The effect of an antenatal physical exercise programme on maternal/perinatal outcomes and quality of life in overweight and obese pregnant women: a randomised clinical trial


Dr FG Surita, CAISM/FCM/UNICAMP, R. Alexander Flemming, 101, 13083-881 Campinas, SP, Brazil. Email


Please cite this paper as: Nascimento S, Surita F, Parpinelli M, Siani S, Pinto e Silva J. The effect of an antenatal physical exercise programme on maternal/perinatal outcomes and quality of life in overweight and obese pregnant women: a randomised clinical trial. BJOG 2011;118:1455–1463.

Objective  To evaluate the effectiveness and safety of physical exercise in terms of maternal/perinatal outcomes and the perception of quality of life (QoL) in pregnant obese and overweight women.

Design  A randomised controlled clinical trial.

Setting  The Prenatal Outpatient Clinic of the Women’s Integral Healthcare Centre (CAISM-UNICAMP) at the University of Campinas, Campinas, Brazil.

Population  Eighty-two pregnant women (age ≥ 18 years; pre-gestational body mass index ≥ 26 kg/m2; gestational age 14–24 weeks).

Methods  Women were randomised into two groups: women in one group exercised under supervision and received home exercise counselling (the ‘study group’; n = 40) and women in the other group followed the routine prenatal care programme (the ‘control group’; n = 42).

Main outcome measures  Primary outcomes were gestational weight gain during the programme and excessive maternal weight gain. Secondary outcomes were increased arterial blood pressure, perinatal outcomes and QoL (WHOQOL-BREF).

Results  In the study group, 47% of pregnant women had weight gains above the recommended limit, compared with 57% of women in the control group (P = 0.43). There was no difference in gestational weight gain between the groups. Overweight pregnant women who exercised gained less weight during the entire pregnancy (10.0 ± 1.7 kg versus 16.4 ± 3.9 kg, respectively; = 0.001) and after entry into the study (5.9 ± 4.3 kg versus 11.9 ± 1.5 kg, respectively; = 0.021) compared with women in the control group. Arterial blood pressure was similar between the groups over time. There was no difference in perinatal outcome or QoL.

Conclusions  The exercise programme was not associated with control of gestational weight gain in our sample as a whole, but was beneficial for lower gestational weight gain in overweight women. Exercise was not associated with adverse perinatal outcomes and did not affect variation in arterial blood pressure or the perception of QoL.


Obesity is increasing in prevalence worldwide, and is now considered a global epidemic. It has become a significant threat to health in all sectors of the population, including women of reproductive age.1 According to the World Health Organization (WHO), the prevalence of obesity, defined as a body mass index (BMI) ≥ 30 kg/m2, during pregnancy ranges from 1.8 to 25.3%.1,2 In Brazil, a study conducted in six cities found that 5.5% of 5564 pregnant women evaluated were obese, and 25% were either overweight or obese.3

The Institute of Medicine (IOM)4 defines obesity during pregnancy as a pre-pregnancy BMI ≥ 30 kg/m2 and considers a range of gestational weight gains for each BMI category. For BMI ≥ 30 kg/m2 this range is 5–9 kg.

Obesity during pregnancy increases the risk of morbidity and mortality in both the mother and the fetus during pregnancy and adversely affects gestational outcome.5,6

During pregnancy and delivery, the maternal complications associated with obesity include gestational diabetes mellitus, gestational arterial hypertension and pre-eclampsia, venous thromboembolic disease, induction of labour and caesarean section. Clinical and surgical complications can also occur, such as infections, haemorrhage, anaemia, urinary tract infection and endometritis; in addition, stress urinary incontinence, depression and even difficulties with breastfeeding have been associated with obesity.2,6,7

An association has also been described between obesity during pregnancy and adverse neonatal outcomes, such as macrosomia, metabolic syndrome and a predisposition to obesity secondary to gestational diabetes in children,8,9 in addition to neural tube defects and congenital anomalies.10,11

While the negative impact of obesity on obstetric and perinatal outcomes is well established in the literature, information on how such adverse effects can be minimised through the use of specific interventions is still limited. Among diverse approaches, physical exercise has been indicated as an alternative for the management of obese pregnant women, although controversy remains regarding its effects during pregnancy.12

