PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of wtpaEurope PMCEurope PMC Funders GroupSubmit a Manuscript
 
Eur J Clin Nutr. Author manuscript; available in PMC 2017 April 12.
Published in final edited form as:
PMCID: PMC5389446
EMSID: EMS37984

Maternal Activity in Relation to Birth Size in Rural India The Pune Maternal Nutrition Study

Abstract

Objective

To describe the relationship of the mother's physical activity to the birth size of her baby in a rural Indian population.

Design

Prospective observational study.

Setting

Six villages near Pune, Maharashtra, India.

Subjects

797 women were studied after excluding abortions and termination of pregnancies (112), fetal anomalies (8), multiple pregnancies (3), incomplete pre-pregnancy anthropometry (14) and pregnancies detected later than 21 weeks of gestation (168).

Method

An activity questionnaire was developed after focus group discussions and incorporated community-specific activities. It was validated against an observer-maintained diary. Activity scores were derived using published data on energy costs to weight the contributions of various activities. It was then administered to assess physical activity at 18 (± 2) and 28 (±2) weeks of gestation.

Outcome measures

Birth outcome, maternal weight gain and neonatal anthropometry.

Results

The activity questionnaire was used to classify women into light, moderate and heavy activity categories. Maternal activity did not influence the incidence of prematurity or stillbirth, or the duration of gestation. It was inversely related to maternal weight gain up to 28 weeks of gestation (p=0.002). Higher maternal activity in early, as well as mid gestation, was associated with lower mean birth weight (p=0.05 and 0.02 respectively), and smaller neonatal head circumference (p=0.005 and 0.009) and mid-arm circumference (p=0.03 and 0.01) after adjusting for the effect of major confounding factors.

Conclusions

The findings suggest that excessive maternal activity during pregnancy is associated with smaller fetal size in rural India. The approach described for developing an activity questionnaire has potential for adoption in other settings.

Sponsorship

Wellcome Trust, London, UK, and the Medical Research Council, UK.

Descriptors: India, Maternal activity, Birth size, Activity questionnaire

Introduction

Women from rural communities in developing countries like India have a high physical workload, including both farm labour and domestic chores. Energy expenditure could therefore be an important factor affecting the relationship between maternal nutrition and birth size. Manual work has been associated with low birth weight in undernourished women in developing countries (Tafari N et al, 1980; Launer LJ et al, 1990). Strenuous physical work has also been associated with increased rates of abortion and premature delivery (Teitelmann AM et al, 1990; Ahlborg G, 1995; Barnes DL et al, 1991). Maternal activity might therefore be a potentially modifiable risk factor for reducing low birth weight prevalence (Kramer MS, 1987). Studies from India have examined the associations between birth weight and maternal age, parity, lack of antenatal care (Trivedi CR & Mavalankar DV, 1986; Theodore K et al, 1992), social variables (Grover I, 1982) and maternal pre-pregnant nutritional status and dietary intake (Vijayalakshmi P & Lakshmi RN, 1985; Bhatia BD et al, 1983a; 1983b, Rao S et al, 2001), but few studies have examined the relationship between maternal physical activity and neonatal size.

We have recently carried out a study to examine the relationship between maternal nutrition, physical activity and birth size among women living in rural Maharashtra State, India. Direct measurement of energy expenditure in this community was not practical because the sophisticated equipment required is culturally unacceptable and disruptive to the rural life style of these women. On the other hand, one of the commonly used indirect methods of measuring energy expenditure, the diary method, was impractical as the majority of women were illiterate and not used to measuring time. Also, these women are too pre-occupied with their daily routine work, to co-operate in other elaborate methods of measuring activity.

Studies reported in the literature indicate that interviewer-assisted activity questionnaires can yield reliable data for assessing habitual physical activity (Phillippaerts RM & Lefevre J, 1998). Such questionnaires however, provide data on frequency and duration of activity but reveal little quantitative information about the intensity of physical activity. We set out to devise a questionnaire approach which would be simple, quick and feasible to administer, to study maternal activity among rural women from Maharashtra. Applying values of energy costs for different physical activities from the literature, we were able to convert the women’s activity into semi-quantitative scores. This helped to classify women into broad categories of light, moderate and heavy activity. We validated our approach against a conventional observer-maintained diary method and then used it to study the relationship of physical activity to birth size.

Methods

Subjects

The study was undertaken in 6 villages, 40-50 kms from Pune city and covered a population of approximately 35,000. Of 2675 married eligible women ( 15-40 y), 2,466 women (92%) agreed to take part. The area is drought prone and most families lived by subsistence farming on small landholdings. The majority of women (75%) worked on their farms or as labourers, in addition to their domestic chores which are tedious and hard. Most families are vegetarian and the usual diet consisted of pearl millet roti and dal or a vegetable curry. Socio-economic status was assessed using a standardized questionnaire (Pareek and Trivedi 1964), which derives a composite score based on occupation and education of the head of the household, caste, type of housing, and family ownership of animals, land and material possessions. Full details of the study including dietary intakes have been reported earlier (Rao S et al., 2001).

