Abstract
Background:
Several features of the neighborhood built environment have been shown to promote leisure-time physical activity (PA) in the general population, but few studies have examined its impact on PA during pregnancy.
Methods:
Data came from 8,362 Nulliparous Pregnancy Outcomes Study: Monitoring Mothers-to-Be (nuMoM2b) cohort participants (2010–2013). Residential address information was linked to three built-environment characteristics: number of gyms and recreation areas within a 3-km radius of residence and census block-level walkability. Self-reported leisure-time PA was measured in each trimester and dichotomized as meeting PA guidelines or not. Relative risks for cross-sectional associations between neighborhood characteristics and meeting PA guidelines were estimated using Poisson regression.
Results:
More gyms and recreation areas were each associated with a greater chance of meeting PA guidelines in models adjusted for sociodemographic characteristics and pre-existing conditions. Associations were strongest in the third trimester, where each doubling in counts of gyms and recreation areas was associated with 10% (95% CI: 1.07, 1.13) and 8% (95% CI: 1.03, 1.12), respectively, greater likelihood of meeting PA guidelines. Associations were similar though weaker for walkability.
Conclusions:
Results from a large, multi-site cohort suggest these built-environment characteristics have similar PA-promoting benefits in pregnant women as seen in more general populations.
INTRODUCTION
Regular leisure-time physical activity is important for overall health and well-being throughout the life course, including during pregnancy. Several studies have shown that women who engage in regular moderate-intensity physical activity (e.g., brisk walking or stationary cycling) have lower risk of excessive gestational weight gain, gestational diabetes mellitus, and symptoms of postpartum depression than those who do not 1–3. However, studies indicate pregnant women are less active than non-pregnant women and that pregnancy tends to lead to a decrease in physical activity 4. According to recent estimates, fewer than 30% of pregnant women meet recommended guidelines for physical activity (150 minutes of moderate activity per week, at least 75 minutes of vigorous activity per week, or an equivalent combination of the two) 5–8. Thus, it is important to understand what factors promote physical activity during pregnancy.
Several features of the neighborhood built environment, which have been shown to promote physical activity in the general population, could also impact physical activity during pregnancy. A recent review on the built environment and physical activity found that higher walkability and quality parks were associated with higher physical activity 9. There is also evidence that greater availability of exercise facilities is associated with higher levels of physical activity 10,11.
Few studies have examined associations of the built environment with physical activity during pregnancy. A multi-ethnic cohort study of 709 pregnant women living in Oslo, Sweden found that women with better objective and perceived access to recreational areas had significantly higher levels of objectively-measured daily physical activity 12. A study of 417 pregnant women living in four counties in North Carolina looked at 11 measures of the neighborhood built environment in relation to self-reported physical activity assessed late in pregnancy and postpartum 13. Those investigators found several features of the neighborhood environment were associated with physical activity, including walkability, access to transit, distance to recreation facilities, and road networks.
In this study, we built on the existing literature by examining which features of the neighborhood physical activity environment were associated with adherence to physical activity guidelines during pregnancy in a larger and more geographically diverse cohort than has been studied previously. We examined cross-sectional associations in each trimester of pregnancy using data from the Nulliparous Pregnancy Outcomes Study: Monitoring Mothers-to-Be (nuMoM2b) cohort of over 10,000 nulliparous women recruited from 8 clinical centers across the US. We hypothesized that women living in neighborhoods with better availability of physical activity-promoting resources would have a higher chance of adherence to physical activity guidelines in all trimesters. Findings from this study will help enhance our understanding of potential barriers to physical activity during pregnancy.
METHODS
Study Sample
The nuMoM2b cohort has been previously described 14. Briefly, 10,038 nulliparous women age 13+ years with singleton pregnancies from 60 to 136 weeks gestational age were enrolled through 8 primary clinical centers, many of which recruited at smaller regional locations (with a total of 18 subsites), from October 2010 through September 2013 and prospectively followed through pregnancy. In-person study visits at four timepoints (including delivery) involved the collection of self-reported sociodemographic, medical history, behavioral, and psychosocial information (including residential addresses); blood and other biospecimen collection; and rigorous chart abstraction of pregnancy characteristics and outcomes. The study protocol was approved by local site IRBs and registered as NCT #01322529. Informed consent was obtained from all participants.