Physical exercise during pregnancy has been part of the recommendations of the American College of Obstetricians and Gynecologists (ACOG) since the mid-1990s. It is recognised as a safe practice, indicated for healthy pregnant women, as long as the intensity, duration and frequency of the exercise are tailored to the requirements of each woman.13 Light to moderate exercise is recommended for all women, even those with a sedentary lifestyle who wish to engage in some type of physical activity during pregnancy.14

Clinical trials have suggested that a change in lifestyle, as well as adherence to a suitable diet and exercise regimen, should be recommended for obese women to prevent excessive gestational weight gain, postpartum weight retention and adverse outcomes associated with obesity and excessive weight gain.12,15,16

In this study, we evaluated the effect of an exercise programme on gestational weight gain, maternal arterial blood pressure and perinatal outcome in overweight/obese pregnant women and their perception of quality of life (QoL).


A randomised controlled clinical trial was conducted in pregnant women seen at the Prenatal Outpatient Clinic of the Women’s Integral Healthcare Centre (CAISM-UNICAMP) from August 2008 to March 2010. Inclusion criteria were pregnancy, pre-gestational BMI categorised as overweight (26.0–29.9 kg/m2) or obese (≥30.0 kg/m2), age ≥ 18 years, and gestational age between 14 and 24 weeks. Exclusion criteria were multiple gestations, exercising regularly and conditions that contraindicate exercise, such as cervical incompetence, severe arterial hypertension, diabetes with vascular disease and risk of abortion.

Selected pregnant women were invited to participate in the study. The full study protocol was explained to these women and written informed consent was obtained. Subsequently, the women were randomly assigned to two groups: in the ‘study group’, the women exercised under supervision and received home exercise counselling, and in the ‘control group’ the women followed the routine prenatal programme provided by CAISM-UNICAMP. The results were analysed by treatment scheduled (intention-to-treat analysis).

The pregnant women were randomised to the groups using the sas statistical program (SAS Institute, Cary, NC, USA), which generated a list of random numbers based on a uniform distribution. To ensure blinding, the sequence was randomly distributed in opaque envelopes, which were sealed and sequentially numbered. Each participant received a sequence number corresponding to a sealed envelope.

After randomisation, sociodemographic and obstetric data for the pregnant women were obtained.

The exercise programme was designed to enable pregnant women to increase their level of physical activity and to improve their QoL through simple exercises that could be performed without supervision and that did not present a risk to the mother or the fetus.

The women were counselled on recommended weight gain for their BMI category; the importance and effects of physical activity during pregnancy; the optimal amount and intensity of home exercise; nutrition; suitable clothing to wear when exercising; the recommended duration of exercise; signs and symptoms to look out for during exercise; and when to interrupt physical activity. The exercise programme consisted of two components.

1 The exercise protocol. Exercise was performed by the women under the guidance of a trained physical therapist, in weekly classes. The protocol consisted of light-intensity to moderate-intensity exercises. According to the ACOG recommendations (2002),13 the woman’s heart rate did not exceed 140 beats per minute. Group or individual exercises consisted of 10 minutes of general stretching, 22 minutes of exercises to strengthen the lower and upper limb muscles, and 10 minutes of supervised relaxation, totalling 40 minutes. The exercises followed a standardised research protocol with a sequence of 22 exercises (see Supporting Information, Table S1).

2 Home exercise counselling. All pregnant women in the study group received counselling on home exercise to be performed five times a week. This exercise could consist of exercises from the protocol described above or walking. The women recorded the type (protocol or walking) and number of minutes of exercise in each session in a monthly exercise journal.

Pregnant women from the control group did not receive physical activity counselling and followed routine prenatal care advice. Both groups received standardised nutritional counselling from the Service of Nutrition and Dietetics (CAISM).

Follow-up data (gestational age, weight and arterial blood pressure) were recorded each time the woman attended the sessions until the end of pregnancy. In the study group, data were recorded on the days on which exercises were performed, at the beginning and end of sessions.