Gestation

Field workers visited the women every month to record the date of their last menstrual period ; women who missed 2 successive periods were examined by ultrasound at 15-18 wk to record sonographic gestational age (Hadlock 1990). Gestational age was derived from the last menstrual period , unless it differed from the sonographic estimate by more than 2 wks, in which case the latter was used. Women entered the study if a singleton pregnancy of less than 21 wk gestation was confirmed. All women reporting missing a period between June 1994 to April 1996 were thus enrolled in the study. Of the 1102 women enrolled, 797 women were studied after excluding abortions and termination of pregnancies (112), fetal anomalies (8), multiple pregnancies (3), incomplete pre-pregnancy anthropometry (14) and pregnancies detected later than 21 week of gestation (168). The method of workload assessment was developed during the pre-recruitment phase.

Weight

Women were measured every three months to record their weight. The last weight recorded before confirmation of pregnancy was used as pre-pregnant weight and the measurement was repeated during pregnancy at 18 ± 2 and 28 ± 2 wk gestation to get estimates of weight gain during pregnancy.

Nutritional intakes

The conventional 24-h recall method was modified and made more objective by incorporating information on portion sizes, which were weighed at each mealtime by a trained fieldworker. Women were interviewed at 18 and 28 wk of gestation by one of four nutritionists to record consumption of food items in chronological order from morning until dinnertime. At the time of diet survey interviewers ensured that the woman was not fasting or suffering from any illness and had reported foods consumed outside home on the day of visit. It was observed that, mean energy and protein intakes at 18 and 28 wk were: energy: 7.4 ± 2.1 MJ and 7.0 ±2.0 MJ ; protein: 45.4 ± 14.1g and 43.5 ±13.5 g respectively; were low - approximately 70% to 75% of recommended intakes (ICMR 1998) at both time points.

Development of the activity questionnaire

Focus Group Discussions (FGD's)

Use of FGD’s in social science research is well known but they have not been used in community-based nutrition studies in India. In all four FGD’s were held with groups of 8 to 10 women, representing young pregnant and elderly women. Women voluntarily participated in FGD’s. A tape recorder was kept at the centre with prior consent of the women and the discussions were moderated around 8 to 10 important relevant questions related to activity of rural women. The FGD’s aimed to obtain information about the different activities that women undertake in this rural community, to understand their perception of light and strenuous activities and to know their perception of the distance and time for activities involving walking. We also enquired if physical workload is altered during pregnancy and by season.

Activity questionnaire

The information gained from FGD’s provided insights into the lifestyle of rural women. The data obtained from the FGD assisted in developing the activity questionnaire, which was field tested for discriminating active women from others and was finalised (Appendix). Details of their typical daily activity from morning until evening were recorded under three broad categories as ‘resting’, ‘domestic’, and ‘other’ activities including farming work. Women who were not working on the farm were either engaged as construction workers or were running a small shop or stitching clothes, and were included under 'other' activities and constituted 13.9 % of the population.

In the ‘resting’ category women were asked about the time they went to bed and got up, whether and for how long they rested during the day and how many TV programs they watched in the evening. The ‘domestic’ category involved the variety of activities besides cooking, such as fetching water, washing utensils and clothes, fetching firewood, and cleaning animal sheds. These activities were recorded in terms of simple numeric measures such as number of roties prepared each meal time, number of trips to the well and number of animals cared for. Perception of the distance from home to the farm or to the well was recorded as `near’, `far’ or ‘mid-way’, while the time spent in actual farm work was recorded in terms of half day or full day and number of days in a week.

Activity scores

Using published data for the energy expenditure of activities (Lawrence M et al, 1985) a weighted score was derived, which reflected as a base unit an activity level of 1 kcal per minute for a 30 minute slot of time. For example, the time required to make up to 10 roties was around 30 minutes and considering the energy costs for bread making (2.3 kcal) the activity score for making 10 roties works out at 2.3 x 1 (one slot of 30 minutes). For making an additional 10 roties the score was increased by 25% as the initial time for dough preparation is saved. Similarly, farm work is listed as expending 3.5 kcal energy/minute; 6 hours of farm work had a score of 3.5 X 12 [6 hr x 2] = 42. For sleeping and resting activities a different weighting was used to reflect the fact that if a woman spends less time sleeping, she would be spending more time working. For all the activities in the questionnaire, the scores were computed (Appendix). A total daily activity score and separates scores for the components: resting, domestic and other were derived.

