Repeated measurement of maternal weight during pregnancy. Is this a

0 downloads 0 Views 410KB Size Report
Is this a useful practice? M. G. DAWES, J. G. GRUDZINSKAS. Summary. A retrospective study of 1092 pregnant women showed that the use of a centile chart of ...
British Journal of Obstetrics and 'Gynaecology Fcbruary 1991, Vol. 98, pp. 189194

ADONIS 03(xi545691OOW5P

Repeated measurement of maternal weight during pregnancy. Is this a useful practice? M. G. DAWES, J. G. GRUDZINSKAS

Summary. A retrospective study of 1092 pregnant women showed that the use of a centile chart of maternal weight gain was not effective at detecting women who give birth to small for gestational age (SGA) infants. Multiple regression analysis showed that of the various weight measurements recorded, weekly weight gain and maternal size at booking were the only factors which had a significant association with infant birthweight after taking into account maternal gestation, age, smoking habit and parity. Low maternal booking weight (< 51 kg) was the most effective maternal weight measurement for antenatal detection of SGA infants (positive predictive value 20.0%). Low average weekly maternal weight gain (< 0.20 kg) had apositive predictive value of only 12.9% for detecting these pregnancies. Weight loss or failure to gain weight over a 2-week interval occurring in the third trimester was observed in 46% of all women studied. Maternal smoking had a positive predictive value of 16.3% in antenatal detection of SGA infants. Maternal weight need be recorded only at booking, with the exception of patients in whom nutrition is of concern. The subsequent routine weighing of patients may produce unnecessary anxiety and should cease.

Small for gestational age (SGA) and low birthweight (LBW) infants contribute to a major proportion of pcrinatal deaths and childhood morbidity and thc antenatal identification of these infants rcmains a priority for antenatal care. Maternal weight gain has been estimated routinely in antenatal clinics since 1941 (Scott & Benjamin 1948) when it was introduced to monitor maternal nutrition. There is little evidence that maternal weight gain is an effective anteDepartment of Obstetrics, John Radcliffe Hospital, Oxford DHA 087 M. G . DAWES Academic Unit of Obstetrics and Cynaecology, The London Hospital Medical College, Whitechapel, London El 2AD I. G . GRUDZINSKAS

Correspondcnce: M. G. Dawes

natal screening test for SGA infants (Gordon e ul. 1978; Elder eral. 1970) although studies hav, shown a positive association between materna weight gain and infant birthweight (Beilly f Kurland, 1945; Thomson & Billewicz 1957 Nyirjesy el al. 1968; Singer et al. 1968; Niswan dcr et al. 1969; Simpson et al. 1975; Miller et U I 1976; Rosso & Luke 1978; Harrison et al. 1981 Luke et al. 1981; Ounsted & Scott 1984). In this study we have examined systematical1 pregnancy outcome and complications i relation to the extremes of maternal weight and its change throughout pregnancy. First, we have explored the effectiveness of using a centile chart of maternal weight gain in dctecting womcn who subsequently give birth to SGA infants. Second, we have examined the booking weight, weight gain between 28 and 36 weeks, and average weekly weight gain over the whole prcgnancy and the association between these 189

190

M . G . Dawes & J. G . Grudzinskas

measures and infant weight allowing for gestation, maternal age, snioking and panty. Third, we have examined the association between excess maternal weight gain and hypertension. Subjects and methods

The study population ( n = 2580) was derived from women delivered in the John Radcliffe Hospital in Oxford during a 12-month study period who gave addresses in areas from where morc than 85% of the population are delivered in the John Radcliffe Hospital. From this population every alternate patient ( n = 1290) was entered into the study. The maternal sociodemographic characteristics, details of the course of the pregnancy, and its outcome were obtained retrospectively from hospital records and general practitioner co-operation cards. The data were entered onto the Oxford University computer twice and an edit program was used to screen for errors in data collection and entry. Matcrnal weight gain was calculated by subtracting weight at the booking visit from all weights recorded during pregnancy. Small-forgestational age was defined as a birthweight < 10th ccntilc for gestational age and sex in normal pregnancies based on standards derived from women delivered in Oxford (Yudkin et al. 1987). The scales used for weighing patients were predominantly beam and spring balanccs. The validation of these scales has been describcd previously (Dawes & Grudzinskas 1991). The data from 198 of the 1290 women wcrc not included for analysis to produce the centile chart as there was no record of infant sex or birthweight, or fewer than three weights recorded during pregnancy. The mean number of clinic attendances after booking with weight recorded was 8.4 (SD 3.6). The average wcight gain between 28 and 32 weeks, 32 and 36 wccks, and 28 and 36 weeks was calculated as was average weekly wcight gain. The wcckly rate of maternal weight gain was computed by dividing the total weight gain by the number of weeks between booking and the gestation at which the last weight was recorded. The body mass index (BMI) was calculated from the formula weight/height2 (kg/m’). The Central Oxford Research Ethics Committee approved thc study. Results