All pregnant women completed the WHOQOL-BREF questionnaire, on two occasions: at study inclusion and at the completion of 36 weeks of gestation. The domains of this questionnaire were scored on a scale of 0–100 points. Values closer to 0 were indicative of worse QoL and values closer to 100 reflected a better QoL.17

Data related to delivery (mode of delivery) and perinatal outcomes (weight of the newborn, Apgar index and adequacy of weight for gestational age) were collected using charts after the end of pregnancy.

The sample size was 41 women in each group. Sample size was calculated using a comparison between two proportions of excessive weight gain in obese women with gestational diabetes following a diet and exercise regimen (53.8 versus 78.9%)18 and the chi-square test; assuming a ratio between groups of 1:1, a significance level of 5% and power of test of 70%.

Demographic characteristics were described using frequencies, percentages, means and standard deviations. The groups were compared in terms of homogeneity using Student’s t test or the Mann–Whitney U test for continuous variables and the chi-square test for categorical variables. To determine the effectiveness of the intervention, the significance of differences between the groups in total weight gain (the difference between the pre-gestational weight reported by the pregnant woman and the weight measured at the last visit before the end of the pregnancy) and weight gain during the programme (the difference between the weight measured at study entry and the weight at the final visit, determined using a mechanical scale) was evaluated using Student’s t test or the Mann–Whitney U test.

Excessive gestational weight gain was defined as a weight gain of >11.5 kg for overweight pregnant women and >9.0 kg for obese pregnant women, according to IOM recommendations.4 Arterial blood pressure was measured using a mercury column sphygmomanometer and a stethoscope with the pregnant woman lying on her left side; the systolic and diastolic blood pressures were recorded. The adequacy of birthweight for gestational age was assessed according to the Alexander curve, and categorised as adequate, small or large (LGA) for gestational age, respectively, for measurements between the 10th and 90th percentiles.19 The effect of exercise in WHOQOL domains was evaluated by analysis of variance for repeated measures.

To evaluate the normal distribution in variables, the Kolmogorov–Smirnov test and histograms were used. The significance level was set at 5% and the software used in the analysis was sas version 9.1.

Pregnant women who abandoned the study over time were discontinued from the study. Some pregnant women delivered in other maternity hospitals and it was difficult to capture data for these women so these data are not included in the tables related to perinatal outcome.

This study was approved by the Institutional Review Board of the UNICAMP Medical School (FCM-UNICAMP) under registration number 542/2008.


From August 2008 to October 2009, 93 pregnant women were considered eligible for inclusion in the study. Eighty-two pregnant women were randomised to the groups, of whom two later discontinued their participation in the study. Therefore, 39 women in the study group and 41 in the control group completed the follow up (Figure 1).

Figure 1.

 Flow chart showing the inclusion and follow up of pregnant women in the study.

The groups did not differ significantly with respect to sociodemographic characteristics, weight, height, BMI, gestational age and the presence of co-morbidities at study inclusion (Table 1).

Table 1.   Sociodemographic characteristics, ponderal index, obstetric history and pathological conditions of overweight or obese pregnant women who did and did not participate in the exercise programme
VariableStudy groupControl groupP value
(n = 40)(n = 42)
  1. BMI, body mass index; GA, gestational age; PNC, prenatal care.

Age (years) (mean ± SD)29.7 ± 6.830.9 ± 5.90.479
School education (%)
<8 years15.023.80.603
8–11 years42.538.1
>12 years42.538.1
Occupation (%)
With remuneration40.042.50.820
Without remuneration60.057.5
Parity (%)
Diabetes (%)
Hypertension (%)37.552.40.176
Lower back pain (%)82.590.50.289
GA at initiation of PNC (weeks) (mean ± SD)14.3 ± 4.513.6 ± 3.50.431
GA at initiation of programme (mean ± SD)17.6 ± 4.217.8 ± 3.70.472
Height (m) (mean ± SD)1.63 ± 0.071.60 ± 0.060.116
Pre-gestational weight (kg) (mean ± SD)92.6 ± 18.994.0 ± 19.20.660
Pre-gestational BMI (kg) (mean ± SD)34.8 ± 6.636.4 ± 6.90.259

There was no significant difference between the groups in the number of visits, with a mean number of 8.0 visits in the study group and 7.1 in the control group. The follow-up period lasted on average 19 weeks in both groups.