Validation

The time spent in each activity, and the energy expenditure scores obtained using the questionnaire were compared in 42 women from the same study population with those derived from a detailed observer-maintained diary. A trained nutritionist stayed with each woman throughout the day and recorded her minute-to-minute activity (including posture) from waking (7 a.m.) until bedtime (9 p.m.). Distances walked by the women were measured using a pedometer (CMS Instruments Ltd., London, UK). We were not able to select the subjects randomly as it depended on the willingness of the family for the women to be followed closely throughout the day. The short activity questionnaire was administered the following day by a different nutritionist.

The validation study aimed to determine i) whether the specific tasks recorded in the questionnaire reflected observed behaviour, ii) whether the simple numeric measures of various activities reflected the time spent in these activities and iii) to test and validate the perceptions of women about the `near’ or `far’ distance using a pedometer.

The activity questionnaire thus developed was administered at 18 ± 2 weeks and 28 ± 2 weeks gestation to 797 women in the study.

Birth outcome

Birth outcome was recorded as prematurity (gestation <37 weeks), stillbirth or livebirth. Birth size was measured in terms of several anthropometric measurements. Babies were measured by one of 5 trained fieldworkers within 72 hours of birth. Birthweight was measured to the nearest 50 g using a Salter spring balance (Salter Abbey, Suffolk, UK); crown-heel length to the nearest 0.1 cm using a portable Pedobaby Babymeter (ETS J.M.B., Brussels, Belgium). Triceps and subscapular skinfold were measured to the nearest 0.2 mm, on the left side of the body, using Harpenden skinfold callipers (CMS Instruments, London, UK). Occipito-frontal head circumference and mid-upper-arm circumference (MUAC) were measured to the nearest 0.1 cm using fiberglass tapes (CMS Instruments). Abdominal circumference was measured at the level of umbilicus in expiration. Placental weight was recorded to the nearest 5 g using Ishida scales, after trimming of the umbilical cord and membranes. Inter- and intra-observer variation studies were conducted 3-monthly to ensure quality control of these measurements.

Statistical Methods

Proportions of pre-term and low birth weights in activity groups were tested by χ2 test . Multiple regression analysis was carried out for examining the relationship of maternal activity with birth size after adjusting for major confounding variables. Maternal weight, sex of the new born, gestation, parity, and weight gain (28th wk) were direct predictors while social class (as a group variable), and protein and energy intakes (as a continuous variable) were indirect predictors. Comparisons of mean activity scores for farming and non- farming women were done using t-tests for independent samples. The data were analysed using the software package SPSS/PC v5.0.

Results

Focus group discussions revealed that almost all the women performed domestic activities that included cooking, washing clothes and utensils, sweeping the house and fetching water and firewood. Women perceived fetching water and washing clothes as the most strenuous activities. They often carried three containers filled with water, placing two on their head and one on their waist/hip. The capacity of each container was 10-15 litres (weighing in total about 45 kg). While fetching water, women were required to bend and pull a bucketful of water from the well, often without even using a simple pulley.

In addition, many women did farm work in which specifically ‘female’ tasks tended to be those requiring prolonged bending, squatting or standing. These included weeding, onion planting, harvesting groundnuts and threshing grains. Some women were also engaged in caring for and milking animals. Women had no recreation other than chit-chatting. Going to the market to sell or buy things was often done by an elderly man or woman in the house. During slack periods farming activities were replaced by other tasks such as stitching quilts, making pappadam and vermicelli stocks for using throughout the year. There was a strong belief among these women that working until late gestation led to an easier delivery.

It was clear from the focus group discussions that women were able to state with confidence the number of roties they prepared in a day or the number of trips they made to the well. They were not, however, able to describe the time taken in particular activities. Similarly, women were not able to assess distances accurately but perceived the farm or the well as `near’ or `far’. FGD’S were thus helpful in indicating possible simple numeric measures proportional to time or distance for various activities.

Validation results

The women who took part in the validation study were similar to those enrolled in the main study in respect of age (20.8 ± 2.9 years v 21.4 ± 3.5 years respectively) and pre-pregnancy weight (42.8 ± 5.3 kg v 41.6 ± 5.1 kg respectively) but were slightly taller (153.8 ± 4.5 cm Vs 151.9 ± 5.0 cm respectively, p< 0.05).

The actual observed time for several activities confirmed the assumptions made in formulating the activity questionnaire. Thus mean time observed for making up to 5 roties (measured on 15 women) was 24.4 minutes while that observed for 6 to 8 roties (measured on 11 women) was 40.1 minutes. Similarly, time spent in washing clothes (for 5 persons) was 28.3 minutes (measured on 15 women) while that for 5 to 10 persons (observed on 11 women) was 70.6 minutes. These observations revealed that time spent cooking was proportional to the number of roties/chapaties, that spent in washing clothes or utensils was proportional to the number of persons, and time spent caring for animals was proportional to the number of animals.