This secondary data analysis included nuMoM2b participants of age 18 years or greater who provided valid physical activity reports and high-quality address data at one or more timepoints during pregnancy, who carried the pregnancy to at least 20 weeks gestational age, and who were not missing covariate data. Participants under the age of 18 were excluded because the nature and determinants of their leisure-time physical activity behaviors (e.g., compulsory physical education class in a school setting) is likely different from those who are 18 and older. Physical activity reports were considered invalid if the reports of exercise performed were at levels not considered physically achievable by humans for the exercise reported (e.g., reports of more than 24 hours of exercise in a day were considered invalid). Address information was considered low-quality if it could not be successfully matched to a known building address or street. Figure 1 details the participant inclusion flow for this analysis.
Figure 1-.
Flow diagram for participation in analysis.
Measures
The outcome of interest was leisure-time physical activity during pregnancy. This was measured during in-person study visits at three timepoints (early pregnancy, 60 – 136 weeks; 2nd trimester, 160 – 216 weeks; 3rd trimester, 220 – 296 weeks) by a trained interviewer using standardized physical activity questions adapted from the Behavior Risk Factor Surveillance System (BRFSS) 15. At each timepoint, women were asked whether they participated in any leisure-time physical activity over the prior four weeks. If they responded yes, they were asked to describe the three activities they spent the most time doing. They were asked to provide the number of times per week they engaged in that activity and how many minutes per time. For running, jogging, walking, cycling, and swimming, there were also asked about distance. See Supplemental Display 1 for the physical activity questionnaire. The types of physical activity were coded based on the Physical Activity Compendium 16 to assign intensity in metabolic equivalents (METs). Participants engaging in at least 150 minutes of moderate activity per week, at least 75 minutes of vigorous activity per week, or an equivalent combination of the two were considered to have met the U.S. Department of Health and Human Services guidelines for physical activity 5,7.
Walkability, number of gyms, and number of recreation areas were the three built-environment exposures examined this study. Participant residential addresses at each timepoint were converted to standard formats and geocoded using ESRI ArcGIS 17 (Version 10.5, Redlands CA). Locations were then linked to the three built-environment characteristics. Neighborhood walkability was measured using the 2010 National Walkability Index derived at the census block group level from the Smart Location Database of the U.S. Environmental Protection Agency 18. The National Walkability Index ranks block groups according to how much their features promote walking trips. This includes land use mix, mix of employment types (e.g., retail, office, and industrial), commute mode, and street intersection density. Scores can range from 1 to 20, with a higher score indicating greater likelihood of walking trips. The numbers of gyms within a 3-km radius of each residential address were calculated from 2014 data using ESRI Business Data / ArcGIS Business Analyst (North American Industry Classification System code 713940) using ArcGIS software 10. The numbers of recreation areas within a 3-km radius of each residential address were calculated from 2014 data using the Recreation Areas layer of TeleAtlas in ArcGIS. All of the following were considered recreation areas: golf courses, amusement parks, beaches, and park and recreation areas.
Analytic covariates of baseline maternal age, self-reported race/ethnicity, insurance status, and the presences of chronic hypertension and pre-gestational diabetes were assessed during the nuMoM2b early pregnancy study visit 14. Briefly, for this analysis, age was reported using categories of 18–21 years, 22–35 years, and > 35 years for descriptive tables and included as a continuous covariate in regression models; self-reported race/ethnicity was coded as white non-Hispanic, black non-Hispanic, Hispanic, Asian, and other; and insurance status as private (commercial), public (government or military), and none (self-pay or other). The presence of chronic hypertension was ascertained by medical record review for the following criteria: diagnosis of hypertension prior to pregnancy OR measured systolic ≥ 140 mmHg or diastolic ≥ 90 mmHg on two occasions at least 6 hours apart, or on one occasion followed by antihypertensive medication, prior to 200 weeks of gestation. Pre-gestational diabetes was ascertained by medical record abstraction.
Statistical Analysis
Descriptive statistics.