There was no significant difference between the groups in terms of excessive weight gain. In the study group, 47.5% of the women gained more weight than recommended by the IOM, compared with 57.2% in the control group (= 0.43).

Regarding gestational weight gain, there was no significant difference between the groups. Total weight gain, weight gain after enrolment in the study and weekly weight gain were similar in the two groups. However, the women in the study group had a lower BMI average (38.6 ± 6.2 kg/m2) than those in the control group (41.4 ± 6.6 kg/m2) at the end of pregnancy (= 0.04). (Table 2)

Table 2.   Weight and BMI in overweight and obese pregnant women, according to whether they participated in the exercise programme
VariableStudy groupControl groupP value
(= 39) (mean ± SD)(= 41) (mean ± SD)
  1. *P value significant according to Student’s t test.

Final weight (kg)103.4 ± 18.9106.0 ± 19.60.621
Final BMI38.6 ± 6.241.4 ± 6.60.004
Total weight gain (kg)10.3 ± 5.011.5 ± 7.40.543
Weight gain in programme (kg)7.7 ± 4.38.1 ± 4.30.947
Weekly weight gain (kg)0.36 ± 0.220.38 ± 0.210.974
Obese(= 30)(= 36) 
Total weight gain (kg)10.4 ± 5.610.9 ± 7.60.757
Weight gain programme (kg)8.2 ± 4.37.7 ± 4.30.646
Weekly weight gain (kg)0.39 ± 0.220.36 ± 0.210.577
Overweight(= 9)(= 5) 
Total weight gain (kg)10.0 ± 1.716.4 ± 3.90.001*
Weight gain in programme (kg)5.9 ± 4.311.9 ± 1.50.021*
Weekly weight gain (kg)0.28 ± 0.220.57 ± 0.170.038*

The data were also analysed after stratification for pre-gestational BMI, with women being categorised as overweight (BMI 26.0–29.9 kg/m2) or obese (BMI ≥ 30.0 kg/m2). Among obese pregnant women, none of the variables showed a significant difference between the study and control groups. In contrast, overweight women from the study group benefited from exercise, gaining significantly less weight during their entire pregnancy when compared with women from the control group (10.0 ± 1.7 versus 16.4 ± 3.9 kg, respectively) and after enrolment in the study (5.9 ± 4.3 versus 11.9 ± 1.5 kg, respectively), with a mean weekly weight gain that was also smaller (0.28 ± 0.22 versus 0.57 ± 0.17 kg/week, respectively) (Table 2).

During pregnancy, no differences were found in the variables weight, BMI and arterial blood pressure between groups, as shown in Figure 2.

Figure 2.

 Changes in arterial blood pressure, weight and body mass index (BMI) over the course of pregnancy in women in the study and control groups. (A) Systolic blood pressure, (B) weight, (C) diastolic blood pressure and (D) BMI.

Concerning perinatal outcomes, no significant differences attributable to exercise were found between the groups. Both groups had high rates of caesarean section and LGA newborns (Table 3).

Table 3.   Perinatal outcomes in women who did and did not participate in the exercise programme
VariableStudy groupControl groupP value
Caesarean rate, n (%)25 (65.8)29 (72.5)0.521
Newborn weight (g) (mean ± SD)3267.4 ± 700.43228.4 ± 591.30.790
Gestational age at birth (mean ± SD)38.5 ± 2.638.5 ± 1.50.488
Apgar, 1 minute, n (%)
<76 (16.7)4 (10.8)0.515
>730 (83.3)33 (89.2)
Apgar, 5 minutes, n (%)
<71 (2.8)0 (0)0.493
>735 (97.2)37 (100)
Adequacy, n (%)
Adequate for gestational age23 (65.7)24 (58.5)1.000
Large for gestational age8 (24.2)8 (24.2)
Small for gestational age2 (6.1)1(3.0)

Adherence to home exercise counselling was 62.5% (n = 25), based on the percentage of pregnant women in the study group who completed an exercise journal. These women carried out a mean of 12.3 weeks of exercise, with means (±SD) of 57 ± 22.2 and 79.8 ± 48.6 minutes/week for exercises from the study protocol and walking, respectively. The majority (60%) of pregnant women did from 9 to 16 weeks of home exercise.