Perceptions of distance varied according to whether the woman was carrying a load or not. Thus, when examined with pedometers, it was observed that the average distance for walking without a load perceived as ‘far’ was higher (4.07± 2.67 km observed on 8 women) than that for walking with a load (1.5 ± 0.36 km observed on 3 women).

Mean time recorded during validation for various broad categories i.e. farming, domestic activity and resting were 323.4±127.8, 616.8±148.4 and 479.3±48.9 minutes respectively (Table 1). The mean activity scores estimated from the activity questionnaire show that despite large differences in time spent in domestic and farming work their scores were similar owing to the higher energy costs of farming activities. The correlations of the actual time observed with the activity scores were significant in all three broad categories of activities, as well as with the total activities. The negative correlation for the resting category indicates that women spending a long time resting had a lower activity score. Data collected on total daily observed time in various activities and daily activity score were cross-classified (2x2) on the basis of the median values, to examine sensitivity and specificity. The median was preferred as the majority of women were from farming families and the distribution of activity scores was skewed. It was observed that the activity scores had a high sensitivity (70%) and specificity (70%) for identifying women engaged in high levels of activity.

Table 1
Results of the validation study

Activity patterns of rural mothers (main study)

The activity pattern of the rural mothers in the main study based on the validated questionnaire (Table 2) shows that the majority of the women did cooking and washing clothes and utensils as their main domestic activities. Only 3.3 % of women did not report any cooking activity; this was due to an additional helping hand, such as an elderly woman, a sister-in law or a mother-in-law in the house. Women who did not cook were often involved in animal care or milking. Over 85 % of women fetched water and made an average of two trips to the well carrying 3 containers at a time, while 64.8% of women had to collect firewood. Fetching water contributed on average 17.6% to the total daily score while that for firewood collection contributed 2.7%. Washing utensils and clothes also made a considerable contribution (7.7%).

Table 2
Activity pattern of rural women at 18 weeks gestation

Sixty seven percent of women worked on the farm and of these 90% worked a full day. This activity had the highest mean contribution to the total daily score (38.7%). They had some rest (about 30 min.) during the lunch hour, while women not working on the farm could enjoy an afternoon nap in the house. Only 3% of women reported leisure activities such as chit-chatting. It was noticeable that 18% of women were breast feeding their youngest child, although many stopped in late gestation.

Mean activity scores for broad categories of activities are given in Table 3. Mean daily activity scores differed significantly (p < 0.001) between farming and non-farming groups. Activity scores in winter (harvesting season) were higher (89.9± 15.3) than in summer (85.8±15.8). Farming women performed similar levels of domestic activities to non-farming women in addition to hard work on the farm. Further, there was no significant difference between mean activity scores at 18 and 28 weeks gestation, in both farming and non-farming groups, indicating that physical activity at later gestation was not substantially reduced compared to that in earlier pregnancy.

Table 3
Mean (95% confidence intervals) activity scores in early & late gestation for women from farming(F) and non-farming (NF) families.

Physical activity, maternal weight gain and birth outcome

Of 797 women, 14 had late termination of pregnancy, while 1 died of pregnancy induced hypertension and 12 had spontaneous abortions. Of 770 singleton deliveries, 8 babies were stillborn, 9 had major congenital abnormalities and 51 did not have birth measurements. Sixty nine of the remaining 702 babies were born premature (<37 weeks gestation).

Maternal weight gain was not related to activity score at 18 weeks but was inversely related at 28 weeks gestation. Women in the lowest third of activity gained more weight (6.1± 2.9 kg) up to 28 weeks than women in the highest third (5.1± 2.7 kg; p<0.01).

Birth outcome was compared in thirds of total activity score at 18 and 28 weeks gestation (Table 4). There was no association between maternal activity and the incidence of prematurity or stillbirth. We were unable to examine associations with spontaneous abortion because of inadequate information (the earliest sonographic confirmation of pregnancy took place at 14± 2 weeks).

Table 4
Birth outcome and birth size by levels of maternal activity

Physical activity and birth size

This analysis was limited to the 633 women who delivered at term. The proportion of low birth weight ( LBW) babies (< 2500 g) was significantly (p <0.05) lower (28.6%) among women in the lowest third of activity at 18 weeks gestation than among those in the medium ( 38.4 %) or highest third (30.3 %). At 28 weeks however, proportion of LBW did not differ.

