Table 1 Main characteristics of the selected studies, ordered by year of publication
From: Physical activity and gestational weight gain: a systematic review of observational studies
Authors, Years | Study design, period location | Population size | Outcomes | PA dimensions assessed | Statistical methods | Confounders / stratification | Main findings |
|---|---|---|---|---|---|---|---|
Olson et al. 2003 [19] | Cohort study, no information on the period New York (US) | 622 pregnant women | Excessive GWG (according 2009 IOM guidelines) Inadequate GWS (according 2009 IOM guidelines) | Self-reported PA: Level PA | Multiple linear and logistic regression model | Maternal characteristics-BMI, the trimester that the prenatal questionnaire was completed, the weeks of gestation, the weeks from the first to the last weight measurement, the weeks from the last measurement to delivery. | Physical activity was significantly related to excessive but not inadequate GWG. |
Haakstad et al. 2007 [20] | Cross-sectional survey no information on the period Oslo (Norway) | 467 pregnant women | Excessive GWG (According 2009 IOM guidelines) Inadequate GWG (< 16 kg) Overweight (BMI > 25) | Self-reported PA PA level, Sedentary activities, PA duration | The x2-test | Maternal characteristic- none | Women who exercised regularly had lower weight gain than inactive women. |
Stuebe et al. 2009 [21] | Cohort study no information on the period Massachusetts (US) | 1388 pregnant women | Total GWG Excessive GWG (According 2009 IOM guidelines) | Self-reported PA -PA duration -PA level -Type of PA - Sedentary | Multivariable logistic and linear regression | Maternal characteristic- Pre-pregnancy BMI, age, race/ethnicity, smoking status, gestational age at delivery, and nausea in the first trimester of pregnancy; | Vigorous activity, walking, and total activity during pregnancy were inversely associated with excessive GWG Walking and vigorous activity were also inversely associated with total GWG. |
Cohen et al. 2009 [22] | Ad hoc recruitment of pregnant women From August 2008 to December 2008 Ottawa and Montreal (Canada) | 81 pregnant women | Achieving recommanded GWG (According 2009 IOM guidelines) | Self-reported PA -PA duration | Univariate logistic regressions | Maternal characteristics: none | The chance for pregnant women to achieved their recommended GWG increase significantly for those who accumulated > 8.5 MET-hr/wk compared to those accumulated < 8.5 MET-hr/wk. |
Melzer et al. 2010 [23] | Observational study no information on the period Geneva, (Switzerland) | 44 pregnant women | Total GWG | Objective PA measure - PA level | t-test | Maternal characteristic: none | There is no difference between Active and inactive women in term of body weight gain |
Abeysena et al. 2011 [24] | Cohort study May 2001 – April 2002 Sri Lanka | 580 pregnant women. | Inadequate GWG (< 2009 IOM guidelines) | Self-reported PA - Type of PA | Multivariate logistic regression | Maternal characteristics- Sleeping during 2nd, 3rd or both trimesters, multiparity, sex of newborn, per capita monthly income, Period of gestation, Period of gestation at recruitment, BMI, gestational age, BMI*Sleeping | Standing and walking more than 5 h per day during the second trimester increase the risk of inadequate weight gain during pregnancy. |
Hong Jiang et al. 2012 [25] | Cohort study 2005 to 2007 Changzhou, Jiangsu Province, (China) | 862 pregnant women | Total GWG Excessive GWG (According 2009 IOM guidelines) | Objective measure of PA - PA level | Multiple linear and logistic regression | Maternal characteristics- Age, educational level, job type, the families’ income, pre-pregnancy BMI, passive tobacco exposure and food energy intake others: gestational age, newborns sex | The GWG decrease among active women compared to the sedentary women during the 2nd and the 3rd trimesters. The risk of excessive GWG decrease significantly among the active women compared to the sedentary women during the 2nd and the 3rd trimester. |
Monpetit et al. 2012 [26] | Prospective study From August to December 2008 Ottawa and Montreal, (Canada) | 59 pregnant women | Total GWG | Self-reported PA: - PA level Objective PA measure -Daily steps | Hierarchical multiple regression analyses Pearson correlation coefficients | Maternal characteristic- Energy intake Pre-pregnancy BMI | The step is no significant predictor of GWG. no significant correlation between GWG and steps. |
Cohen et al. 2013 [27] | Prospective study no information on the period Ottawa and Montreal (Canada) | 61 pregnant women | Total GWG Rate of weight gain (kg/week) | Self-reported PA - PA duration - PA intensity - Type of PA - Sedentary Objective PA measure Daily steps | Pearson correlation coefficients, PCA | Maternal characteristics: none | Results suggest that walking and pedometer steps were associated with the rate of GWG |
Kraschnewski et al. 2013 [28] | Cohort study From January 2009 to April 2011 Pennsylvania (US) | 2603 Pregnant women | Excessive GWG (According 2009 IOM guidelines) | Self-reported PA - PA duration | Multivariable logistic regression | Maternal characteristic- Prepregnancy weight category, age, Race/Ethnicity, Education, Poverty Status, Marital Status, Gestational age at delivery, Smokes Daily | Results show that meeting the physical activity guidelines during pregnancy was significantly associated with a decrease risk of exceeding GWG recommendations. |