The percent of participants adhering to physical activity guidelines was compared for categories of baseline characteristics during each of the three timepoints during pregnancy using chi-square tests. Neighborhood characteristics were summarized using medians and interquartile ranges within categories of participants’ baseline characteristics and compared across categories using Kruskal-Wallis tests. Neighborhood characteristics were also summarized and compared according to adherence (yes/no) using Wilcoxon rank-sum tests at each timepoint.
Models of association.
Relative risk of adherence to physical activity guidelines was estimated for units of change in neighborhood characteristics of the participant. Models were fit separately for each of the three timepoints during pregnancy when physical activity and residential address were collected and for each of the three neighborhood characteristics: walkability index, number of gyms, and number of recreation areas. The models included fixed effects for maternal age (continuous), race/ethnicity, insurance status, chronic hypertension, and pre-gestational diabetes as adjustment covariates. A random intercept for recruitment subsite was included to account for association between the measures of the built environment and physical activity that might be driven by the general geographic location. The number of gyms and the number of recreation areas were log2 transformed in the models due to extreme skewing; accordingly, relative risks for these exposures are interpreted on a multiplicative scale. Poisson regression was used to compute the estimates using a spatial power covariance structure 19, which scales the covariance between observations by a function of their distance to account for clustering of the observations, and robust (sandwich) variance estimation for fixed effects 20. Poisson regression for the estimation of relative risks, rather than logistic regression for odds ratios, was used due to the relatively high prevalence of the outcome; under this condition, the odds ratio is a poor estimator of relative risk 21.
Statistical analyses were performed using SAS V9.4.
RESULTS
A total of 8,362 women were included in the analysis at one or more timepoints. METS ranged from 0 to 33.8 in the first trimester, 0 to 32.8 in the second trimester, and 0 to 28.3 in the third trimester. Adherence to physical activity guidelines was lower in the third trimester than it was in the first trimester. The prevalence of adherence in the full sample was 34% (2785/8241) in the first trimester, 33% (2643/7975) in the second trimester, and 30% (2343/7912) in the third trimester. Among the 7,661 participants who contributed data at all three timepoints, 34% (2636/7661) were guideline adherent during the first trimester, 33% (2548/7661) during the second trimester, and 30% (2279/7661) during the third trimester. Table 1 shows the prevalence of physical activity level guideline adherence by participant characteristics. Most patterns were consistent across trimesters. Guideline adherence was consistently highest among women over 35 years old (ranging from 39.5% to 43.2%). Non-Hispanic black and Hispanic women and women with public insurance were less likely to adhere to physical activity guidelines than those in other race/ethnic and insurance groups. Women with chronic hypertension and pre-gestational diabetes were also less likely to adhere to physical activity guidelines.
Table 1.
Prevalence of physical activity level guideline adherence within categories of participant characteristics at three timepoints during pregnancy in nuMoM2ba
| Baseline Participant Characteristic | Visit 1: 6 to <14 Weeks | Visit 2: 16 to <22 weeks | Visit 3: 22 to <30 weeks |
|---|---|---|---|
| N=8241 | N=7975 | N=7912 | |
|
| |||
| Statistics: | n/N (%) | n/N (%) | n/N (%) |
|
| |||
| Maternal age, years | |||
| 18–21 | 326/1543 (21.1) * | 258/1454 (17.7) * | 244/1440 (16.9) * |
| 22–35 | 2221/6146 (36.1) * | 2152/5982 (36.0) * | 1892/5948 (31.8) * |
| > 35 | 238/552 (43.1) * | 233/539 (43.2) * | 207/524 (39.5) * |
| Maternal race/ethnicity | |||
| White Non-Hispanic | 2024/5084 (39.8) * | 1962/4995 (39.3) * | 1743/4974 (35.0) * |
| Black Non-Hispanic | 218/1057 (20.6) * | 176/994 (17.7) * | 176/989 (17.8) * |
| Hispanic | 282/1369 (20.6) * | 270/1272 (21.2) * | 222/1245 (17.8) * |
| Asian | 119/329 (36.2) * | 113/322 (35.1) * | 87/316 (27.5) * |
| Other | 142/402 (35.3) * | 122/392 (31.1) * | 115/388 (29.6) * |
| Insurance status | |||
| Private | 2248/5778 (38.9) * | 2190/5661 (38.7) * | 1917/5636 (34.0) * |
| Public | 440/2144 (20.5) * | 375/2006 (18.7) * | 353/1969 (17.9) * |
| None | 97/319 (30.4) * | 78/308 (25.3) * | 73/307 (23.8) * |
| Chronic hypertension: | |||
| Yes | 46/204 (22.5) * | 37/200 (18.5) * | 36/199 (18.1) * |
| No | 2739/8037 (34.1) * | 2606/7775 (33.5) * | 2307/7713 (29.9) * |
| Pre-gestational diabetes: | |||
| Yes | 31/126 (24.6) * | 21/122 (17.2) * | 26/115 (22.6) |
| No | 2754/8115 (33.9) * | 2622/7853 (33.4) * | 2317/7797 (29.7) |
Statistical significance reported as:
= p < 0.05. P-values obtained from chi-square tests comparing the prevalence of adherence to physical activity guidelines among different categories of participant characteristics by study visit.