Associations of maternal daily activity scores with the babies’ measurements at birth were also examined (Table 4). After adjusting for direct as well as indirect predictors, activity score at 18 weeks was inversely related to birth weight (p=0.05), head circumference (p=0.005), mid-arm circumference (p=0.03) and placental weight (p=0.02). Similarly, activity score at 28 weeks was inversely related to birth weight (p=0.02), head circumference (p=0.009) and mid-arm circumference (p=0.01). Maternal activity was not related to neonatal body fat, as measured by triceps skinfold thickness.

The above analysis was performed separately for women with prepregnant weight below and above 45 Kg to examine how differences in prepregnant nutritional status affect associations of activity with various neonatal measurements (Table 5). None of the trends were significant for women with weight above 45 kg. In contrast, in the other group (i.e. < 45 Kg) they were significant and negative for head circumference (p=0.01) and placental weight (p=0.02) at 18th week of gestation and for weight gain (p=0.002) at 28th week of gestation.

Table 5
Birth outcome and birth size by levels of activity for pre-pregnant weight <45 kg of rural mothers

Separate analyses were carried out for the subcategories of total activity, ‘domestic’ and ‘other’, which included farming activities (Table 6). Domestic activity was inversely associated with birth weight (p=0.01), head circumference (p=0.006), mid arm circumference (p=0.05) and placental weight (p=0.03) only at 28 weeks. However, farming activity at 18 weeks was inversely related to birth weight (p=0.02), head circumference (p<0.01) and mid arm circumference (p=0.03), and these associations remained significant at 28 weeks.

Table 6
Birth outcome and birth size by levels of domestic and farming activity

According to the women’s perception, fetching water was the most strenuous activity. At 28 weeks gestation this single activity was inversely associated with birth weight (p<0.001), head circumference (p=0.01), mid-arm circumference (p=0.03) and placental weight (p< 0.01) after adjusting for the confounding variables (Table 6). Inverse association of baby’s triceps skinfold thickness was seen (p=0.02) only when activity of fetching water was considered separately which was not seen for the total activity score. Babies born to women who were not performing this strenuous activity were heavier by 112 g than those born to women who fetched water.

Discussion

We have earlier studied the maternal intakes in relation to birth size among these rural mothers (Rao et al 2001). Maternal energy intakes showed no significant relationships with neonatal size. However, the relationship between maternal nutrition and fetal growth cannot be fully understood by measurement of maternal food intake alone. In communities where women are involved in hard work like farming, consideration of physical activity becomes essential. Direct methods of measuring energy expenditure are expensive and complicated to perform in field conditions especially in a place like rural India. A large number of alternative approaches have emerged in the literature, including simple categorisation of subjects as ‘active’ and ‘non-active’ (Washburn RA et al, 1990), 7 day activity recall (Warwick PM & Macqueen SE, 1988; Schoeller DA & Racette SB, 1990) or 24-hour recall (Bernstein M et al, 1998), using activity questionnaires. In simplifying conventional methods, it is advantageous to take into consideration community peculiarities. We therefore developed a community-specific activity questionnaire. We observed that maternal activity was inversely related to maternal weight gain up to 28 weeks, birth weight, head circumference and mid-arm circumference of the newborn.

Focus group discussions provided valuable information about the women's' lifestyle and more importantly helped identify simple numeric variables to quantify the intensity of various activities. Our questionnaire was quick and easy to administer and was therefore applicable with minimal disruption to their daily activity. It performed well in validation when compared with minute-to-minute observer maintained diary and was able to reveal the differences in activity patterns of farming and non-farming groups.

Our questionnaire revealed that domestic tasks contribute a high proportion to the woman’s daily activity and that farming women had a similar domestic workload to that of non-farming women. Time spent in farming activities was comparable to that observed by Bleiberg F et al, (1980) in female farmers from Upper Volta. Time spent in domestic work was however, considerably more than that reported in other studies, effectively reducing their resting time which was less than that reported in Upper-Volta (Bleiberg F et al, 1980; 1981). In developing countries excess physical activity may reduce maternal weight gain (Langhoff RJ et al, 1987). In our study, among women in the highest third of activity score, weight gain up to 28 weeks gestation was significantly lower than that for women in the lowest third.

Weight gain during pregnancy is known to be a strong determinant of birth size and several factors are known to influence weight gain in pregnancy. Pre-pregnancy weight has been shown to be negatively associated with weight gain in poor and undernourished women (Kirchengast S & Hartmann B 1998). In women from Sri Lanka, maternal weight gain was inversely associated with parity but positively with income and maternal education (Wanalawansa S J & Wikramanayake T W 1987). Dietary energy intakes were associated with weight gain in pregnant women from Bangla Desh (Tawfeek et al., 1999) while restrained eaters were observed to experience significantly lower weight gain compared to their recommended range based on pre-pregnant BMI (Lonway R. et al., 1999). Our data also shows association of weight gain (28th wk) with pre-pregnant weight (p < 0.000), social class (p = 0.02), gestation (p = 0.03), parity (p = 0.003), energy intake (p = 0.01) and protein intake (p = 0.01). However, the association between maternal activity and weight gain has been investigated only by few (Dewey et al., 1994; Agarwal et al., 2001). Our study showed that maternal activity, especially during mid gestation had an inverse effect on weight gain even after controlling for all above mentioned maternal factors.