Restall et al. 2014 [29] | Cohort study From November 2004 and February 2011 Australia, New Zealand, Ireland, | 1950 pregnant women | Excessive GWG (According 2009 IOM guidelines) | Self-reported PA Descline exercise during pregnancy | Multivariate logistic regression | Maternal characteristics- Age, BMI, smoke, Mother’s birth weight, Immigrant in past 5 years, fertility treatment, fish or seafood intake, limiting behavior score, sleep Others: Centre | There is a significant increase risk of GWG among women who decreased their level of exercise during pregnancy compared to those who unchanged. |
Ruifrok et al., 2014 [30] | Randomized controlled trials analysed as a cohort From 2005 2006 Amsterdam (Netherlands) | 111 pregnant women | Rate of Weight gain (kg/week) | Objective PA measure - PA level - Sedentary | Multivariate regression models | Maternal characteristic- BMI, parity, gestational age Others: intervention group | There is no significant association between MVPA or sedentary behavior at 15 weeks with GWG. No significant associations were found for changes in PA and sedentary behavior from 15 to 32–35 weeks of gestation. |
Chasan et al. 2014 [31] | Cohort study From 2006 to 2011 Western Massachusetts (US) | 1297 pregnant women | -Total GWG - Rate of Weight gain (kg/week) -Inadequate GWG - Excessive GWG (According 2009 IOM guidelines) | Self-reported PA -PA duration -PA intensity - Type of PA - Sedentary - Met PA guidelines | Multinomial logistic regression Linear regression models | Maternal characteristics-pre-pregnancy BMI, age, parity, smoking | There is no significant association between inadequate and excessive GWG and late pregnancy physical activity. However, the total and rate of GWG increase significantly with total physical activity and with physical activity guideline. |
Schlaff et al. 2014 [32] | Cohort study From 2008 to2012 Michigan (US) | 135 pregnant women | Excessive GWG (According 2009 IOM guidelines) | Self-reported PA - LTPA level | Multivariate logistic regression model | Maternal characteristics: WIC | Results suggest that LTPA level was not significantly related to appropriateness of GWG. |
Schlaff et al. 2014 [33] | Cohort study from September 1998 to June 2004 Michigan (US) | 449 pregnant women | Inadequate GWG Excessive GWG (According 2009 IOM guidelines) | Self-reported PA -LTPA intensity | Polytomous logistic regression | Maternal characteristics: parity, BMI | Results suggest that LTPA and GWG are not significantly associated. |
Merkx et al. 2015 [34] | Cross-sectional survey From September to November 2012 Netherlands | 396 pregnant women | Inadequate GWG - Excessive GWG (According 2009 IOM guidelines) | Self-reported PA - motivation healthy PA, - Decline in PA | Multinomial logistic regression | Maternal characteristics- Vegetable consumption, age, gestational age, parity, family income education level, smoking behavior,s atisfied pre-pregnancyweight, perceived BMI | A decline in PA was associated with Excessive GWG. |
Ebrahimi et al. 2015 [35] | Cross-sectional study no information on the period Rafsanjan city (Iran) | 308 pregnant women | Total GWG Inadequate GWG Excessive GWG (According 2009 IOM guidelines) | Self-reported PA - PA duration - Sedentary | Multivariate Logistic regression models and cumulative logit model | Maternal characteristics- age, education level, and household income, dietary intake, BMI, number of pregnancy. | There is no significant association between PA duration and GWG. Sitting time was positively. associated with gestational weight gain, but the association did not persist in the cumulative logit analysis. |
Yong et al. 2016 [36] | Cross-sectional study From November 2010 and April 2012 Selangor and Negeri Sembilan (Malaysia) | 589 pregnant women | - Inadequate rate of GWG - Excessive rate of GWG (According 2009 IOM guidelines) | Self-reported PA - PA level | Multinomial logistic regression | Maternal characteristics- age, ethnicity, parity, | Women with low PA level were more likely to have excessive GWG, but the result were no significant. |
Collings et al. 2020 [37] | Cohort study From Mars 2007 to December 2010 England | 2702 pregnant women A | Total GWG | Self-reported PA - PA level | Multivariate Linear regression | Maternal characteristics- age, gestational age at measurement, socioeconomic status, parity, smoking, alcohol consumption, cafeine intake, sleep quality, use of dietary supplements, early-pregnancy BMI, and the number of weeks between mid- and late- pregnancy weight measurements. Stratified: for white British and Pakistani-origin women, separately | No association was found between PA level and GWG. |
Anh Vo Van Ha et al. 2020 [38] | Cohort study From 2015 to 2017 Vietnam | 1873 pregnant women | Total GWG | Self-reported PA - PA duration - PA intensity - Sedentary | Multiple linear regression models | Maternal characteristic- age, education, gestational diabetes mellitus, history of health-related problems, total energy intake during pregnancy, parity, employment, gestational age, and pre-pregnancy BMI | Women with high PA level, intensity and household/caregiving activities, and occupational PA have significantly less GWG. Result suggest also women with longer sitting time have significant increase GWG. |
Sun et al., 2021 [39] | Cohort study From August 2016 to April 2017 Taiwan | 747 pregnant women | Excessive GWG (According 2009 IOM guidelines) | Self-reported PA - Decline in PA | Multivariate logistic regression model | Maternal characteristic- age, Pre-pregnancy BMI | A decline in PA was associated with Excessive GWG. |