There were significant differences in all neighborhood characteristics across baseline participant characteristics (Table 2). Women > 35 years of age had the highest median walkability score (14.17 [IQR 11.83, 16.17] vs. 13.67 [11.33, 15.33] for age 18–21 years) and numbers of gyms (19 [8, 59] vs. 8 [4, 14]) and recreation areas (15 [6, 34] vs. 8 [4, 14]) within a 3 km radius of participants’ homes. These characteristics also varied by race/ethnicity. Hispanic and Asian women had approximately double the numbers of gyms and recreation areas available to them, compared to white non-Hispanic and Black non-Hispanic women. The median numbers of gyms and recreation areas were lowest among those women without insurance. Availability of gyms and recreation areas were lowest for women with chronic hypertension and pre-gestational diabetes compared with those who did not have those conditions, and walkability scores were also significantly lower among women with chronic hypertension.
Table 2.
Baseline neighborhood characteristics by baseline study characteristics in nuMoM2b (n=8,241)a
| Baseline Participant Characteristics | Walkability Indexb | Number of Gymsb | Number of Recreation Areasb |
|---|---|---|---|
| Median (IQR) | Median (IQR) | Median (IQR) | |
|
| |||
| Maternal Age Category: | |||
| 18–21 | 13.7 (11.3, 15.3) * | 8 (4, 14) * | 8 (4, 14) * |
| 22–35 | 13.8 (11.2, 15.8) * | 11 (5, 30) * | 9 (4, 23) * |
| >35 | 14.2 (11.8, 16.2) * | 19 (8, 59) * | 15 (6, 34) * |
| Maternal race: | |||
| White Non-Hispanic | 13.8 (10.7, 15.8) * | 9 (5, 22) * | 8 (3, 18) * |
| Black Non-Hispanic | 13.8 (12.0, 15.5) * | 9 (5, 15) * | 9 (5, 16) * |
| Hispanic | 13.7 (11.7, 15.7) * | 20 (8, 39) * | 16 (8, 33) * |
| Asian | 14.3 (12.2, 16.3) * | 19 (9, 68) * | 15 (8, 32) * |
| Other | 13.8 (11.2, 16.0) * | 10 (5, 20) * | 9 (4, 17) * |
| Insurance status: | |||
| Private | 14.0 (11.0, 16.0) * | 11 (5, 31) * | 9 (4, 24) * |
| Public | 13.7 (11.5, 15.3) * | 10 (5, 22) * | 10 (5, 19) * |
| None | 13.8 (11.0, 15.8) * | 8 (5, 16) * | 7 (4, 13) * |
| Chronic hypertension: | |||
| Yes | 12.9 (10.3, 15.7) * | 9 (4, 20) * | 8 (3, 16) * |
| No | 13.8 (11.3, 15.8) * | 11 (5, 27) * | 9 (4, 22) * |
| Pre-gestational diabetes: | |||
| Yes | 13.3 (11.0, 15.7) | 8 (5, 16) * | 8 (3, 16) * |
| No | 13.8 (11.2, 15.8) | 11 (5, 27) * | 9 (4, 22) * |
Statistical significance reported as:
= p < 0.05. P-values obtained from Kruskal Wallis tests comparing the distribution of neighborhood characteristics by different categories of participant characteristics at baseline.