We observed that maternal activity was not associated with the occurrence of prematurity or stillbirth, or with length of gestation. Some studies offer supportive evidence for this relationship ( Naeye R L and Peters E C 1982, Teitelman A M et al 1990) ) while others donot show relationship between employment category and pregnancy outcome (Berkowitz, et al. 1983; Mayer et al., 1985; Hartikanen-Sorri et al., 1989). The variation in the result could be partly due to the fact that most studies have not adequately controlled for potentially confounding variables (Dewey K.G., 1994) or due to variations in categorisation of jobs as sedentary or active.

Maternal activity was however, associated with the incidence of low birth weight in our study. In fact, total daily maternal activity at 18 and 28 weeks was negatively associated with almost all the birth measurements except length. Separate analysis for farming activity scores showed similar associations. The odds ratio for delivering a low birth weight baby was 1.93 (95% CI: 1.47 to 2.39) at 18th wk and 1.63 (95% CI: 1.21 to 2.05) at 28th wk among women engaged in farming compared with mothers not doing farming (as the reference category). Lima M. et al., 1999, had also reported that heavy agricultural work throughout pregnancy significantly reduced birth weight in low income north-east Brazilian women. In the case of 'domestic' activity, we observed inverse association with birth size only at 28 weeks.

Separate data analysis for women with pre-pregnant weight below and above 45 kg showed that the trends with the activity at 18th week were significant only for head circumference and placental weight and that at 28th week with weight gain, mainly among the undernourished women (< 45 kg). The adverse effects of excessive or strenuous activity therefore, are worse in undernourished women. Agarwal et al., 2001, too have reported that hard activities in undernourished Indian rural women in later pregnancy reduced fetal weight as well as length. In relatively well-nourished US women, moderate exercise has only a small effect on birth weight and in some reports it is associated with higher birth weight possibly because of improved placental blood flow. Strenuous exercise on the other hand, appears to reduce birth weight in both populations (Clapp J.F., 2000).

The relationship between maternal activity and birth size was strong for one specific activity namely fetching water, which was perceived as a strenuous activity by rural mothers. Lima et al. (1999) has also found lower mean birth weights among mothers who fetched water during pregnancy. An adverse influence of prolonged standing on birth outcome and birth size has been reported in Filipino women (Barnes DL et al., 1991). Many investigators (Simson JL, 1993; Keith L & Luke B., 1991) have speculated that strenuous occupations increased uterine contraction and therefore increased the risk of pre-mature birth. The biological basis for a harmful effect of heavy work on pregnancy outcome has been identified as it decreases uterian and placental blood flow, thereby reducing the fetal supply of oxygen and nutrients which restricts intra-uterian growth ( Lima M et al.1999). In contrast, leisure time exercise, especially in the second trimester was observed to protect against pre-term delivery (Misra DP et al., 1998; Berkowitz GS et al., 1983).

We found that higher maternal activity scores in earlier as well as later pregnancy were associated with lower mean birth weight, head circumference and mid-arm circumference The fact that these relationships were even stronger for the strenuous activity of fetching water, points towards the possible effects of certain postures such as bending in this case. It is speculated that non supine postures may affect utero placental blood flow (Briend A 1979, Suonio S A et al, 1976) and therefore birth size. Our observation that higher activity in early gestation resulted in lower placental weight gives some support to this speculation. Physiological consequences of maternal activities involving certain postures thus needs further investigation. Interestingly, measures of neonatal fat were related only to this strenuous activity. A negative association of maternal physical activity with head circumference has not been reported before and raises the question whether this indicates an adverse impact on brain growth.

In conclusion, reduction in maternal physical activity offers significant means for improving neonatal size in this rural Indian community. It is often assumed that pregnant women can economize their energy expenditure by curtailing some activities (Ferro- Luzzi A 1980). However, the extent to which such economization is possible may be limited by social, seasonal or other constraints (Panter-Brick C., 1993). For example, in rural India , social beliefs, such as desire for more sleep during pregnancy is interpreted to be a sign of a female fetus, or working until late gestation is thought to result in an easy delivery, may create difficulties in reducing maternal activity. Programmes aimed at reducing strenuous workload during pregnancy will therefore, need to adhere to these community beliefs before it is likely that recommended changes will be adopted. The impact of changes in workload on household food security will also need to be considered.