There were 507 primigravidae, and in 1035preg-

nancy lasted between 37 and 41 weeks. The loth, 50th and 90th centiles of maternal weight gain of 988 women booked before 20 weeks and delivered between 37 and 41 weeks were calculated for each gestational week from week 10 to week 40. There was a very large range of maternal weight gain in the 103 women who gavc birth to SGA infants (Fig. 1). Thc proportion of women (booking beforc 16weeks) who had SGA infants was not statistically significantly different between the group of women whose weight gain was < 10th centile at any time between 20 and 40 weeks gestation and the group whose weight gain never fell below the 10th centile. The sensitivity, specificity, positive and ncgative predictive values of using the test ‘maternal weight gain falling below the 10th ccntile between 20 and 40 weeks gestation in women booking before 16 weeks’ to detect women subsequently giving birth to SGA infants was 19.4%, 87.1%, 12.7%, 91.8% respectivcly. The correlation cocfficient between total weight gain and infant weight was 0426. The correlation coefficient betwccn infant birthweight and maternal vanablcs were calculated (Table 1). To determine whether maternal weight gain contributed to the variation of birthweight independently of gestation, maternal age, parity, and smoking habits, we performed a stepwise multiple linear regression analysis in which the measurements of wcight gain were cntered after these variables. Stepwise regression was performed with all variables included in thc model and then removing each non significant variablc at a time until only variables that were significant predictors of birthweight were

I - 1 1 1 1

I

I

I

I

I

I

I

I

I

I

j

10 12 14 16 18 20 22 24 26 28 3032 3436 3840

Week of gestation Fig. 1. Maternal weight gain in womcn (n = 103) who gavc birth to small-for-gestationalage (SGA) infants. Superimposcd are the 10th. 50th and 90th centiles of 988 women who booked beforc 20 weeks, and wcrc dclivered between 37 and 41 wccks gcstation.

Weight gain in pregnancy

191

Table 1. Corrclation coefficient between infant birthweight and maternal variables

Birthweight (kg) Gestational age (weeks) Maternal age (years) Parity Smoking Booking weight (kg) BMI Average weekly weight gain for total pregnancy (kg) Average weekly weight change (kg) between: 28-32 weeks 32-36 weeks 28-36 weeks

Mean

(SD)

Correlation coefficient

3.32 39.5 26.6 0436 3.23 63.1 24.3

(0.60) (2.52) (5.1) (1.0) (6.5) (11-3) (4.1)

0.06 0.10 -0-15 0.23 0.15

0.38

(0.15)

0.20

0.41 0.48 0.46

(0.32) (0.38) (0.29)

0-08 0.15 0.20

0-56

between 20 and 40 weeks gestation in women booking before 16 weeks) was 26%, 80%, 15%, and 89% respectively. The 10th centile of maternal weight gain was also assessed as a screening test for women who subsequently give birth to infants with low Apgar scores (64) or infants requiring admission to a special care baby unit. The sensitivities of this test for these conditions were 16% and 21% respectively with positive predictive values of 4% and 5% respectively. In women who smoked at all during pregnancy (n = 269, 16.4%) the frequency of SGA infants was statistically significantly higher than in women who did not smoke (n = 828, 7.1%) (P< 0.001). There was no statistically significant difference in maternal booking weight between smokers and non-smokers but smokers were statistically significantly younger (P< 0.001). Smoking was assessed as a predic-

included. Average weekly weight gain, BMI and booking weight were the only weight related variables that were left after this analysis (Table 2). Low booking weight, low average weekly weight gain and low booking BMI were all assessed as screening tests to determine women at risk of having SGA infants (Table 3). The frequency of raised blood pressure (BP > 140/90 on one occasion) was statistically significantly higher in those women whose maternal weight gain exceeded the 90th centile for maternal weight gain between 10 and 40 weeks ( n = 236, 16%) than in women whose weight gain was not greater than the 90th centile ( n = 855, lo%, P < 0.05). There was, however, a large variation in weight gain in women with high blood pressure (Fig. 2). The sensitivity, specificity, and predictive value of a positive and negative result of the screening test (maternal weight gain rising above the 90th centile