Baseline minimum and maximum values for each exposure: walkability (1.5, 20.0), gyms (0, 373), and recreation areas (0, 62).
Median walkability index, number of gyms, and number of recreation areas were all significantly higher for participants who met physical activity guidelines compared with those who did not (Table 3). Median scores were consistent across trimesters. In sensitivity analyses (Supplemental Table 1) restricted to women who contributed data at all three timepoints, the results were comparable.
Table 3.
Neighborhood characteristics according to physical activity guideline adherence at three timepoints during pregnancy in nuMoM2ba
| Neighborhood characteristic, median (IQR) | Visit 1: 6 to < 14 Weeks |
Visit 2: 16 to < 22 weeks |
Visit 3: 22 to < 30 weeks |
|||
|---|---|---|---|---|---|---|
| (N=8241) | (N=7975) | (N=7912) | ||||
| Achieved guidelines |
Did not achieve guidelines |
Achieved guidelines |
Did not achieve guidelines |
Achieved guidelines |
Did not achieve guidelines |
|
| (n=2785) | (n=5456) | (n=2643) | (n=5332) | (n=2343) | (n=5569) | |
|
| ||||||
| Walkability score | 14.2 (11.8–16.2)* | 13.7 (11.0–15.7) | 14.3 (11.8–16.3)* | 13.7 (11.0–15.5) | 14.3 (11.8–16.3)* | 13.7 (10.8–15.5) |
| Gyms count | 14 (6–41)* | 10 (5–21) | 14 (6–49)* | 9 (5–19) | 15 (6–51)* | 9 (5–19) |
| Recreation areas count | 11 (5–30)* | 9 (4–18) | 12 (5–31)* | 8 (4–17) | 12 (5–32)* | 8 (3–17) |
Statistical significance reported as:
= p < 0.05 for difference in the neighborhood characteristic among those who achieved the physical activity guidelines versus those who did not. P-values obtained from Wilcoxon rank-sum tests
Participants living in neighborhoods that were more walkable were slightly more likely to meet physical activity guidelines than those who lived in less walkable neighborhood, in models adjusted for maternal age, maternal race/ethnicity, insurance status, chronic hypertension, and pre-gestational diabetes mellitus (Table 4). This association was strongest and statistically significant in the second and third trimester, where for each 3-unit increase in walkability index score, women were 4% (95% confidence interval (CI): 1.00, 1.08) more likely to meet the guidelines. All other associations between the built environment and physical activity were strongest in the third trimester as well. With each doubling in numbers of gyms and recreation areas available within a 3 km radius, women in their third trimester were 10% (95% CI: 1.07, 1.13) and 8% (95% CI: 1.03, 1.12) more likely to meet physical activity guidelines. Results were comparable in sensitivity analyses (Supplemental Table 2) restricted to women who contributed data at all three timepoints.
Table 4.
Relative risk of meeting physical activity guidelines per units of change in neighborhood characteristics at three timepoints during pregnancy in nuMoM2ba
| Neighborhood Characteristic |
Visit 1: 6 to <14 weeks | Visit 2: 16 to <22 weeks | Visit 3: 22 to <30 weeks |
|---|---|---|---|
|
| |||
| (N=8241) | (N=7975) | (N=7912) | |
|
| |||
| RR (95% CI) | RR (95% CI) | RR (95% CI) | |
|
| |||
| Walkability index score, per 3 units | 1.03 (1.00, 1.07) | 1.04 (1.00, 1.08)c | 1.04 (1.00, 1.08)c |
| Number of gyms, per doubling in valueb | 1.07 (1.04, 1.10) | 1.08 (1.05, 1.11) | 1.10 (1.07, 1.13) |
| Number of recreation areas, per doubling in valueb | 1.04 (1.00, 1.08) | 1.05 (1.01, 1.09) | 1.08 (1.03, 1.12) |
Models are adjusted for maternal age, maternal race/ethnicity, insurance status, chronic hypertension, and pre-gestational diabetes mellitus.
This characteristic (x) is included in statistical models after a log2 transformation (log2(x+1)). Relative risks for log-transformed quantities are interpreted on the multiplicative scale.
The non-rounded CI excludes 1.0.