Our findings demonstrate the strength of FGDs in designing a community specific activity questionnaire. This may improve the chances of demonstrating a relationship between physical activity and fetal growth when sophisticated methods are not practical. We therefore feel that the approach described in this study has the potential for adaptation for other settings, especially for rural communities in developing countries where women have monotonous lives with clearly defined routine tasks. Finally, the findings in this study suggest that limiting maternal strenuous activities could be a potential intervention for improving birth size in rural Indian community.

Supplementary Material

Appendix

Acknowledgement

We are grateful to the community, and to the pregnant women and their families for their co-operation. The study was supported by the Wellcome Trust, London, and the Medical Research Council, UK. We thank Dr AD Agate, Director, Agharkar Research Institute and Dr VN Rao, Director, the KEM Hospital Research Center for providing facilities for this collaborative research.

References

  • Agarwal S, Agarwak A, Agarwal KN, Agarwal DK, Bansal A. Brief Reports. Indian Pediatrics. 2001;38:1017–1022. [PubMed]
  • Ahlborg G. Physical workload and pregnancy outcome. J Occ Envir Med. 1995;37:941–944. [PubMed]
  • Barnes DL, Adair LS, Popkin BM. Women’s physical activity and pregnancy outcome : A longitudinal analysis from the Philippines. Int J Epid. 1991;20(1):162–172. [PubMed]
  • Berkowitz GS, Kelsey JL, Holford DR. Physical activity and risk of spontaneous pre-term delivery. J Reprod Med. 1983;28:581–588. [PubMed]
  • Bernstein M, Sloutskis D, Kumanyika S, Sparti A, Schutz Y, Morabia A. Data-based approach for developing a physical activity frequency questionnaire. Am J Epidemiol. 1998;147(2):147–154. [PubMed]
  • Bhatia BD, Banarjee D, Agarwal DK, Agarwal KN. Fetal growth : Importance of maternal body size and dietary intakes during third trimester. Ind J Pediat. 1983a;50(402):1–8. [PubMed]
  • Bhatia BD, Banarjee D, Agarwal DK, Agarwal KN. Fetal growth : Relationship with maternal dietary intakes. Ind J Pediat. 1983b;50(403):113–120. [PubMed]
  • Bleiberg F, Brun AT, Goihman S, Lippman D. Food intake and energy expenditure of male and female farmers from Upper Volta. J Nutr. 1981;45:505–515. [PubMed]
  • Bleiberg FM, Brun TA, Goihman S. Duration of activities and energy expenditure of female farmers in dry and raining seasons in Upper Volta. Br J Nutr. 1980;43:71–81. [PubMed]
  • Briend A. Fetal malnutrition - the price of upright posture? Br Med J. 1979;2:317–319. [PMC free article] [PubMed]
  • Clapp JF. Exercise during pregnancy; a clinical update. Clinics in Sports Medicine. 2000;19:273–286. [PubMed]
  • Dewey KG, McCrony MA. Effects of dieting and physical activity on pregnancy and lactation. Ame Jour Clin Nutr. 1994;59(Suppl):446S–453S. [PubMed]
  • Ferro-Luzzi A. Work capacity and productivity in long term adaptation to low energy intakes. In: Blaxter K, Waterlow JC, editors. Nutritional adaptation in Man. London: John Libbey; 1985. pp. 61–67.
  • Grover I. Effects of dietary intake, maternal factors and socio-economic factors on birth weight of infants in rural Haryana. The Ind J Nutr Dietet. 1982;19:80–86.
  • Hadlock FP. Sonographic Estimation of Fetal Age and Weight. Fetal Ultrasound. 1990;28:39–50. [PubMed]
  • Hartikainen-Sorri AL, Sorri M. Occupational and socio-medical factors in pre-term birth. Obstet Gynecol. 1989;74:13–16. [PubMed]
  • Indian Council of Medical Research. Nutrient requirements and recommended dietary intakes for Indians. ICMR; New Delhi: 1998.
  • Jarski RW, Trippett DN. The risks and benefits of exercise during pregnancy. J Fam Pract. 1990;30:185–189. [PubMed]
  • Keith L, Luke B. The association between women's work, working conditions and adverse pregnancy outcomes: a review of the literature and directions for future research. Women's Health Inst. 1991;1:113–119. [PubMed]
  • Kirchengast S, Hartmann B. Maternal prepregnancy weight status and pregnancy weight gain as major determinants for newborn weight and size. Ann Hum Bio. 1998;25(1):17–28. [PubMed]