Quality of life was evaluated according to four domains (physical, psychological, social and environmental) in addition to two general questions about perception of QoL and satisfaction with health. Concerning the effect of time, the physical and social domains had significantly lower scores at the end of pregnancy, indicating worsening QoL in these aspects. There were no significant differences between the groups (Table 4).

Table 4.   Assessment of the effects of time and intervention on the perception of quality of life in pregnant women at the beginning and end of pregnancy using WHOQOL-Bref
DomainStudy groupControl groupTime effectIntervention effect
  1. *Repeated measures analysis of variance; initial, n = 79; final, n = 71.

Perception of quality of life
Satisfaction with health

During the performance of the exercise protocol under supervision, unexpected events, such as hypotension, falls and musculoskeletal lesions, were not observed.


The results of this study indicated that physical activity was not associated with control of gestational weight gain when the data were analysed as a whole. However, when the data were stratified according to BMI, with the women being divided into overweight and obese categories, it was found that the exercise programme was effective in terms of control of weight gain in the overweight women, despite the reduced sample size.

The statistical power of the overweight subgroup results for the variables total weight gain, weight gain during protocol exercises and weekly weight gain was 98, 81 and 62%, respectively, showing that only the results for weekly weight gain had low statistical power. However, the results for the overweight subgroup should be treated with caution because of the small number of women in the sample.

With respect to excessive weight gain based on IOM recommendations, there was a smaller proportion of women with excessive weight gain in the study group (47.5 versus 57.2%), but the difference was not significant, and so was insufficient to show that exercise had an effect on weight gain control.

Similar findings were obtained in studies using only educational programmes to encourage physical activity, which failed to prevent excessive gestational weight gain.16,20–23 In two randomised studies, low adherence rates to IOM recommendations were observed among obese pregnant women.16,23 In a study by Polley et al.,16 the opposite result from that expected was found in pregnant women with a BMI > 26 kg/m2: 59% of pregnant women in the intervention group exceeded the recommended weight gain limit, compared with 32% in the control group. In the study by Asbee et al.,23 weight gain was within the recommended limits in only 33.3% of women, similar to the results obtained in this study. The findings of these studies and those of the present study highlight the difficulty of achieving satisfactory results concerning weight control in obese pregnant women. Most studies have evaluated interventions that combined diet and exercise, in comparison with control groups without any intervention other than routine prenatal care.16,21–25 In contrast, our study was designed with the intention of isolating the exercise effect, ensuring that both groups received the same nutritional counselling which made it difficult to compare our results with those of other studies.

Comparison of our results with those of previous studies is difficult because of the different cut-off points used to evaluate excessive gestational weight gain. Until the latest revision of the IOM guidelines in 2009, an upper limit for weight gain in obese pregnant women had not been established. Mottola et al.25 considered as excessive weight gain >11.5 kg (the upper limit of the overweight range), these authors found 80% of pregnant women had adequate weight gain. In other studies in which weight gain control was not found to be successful, 6.8 kg was considered the upper limit of normal weight gain.16,22–25 Both of these limits differed from that used in this study, where we analysed excessive weight gain according to the new IOM range for obese women (5–9 kg). Artal et al.26 considered that the IOM recommendations for obese pregnant women are overestimated.

Overweight and obese pregnant women who gained <8 kg during pregnancy had lower rates of LGA newborns, pre-eclampsia, caesarean section and operative vaginal delivery than those who gained a significant amount of weight during pregnancy.27

The exercise programme was not associated with adverse outcomes in either the mother or the fetus, and did not affect systolic or diastolic arterial blood pressure or perinatal outcomes. Exercise performed by previously sedentary overweight and obese pregnant women seemed to be safe.

The route of delivery and the birthweight of the newborn were not influenced by exercising in obese women, which is similar to the findings of other studies examining exercise as a form of intervention during pregnancy.16,18,20,21,24,25 Gestational obesity and excessive maternal weight gain are associated with high rates of low birthweight and LGA newborns, and high rates of caesarean section.9,10,28,29 Kinnunen et al.,22 evaluating a programme to encourage physical activity, found a rate of LGA newborns of 15% in the control group. Nevertheless, gestational age at birth, the Apgar index and the proportion of LGA newborns were similar between the groups in our sample. We believe that the high rates of caesarean section and LGA newborns are related to population characteristics, as these are high-risk pregnant women with co-morbidities, such as diabetes.