DISCUSSION
In this study, we found that several features of the built environment were associated with meeting physical activity guidelines during pregnancy. Specifically, women living in more walkable neighborhoods and neighborhoods with more gyms and recreation areas were more likely to meet physical activity guidelines. This association held across all trimesters, and point estimates were largest in the third trimester.
Our findings for walkability, gyms, and recreation areas are consistent with previous studies of the built environment and physical activity during pregnancy. A study in Oslo, Norway found that women in neighborhoods with better objective access to recreational areas engaged in nine more minutes of moderate-vigorous physical activity per day than those in neighborhoods with limited access 12. In addition, a study of five counties in North Carolina found women living in the most walkable neighborhoods (highest tertile) were over two times more likely to achieve weekly levels of physical activity above the median than those in the least walkable neighborhoods 13. They also found women living furthest from a physical activity facility and those living furthest from a park were less likely to achieve above-median levels of physical activity. However, findings were only statistically significant for the association with physical activity facility.
Previous studies of the built environment and physical activity during pregnancy have not examined associations across all three trimesters. The Oslo study did not find significant differences in associations of the built environment with physical activity between early and mid-pregnancy 12, which is consistent with our findings, but they did not examine these relationships in the third trimester. We not only extended our study into the third trimester, but also found the point estimates of the association with physical-activity-guideline adherence to be greatest during that trimester. The strength of the relationship is noteworthy given that the prevalence of physical activity typically declines in that final trimester 5,22.
The most likely pathway linking the built environment to physical activity during pregnancy is that having these physical activity-promoting resources makes it easier for women to engage in more activities. Few studies have examined where people engage in physical activity, and none to our knowledge have assessed this during pregnancy, but a study of adults living in 5 states found participants’ homes, roads, and parks were the most common locations for engaging in moderate to vigorous physical activity 23. However, there is also research demonstrating that commonly used neighborhood buffers do not accurately capture the spaces where individuals engage in physical activity 24. Another plausible explanation for our significant findings is that our indicators of the built environment are proxies for social factors like economic investment, social cohesion, or political capital 25. These factors can influence physical activity in ways that we did not measure such as through differential exposure to psychosocial stressors (e.g., violence) or unsafe conditions (e.g., poor maintenance of sidewalks or absence of streetlights).
This study is not without limitations. Our cross-sectional study design means that we cannot rule out the possibility of residential self-selection bias due to health-related attitudes or neighborhood preferences that influence both neighborhood choice and health-related outcomes. This could occur in our study if an individual’s desire to be physically active led them to choose to live in a neighborhood with more physical activity-promoting resources. While we cannot rule out this possibility, several previous studies in non-pregnant adults suggest residential self-selection does not solely account for observed associations between the neighborhood environment and physical activity 26–30. Another limitation is that, as alluded to above, we cannot be sure that the geographic boundaries we chose best reflect the spaces where participants engage in physical activity 23,24,31. This potential error in the measurement of the physical activity environment is likely non-differential with respect to physical activity, so this is expected to bias our finding towards the null.
Our measures of gyms and recreation areas are limited in that we were not able to capture quality or the presence of specific features that might promote physical activity. Further work is needed to determine whether these features may be more salient for physical activity promotion than availability. In addition, obtaining information of these features of the built environment from commercial databases allowed us to generate these measures for the diverse places where numom2b participants lived, but it also precluded us from performing detailed quality checks to ensure completeness. Finally, while steps were taken to increase the precision and accuracy of self-reported physical activity including collection of information on type, frequency, and duration 32, it still may not be as accurate as objective measures.
CONCLUSIONS
Our results suggest the built environment has similar physical activity-promoting benefits in pregnant women as it does in general populations. Given the low prevalence of physical activity guideline adherence during pregnancy, more work is needed to understand the extent to which features of the built environment can be leveraged to improve physical activity at this important phase in the life course. Findings from our large, multi-site study suggest recreation areas, walkable areas, and gyms may facilitate physical activity during pregnancy. Further work is needed to determine how best to leverage these resources to promote adherence to physical activity guidelines.
Supplementary Material
ACKNOWLEDGEMENTS
The authors would like to express their gratitude to Dr. Corette B. Parker (RTI International) for her adept scientific and statistical leadership, and Ms. Maggie O’Neal (RTI International) for obtaining geospatial data.