  • Kramer MS. Determinants of low birth weight : methodological assessment and meta-analysis. Bulletin World Health Organisation. 1987;65:663–737. [PubMed]
  • Langhoff-Roos J, Lindmark G, Kylberg E, Gebre-Medhin M. Energy intake and physical activity during pregnancy in relation to maternal fat accretion and infant birth weight. Brit J Obstet Gynaecol. 1987;94(12):1178–1185. [PubMed]
  • Launer LJ, Villar J, Kestler E, de Onis M. The effect of maternal work on fetal growth and duration of pregnancy : a prospective study. Br J Obstet Gynaecol. 1990;97:62–70. [PubMed]
  • Lawrence M, Singh J, Lawrence F, Whitehead RG. The energy cost of common daily activities in African Women: increased expenditure in pregnancy? Am J Clin Nutr. 1985;42:753–763. [PubMed]
  • Lima M, Ismail S, Ashworth A, Morris SS. Influence of heavy agricultural work during pregnancy on birth weight in North-East Brazil. Int J Epidem. 1999;28:469–474. [PubMed]
  • Lonway R, Reddy S, Davies J. Dietary restraint and weight gain during pregnancy. Europ Jour Clin Nutr. 1999;53(11):849–853. [PubMed]
  • Lotgering FK, Gilbert RD, Longo LD. The interactions of exercise and pregnancy : A review. Am J Obstet Gynecol. 1984;149:560–568. [PubMed]
  • Meyer BA, Daling JR. Activity level of mother's usual occupation and low infant birthweight. Jour Occup Med. 1985;27:841–847. [PubMed]
  • Misra DP, Strobino BM, Stashinko EE, Nagey DA, Nanda J. Effects of physical activity on pre-term birth. Am J Epidemiol. 1998;147(7):628–635. [PubMed]
  • Naeye RI, Peters EC. Working during pregnancy: Effect on fetus. Pediatr. 1982;69:724–727. [PubMed]
  • Panter-Brick C. Seasonality of energy expenditure during pregnancy and lactation for rural Nepal women. Am J Clin Nutr. 1993;57:620–628. [PubMed]
  • Pareek U, Trivedi G. Reliability and valibility of socio-economic case. Ind J Appl Psychol. 1964;1:34–40.
  • Phillippaerts RM, Lefevre J. Reliability and validity of three physical activity questionnaires in Flemish males. Am J Epidemiol. 1998;147(10):980–990. [PubMed]
  • Rao S, Yajnik CS, Kanade AN, Fall CHD, Margetts BM, Jackson AA, Shier R, Joshi S, Rege S, Lubree H. Intake of micronutrient-rich foods in rural Indian mothers and size of their babies at birth (Pune Maternal Nutrition Study) Journal of Nutrition. 2001;131:1217–1224. [PubMed]
  • Schoeller DA, Racette SB. A review of field techniques for the assessment of energy expenditure. J Nutr. 1990;120(suppl):1492–1495. [PubMed]
  • Simpson JL. Are physical activity and employment related to pre-term birth and low birth weight ? Am J Obstet Gynecol. 1993;168:1231–1238. [PubMed]
  • Suonio S, Simpanen A-L, Olkkonen H, et al. Effect of the left lateral recumbent position compared with the supine and upright positions on placental blood flow in normal late pregnancy. Ann Clin Res. 1976;8:22–26. [PubMed]
  • Tafari N, Naeye RL, Gobezie A. Effects of maternal undernutrition and heavy physical work during pregnancy on birthweight. Br J Obstet Gynaecol. 1980;87:222–226. [PubMed]
  • Tawfeek HL, Abdulla JN, Rasheed AH. Maternal dietary intakes and pregnancy outcomes in Baghdad, Iraq. Food & Nutrition Bulletin. 1999;20(2):179–182.
  • Teitelmann AM, Welch LS, Hellenbrant KG, Bracken MB. Effect of maternal work activity on preterm birth and low birthweight. Am J Epidemiol. 1990;131:104–113. [PubMed]
  • Theodore K, Abel R, Sampathkumar V. Epidemiological correlates of low birth weight in rural Tamil Nadu. Ind J Pediat. 1992;59:299–304. [PubMed]
  • Trivedi CR, Mavalankar DV. Epidemiology of low birth weight in Ahmedabad. Ind J Pediat. 1986;53:795–800. [PubMed]
  • Vijayalakshmi P, Lakshmi RN. Reproductive performance of expectant mothers. The Ind J Nutr Dietet. 1985;22(2):36–41.
  • Wanalawansa SJ, Wikramanayake TW. Factors affecting weight gain during pregnancy and the growth of the infant. Ceylon J Med Sci. 1987;30(1):21–38.
  • Warwick PM, Macqueen SE. Retrospective assessment of level of activity and energy expenditure - a pilot study. Proceedings of the Nutrition Society of Australia. 1988;13:149.
  • Washburn RA, Goldfield SRW, Smith KW, MacKinlay JB. The validity of self reported exercise induced sweating as a measure of physical activity. Am J Epidemiol. 1990;132(1):107–113. [PubMed]