.02.006
Gaston, A., Wilson, P. M., Mack, D. E., Elliot, S., &
Prapavessis, H. (2013). Understanding physical activity
behavior and cognitions in pregnant women: An appli-
cation of self-determination theory. Psychology of Sport and
Exercise, 14(3), 405–412. https​
://doi.org/10.1016/j.psych​
sport.2012.12.009
Geiger, A. (2018, September 06). Are you in the American mid-
dle class?Retrieved from http://www.pewre​ search.
org/fac t-t a n k/2018/09/06/a re-you-i n-t he-a me r i​
can-middle-class/​
Haakstad, L., Voldner, N., & Bø, K. (2013). Stages of change
model for participation in physical activity during
pregnancy. Journal of Pregnancy, 2013, 1–7. https​ ://doi.
org/10.1155/2013/193170
Hausenblas, H., Giacobbi, P., Cook, B., Rhodes, R., & Cruz,
A. (2011). Prospective examination of pregnant and non-
pregnant women's physical activity beliefs and behaviours.
Journal of Reproductive and Infant Psychology, 29(4), 308–319. https​
://doi.org/10.1080/02646​838.2011.629993
health.gov. (2008). 2008 physical activity guidelines for Americans.
Washington, DC: Author. Retrieved from https​://health.
gov/pagui​delin​es/guide​lines/​
Joannabriggs.org. (2017). Critical appraisal tools-JBI. Retrieved
form https​://joann​abrig​gs.org/criti​cal_appra​isal_tools​
Liberati, A. (2009). The PRISMA statement for report-
ing systematic reviews and meta-analyses of studies
© 2019 Sigma Theta Tau International

that evaluate health care interventions: Explanation
and elaboration. Annals of Internal Medicine, 151(4), 265–
267. https​://doi.org/10.7326/0003-4819-151-4-20090​
8180-00136​
Mullan, B., Henderson, J., Kothe, E., Allom, V., Orbell, S.,
& Hamilton, K. (2016). The role of habit and perceived
control on health behavior among pregnant women.
American Journal of Health Behavior, 40(3), 291–301. https​://doi.
org/10.5993/AJHB.40.3.1
Ohara, M. W. (2009). Postpartum depression: What we know.
Journal of Clinical Psychology, 65(12), 1258–1269. https​
://doi.
org/10.1002/jclp.20644​
Pearce, E., Evenson, K., Downs, D., & Steckler, A. (2013).
Strategies to promote physical activity during pregnancy.
American Journal of Lifestyle Medicine, 7(1), 38–50. https​://doi.
org/10.1177/15598​27612​446416
Redmond, M., Dong, F., & Frazier, L. (2015). Does the extended
parallel process model fear appeal theory explain fears and
barriers to prenatal physical activity? Women's Health Issues,
25(2), 149–154. https​://doi.org/10.1016/j.whi.2014.11.009
Ribeiro, C., & Milanez, H. (2011). Knowledge, attitude
and practice of women in Campinas, São Palo, Brazil
with respect to physical exercise in pregnancy: A de-
scriptive study. Reproductive Health, 8(1), 1–7. https​
://doi.
org/10.1186/1742-4755-8-31
Robbins, C., Boulet, S. L., Morgan, I., D'Angelo, D. V., Zapata,
L. B., Morrow, B., … Kroelinger, C. D. (2018). Disparities
in preconception health indicators — Behavioral Risk
Factor Surveillance System, 2013–2015, and Pregnancy
SUPPORTING INFORMATION
Additional supporting information may be found in the online version of this article at the publisher’s web site:
Table S1. Author, Country, Study Design, Number of Subjects, Theory, and Demographic Correlates of Physical Activity
Table S2. Pregnancy Characteristics, Health Characteristics, and Physical Activity Measure
Table S3. Met Criteria for Methodological Quality by Study and Criteria
Figure S1. Flow chart of search and retrieval process.
Worldviews on Evidence-Based Nursing, 2019; 1–9.
© 2019 Sigma Theta Tau International
Evidence Review
Risk Assessment Monitoring System, 2013–2014. Surveillance
Summaries, 67(01), 1–16.
Rosenthal, R., Rosnow, R. L., & Rubin, D. B. (2000). Contrasts and
effect sizes in behavioral research: A correlational approach. Cambridge,
England: Cambridge University Press.
Sanabria-Martínez, G., García-Hermoso, A., Poyatos-León,
R., Álvarez-Bueno, C., Sánchez-López, M., & Martínez-
Vizcaíno, V. (2015). Effectiveness of physical activity inter-
ventions on preventing gestational diabetes mellitus and
excessive maternal weight gain: A meta-analysis. BJOG: An
International Journal of Obstetrics & Gynaecology, 122(9), 1167–1174.
https​://doi.org/10.1111/1471-0528.13446​
Seneviratne, S. N., McCowan, L. M., Cutfield, W. S., Derraik,
J. G., & Hofman, P. L. (2015). Exercise in pregnancies com-
plicated by obesity: Achieving benefits and overcoming
barriers. American Journal of Obstetrics and Gynecology, 212(4),
442–449. https​://doi.org/10.1016/j.ajog.2014.06.009
Thompson, E. L., Vamos, C. A., & Daley, E. M. (2017). Physical
activity during pregnancy and the role of theory in pro-
moting positive behavior change: A systematic review.
Journal of Sport and Health Science, 6(2), 198–206. https​://doi.
org/10.1016/j.jshs.2015.08.001
Yeo, S., & Davidge, S. (2008). Walking versus stretching exer-
cises to reduce the incidence of preeclampsia: A randomized
clinical trial. Medicine & Science in Sports & Exercise, 40(Supp.), S107.
https​://doi.org/10.1249/01.mss.00003​21915.24035.f3
10.1111/wvn.12391
WVN 2019;0:1–9
9