FUNDING SOURCE
This work was supported by grants (cooperative agreements) from the National Heart, Lung, and Blood Institute and the Eunice Kennedy Shriver National Institute of Child Health and Human Development: U01HL145358; U10-HL119991; U10-HL119989; U10-HL120034; U10-HL119990; U10-HL120006; U10-HL119992; U10-HL120019; U10-HL119993; and U10-HL120018. Support was also provided by the National Institutes of Health: Office of Disease Prevention; Office of Research on Women’s Health; Office of Behavioral and Social Sciences Research; and the National Center for Advancing Translational Sciences – UL-1-TR000124, UL-1-TR000153, UL-1-TR000439, and UL-1-TR001108. In addition, support was provided by respective Clinical and Translational Science Institutes to Indiana University (UL1TR001108) and University of California Irvine (UL1TR000153). The content of this article is solely the responsibility of the authors and does not necessarily represent the official views of the National Heart, Lung, and Blood Institute, the National Institutes of Health, or the U.S. Department of Health and Human Services.
REFERENCES
- 1.Physical Activity and Exercise During Pregnancy and the Postpartum Period: ACOG Committee Opinion Summary, Number 804. Obstetrics & Gynecology. 2020;135(4):991–993. [DOI] [PubMed] [Google Scholar]
- 2.Dipietro L, Evenson KR, Bloodgood B, et al. Benefits of Physical Activity during Pregnancy and Postpartum: An Umbrella Review. Medicine and science in sports and exercise. 2019;51(6):1292–1302. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Downs DS, Chasan-Taber L, Evenson KR, Leiferman J, Yeo S. Physical activity and pregnancy: past and present evidence and future recommendations. Res Q Exerc Sport. 2012;83(4):485–502. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Gaston A, Cramp A. Exercise during pregnancy: a review of patterns and determinants. Journal of science and medicine in sport. 2011;14(4):299–305. [DOI] [PubMed] [Google Scholar]
- 5.Catov JM, Parker CB, Gibbs BB, et al. Patterns of leisure-time physical activity across pregnancy and adverse pregnancy outcomes. Int J Behav Nutr Phys Act. 2018;15(1):68. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Perak AM, Ning H, Khan SS, Van Horn LV, Grobman WA, Lloyd-Jones DM. Cardiovascular Health Among Pregnant Women, Aged 20 to 44 Years, in the United States. J Am Heart Assoc. 2020;9(4):e015123–e015123. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.U.S. Department of Health and Human Services. Physical activity guidelines for Americans. DHHS. https://health.gov/paguidelines/second-edition/. Published 2018. Accessed March 22, 2020. [Google Scholar]
- 8.Hesketh KR, Evenson KR. Prevalence of U.S. Pregnant Women Meeting 2015 ACOG Physical Activity Guidelines. Am J Prev Med. 2016;51(3):e87–e89. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Smith M, Hosking J, Woodward A, et al. Systematic literature review of built environment effects on physical activity and active transport - an update and new findings on health equity. Int J Behav Nutr Phys Act. 2017;14(1):158–158. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Eriksson U, Arvidsson D, Sundquist K. Availability of exercise facilities and physical activity in 2,037 adults: cross-sectional results from the Swedish neighborhood and physical activity (SNAP) study. BMC Public Health. 2012;12(1):607. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Roux AVD, Evenson KR, McGinn AP, et al. Availability of Recreational Resources and Physical Activity in Adults. American Journal of Public Health. 2007;97(3):493–499. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Richardsen KR, Mdala I, Berntsen S, et al. Objectively recorded physical activity in pregnancy and postpartum in a multi-ethnic cohort: association with access to recreational areas in the neighbourhood. Int J Behav Nutr Phys Act. 2016;13:78. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Porter AK, Rodríguez DA, Frizzelle BG, Evenson KR. The Association between Neighborhood Environments and Physical Activity from Pregnancy to Postpartum: a Prospective Cohort Study. Journal of urban health : bulletin of the New York Academy of Medicine. 2019;96(5):703–719. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Haas DM, Parker CB, Wing DA, et al. A description of the methods of the Nulliparous Pregnancy Outcomes Study: monitoring mothers-to-be (nuMoM2b). Am J Obstet Gynecol. 2015;212(4):539.e531–539.e524. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Hibbard JH, Mahoney ER, Stockard J, Tusler M. Development and testing of a short form of the patient activation measure. Health Serv Res. 2005;40(6 Pt 1):1918–1930. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Ainsworth BE, Haskell WL, Herrmann SD, et al. 2011 Compendium of Physical Activities: a second update of codes and MET values. Med Sci Sports Exerc. 2011;43(8):1575–1581. [DOI] [PubMed] [Google Scholar]