The results of the present study suggest that exercise is safe in overweight and obese pregnant women as, contrary to concerns expressed in the long-running debate on the benefits of exercise in these women, exercise was not associated with low birthweight or preterm delivery. However, doubt remains about whether this type and intensity of intervention is effective at decreasing the rate of LGA newborns in high-risk pregnant women and the rate of caesarean sections. This study was not designed to address this question, and specific studies must be carried out to investigate the issue.

Regarding QoL, the significant decrease in the mean scores of the physical and social domains observed in both groups over time could be explained by the inconveniences typical of the end of pregnancy, arising from weight gain, pain and fatigue, as a consequence of the greater load on the musculoskeletal system. The perception of QoL was most strongly affected by the evolution of pregnancy, and was not affected by whether the women exercised during pregnancy. It is probable that QoL is affected by multiple factors, and so it is difficult to determine the effects of individual variables. It is worth mentioning that the literature is scarce on the assessment of QoL in this population.

In relation to overload on the musculoskeletal system in obese pregnant women, a previous study demonstrated that increased weight and a higher pre-gestational or gestational BMI were related to a higher prevalence of lower back pain during and up to 6 months after pregnancy.30 This finding is in agreement with the high prevalence of lower back pain in this sample (82.5% in the study group and 90.5% in the control group). This is one more indication that exercise should prove to be beneficial in this population.

A limitation of the study was that pre-gestational weight as reported by the pregnant women was used in determining pre-gestational BMI and total weight gain, which may have resulted in bias in these variables. However, this is a common limitation present in the majority of such studies because of the lack of records before pregnancy.17,21,23,24,27 Another significant limitation was the small sample size, which yielded low power in the analysis of the results, as already discussed. Nevertheless, our results can be used as a foundation for future clinical trials in this population, for which there is a lack of studies in the literature.

Another limitation of the study relates to difficulties in the management of obese pregnant women. In this study, the women had low adherence to the exercise programme and active lifestyle counselling. We believe that, for the obese women in particular, higher adherence would have increased the effectiveness of the intervention. Resistance to exercise may have been reinforced by the ancient cultural tradition that pregnancy should be a time of seclusion and rest.

In obese pregnant women, the evidence suggests that interventions should focus specifically on behavioural changes. Improving knowledge through educational programmes does not seem to be sufficient, and individualised interventions that combine diet and physical activity counselling seem to be required.18,24

Because weight gain during pregnancy is a strong determinant of postpartum weight retention,17 the prevention of excessive weight gain during pregnancy is fundamental to avoiding the development of obesity in pre-obese women of reproductive age. Furthermore, it can interrupt a vicious cycle of gradual increases in BMI between pregnancies that can lead to complications in subsequent pregnancies and a higher risk of diseases related to obesity in the future.


Although in this study we did not find significant differences in the control of weight gain between women who did and did not exercise, exercise was not related to adverse perinatal outcomes and did not affect arterial blood pressure or the perception of QoL. In future research, obese and morbidly obese women should not be excluded from programmes designed to encourage a healthy lifestyle during pregnancy. These programmes should include supervised exercise, as well as individualised nutritional and weight gain counselling. Pregnancy may be the best time to introduce such lifestyle changes.

Disclosure of interests

The authors have no disclosures of interests to make.

Contribution to authorship

SLN developed the research design, was involved in data collection, interpretation of the results and wrote the paper. FGCS contributed to the preparation of the research project, data collection, interpretation of the results and manuscript preparation. MÂP contributed to the preparation of the research project and the interpretation of the results. SS contributed to the analysis of the results and performed the statistical analysis of the data. JLPS contributed to the final editing of the manuscript.

Details of ethics approval

This study was approved by the Research Ethics Committee of the UNICAMP Medical School (FCM-UNICAMP) under registration number 542/2008 on 27 July 2008.


The principal investigator received financial funding from CAPES (Coordination of Improvement of Higher Education Personnel).