- 17.ArcGIS Desktop: Release 10. [computer program]. Redlands, CA: Environmental Systems Research Institute; 2011. [Google Scholar]
- 18.EPA’s Smart Location Database: A National Dataset for Characterizing Location Sustainability and Urban Form. Draft Technical Report; 2012. https://edg.epa.gov/data/PUBLIC/OP/Smart_Location_DB_v02b.zip. [Google Scholar]
- 19.Stroup WW, Milliken GA, Claassen EA, Wolfinger RD. SAS® for Mixed Models: Introduction and Basic Applications. Cary, NC: SAS Institute Inc.; 2018. [Google Scholar]
- 20.Zou G A modified poisson regression approach to prospective studies with binary data. Am J Epidemiol. 2004;159(7):702–706. [DOI] [PubMed] [Google Scholar]
- 21.Rothman KJ, Greenland S, Lash TL. Modern Epidemiology. Philadelphia, PA: Lippincott Williams & Wilkins; 2008. [Google Scholar]
- 22.Borodulin KM, Evenson KR, Wen F, Herring AH, Benson AM. Physical activity patterns during pregnancy. Medicine and science in sports and exercise. 2008;40(11):1901–1908. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Holliday KM, Howard AG, Emch M, Rodríguez DA, Rosamond WD, Evenson KR. Where Are Adults Active? An Examination of Physical Activity Locations Using GPS in Five US Cities. Journal of urban health : bulletin of the New York Academy of Medicine. 2017;94(4):459–469. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Holliday KM, Howard AG, Emch M, Rodríguez DA, Evenson KR. Are buffers around home representative of physical activity spaces among adults? Health & place. 2017;45:181–188. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.McNeill LH, Kreuter MW, Subramanian SV. Social environment and physical activity: a review of concepts and evidence. Social science & medicine (1982). 2006;63(4):1011–1022. [DOI] [PubMed] [Google Scholar]
- 26.Baar J, Romppel M, Igel U, Brähler E, Grande G. The independent relations of both residential self-selection and the environment to physical activity. International journal of environmental health research. 2015;25(3):288–298. [DOI] [PubMed] [Google Scholar]
- 27.James P, Hart JE, Arcaya MC, Feskanich D, Laden F, Subramanian SV. Neighborhood Self-Selection: The Role of Pre-Move Health Factors on the Built and Socioeconomic Environment. International journal of environmental research and public health. 2015;12(10):12489–12504. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Kaczynski AT, Mowen AJ. Does self-selection influence the relationship between park availability and physical activity? Preventive medicine. 2011;52(1):23–25. [DOI] [PubMed] [Google Scholar]
- 29.Mackenbach JD, Matias de Pinho MG, Faber E, et al. Exploring the cross-sectional association between outdoor recreational facilities and leisure-time physical activity: the role of usage and residential self-selection. Int J Behav Nutr Phys Act. 2018;15(1):55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Van Dyck D, Cardon G, Deforche B, Owen N, De Bourdeaudhuij I. Relationships between neighborhood walkability and adults’ physical activity: How important is residential self-selection? Health & place. 2011;17(4):1011–1014. [DOI] [PubMed] [Google Scholar]
- 31.Perchoux C, Chaix B, Cummins S, Kestens Y. Conceptualization and measurement of environmental exposure in epidemiology: Accounting for activity space related to daily mobility. Health & place. 2013;21:86–93. [DOI] [PubMed] [Google Scholar]
- 32.Evenson KR, Chasan-Taber L, Symons Downs D, Pearce EE. Review of self-reported physical activity assessments for pregnancy: summary of the evidence for validity and reliability. Paediatric and perinatal epidemiology. 2012;26(5):479–494. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.