Table 2. Prediction of birthweight: stepwise multiple linear regression analysis

Step

Partial R2

Variable entered ~~~~

1 2 3 4 5 6 7

Gestation Weight Weekly weight gain Age Smoking

~

Modcl R2

B (partial regression coefficient)

F Statistic

Prob > F

309.6 29.8 26.3 6.5 9.7 7.8 9.9

< 0.001 < 0.001 < 0.001 < 0.02 c 0.01 < 0.01 < 0.01

~

BMI

0.271 0-039 0426 0.016 0-007 0.007

Parity

0.006

0.271 0-31 0.336 0.352 0.359 0.366 0.372

0-116 0.019 0.593 0.009 -0440

-0428

0.059

192

M . G . Dawes & J . G. Grudzinskas

Table 3. Low booking weight, average weekly gain, and low BMI at booking as predictors of SGA infants

Sensitivity Specificity Booking weight < 51 kg Avcrage weekly weight gain Booking BMI < 20

< 0.20 kg

21.3%

92.1%

12.9%

91.3% 90.4%

13.8%

tor of SGA for gestational age infants. The sensitivity, specificity, predictive value of a positive and negative result were 43%, 9370, l6%, and 77% respectively.

Discussion Since the report from the Medical Officer of Health in 1929 suggested that pregnant women should attend antenatal clinics at 16, 24 and 28 weeks gestation, then every 2 weeks until 36 wceks and weekly thereafter until delivery, the pattern of antenatal care in this country has remained the same. The routine weighing to monitor the nutrition of all pregnant women was begun in antenatal clinics in London in 1941 (Scott & Benjamin 1948). Although maternal nutrition remains an important element of antenatal care the antenatal detection of SGA or low birthweight babies has become one of the priorities in antenatal care (Gordon 1977). Dcspite the research describing the association of maternal weight gain and fetal growth, the cffectiveness of routine measurement of maternal weight gain to detect women at risk of having a growth retarded fetus remains unclear (Thomson & Billewicz 1957, Singer el al. 1968, Eldcr et al. 1970, Gordon et al. 1978, Rosso 1985, Abrams & Laros 1986, Lawton et al. 1988). We havc confirmed that weight gain was positively association with infant birthweight but that the correlatioii was low. Our findings are similar to those found by Rosso (1985) and Abrams & Laros (1986) but less than that found in carlier work (Nyirjcsy et al. 1968, Simpson 1975). We have also confirmed that the major detcrminants of birthweight, after taking into consideration gestational age, were maternal size at boolmg, rnatcrnal weight gain and matcrnal ake but that these only accountcd for 9% of the variation in infant birthweight. The higher prevalence of SGA infants born to womcn with low weight gain reflected this association, but examination of the value of the

Predictive Value +VE test

Predictive Value - VE test

20.0% 11.8%

80.0n/o 88.2Y"

13.9?Jo

86.1%

ccntile chart of maternal weight gain in screening for SGA infants indicates that this is not an effective test. In fact thc centile chart only meets two of the criteria set down for a screening test (Grant & Mohide 1982), namely that it is cheap and acceptable. Multiple regression analysis showed that the weight changes occurring between 28 and 32,32 and 36 and 28 and 36 weeks were not significantly associatcd with infant weight aftcr taking into account gcstation, maternal age, smoking and parity. By comparison, booking weight, booking BMI, average weekly weight gain did have a significant effcct on infant birthweight. The sensitivity of using low average weekly weight gain to detect women at risk of having SGA infants was considerably lower (12.9%) compared with booking weight (21.3%). Our findings confirm those by Elder et al. (1 970) that the increascd prevalence of SGA infants is associated with weight loss or static weight in the third trimester, but that this has no practical predictive value. Weight loss or failure to gain weight over a 2-week interval in the third trimester was observed in 46% of all women studied. The factor that had the highest association with weight gain was maternal booking weight

I

20 r

1

l

1

!

1

1

1

1

1

1

1

1

1

1

]

10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 Week of gestation

Fig. 2. Maternal weight gain in womcn (n = 236) found to havc raised blood pressure during pregnancy. Supcrimpxed are thc 10th. 50th and 90th centilcs of 988 worncn who bookcd before 20 weeks, and were delivered at bctwccn 37 and 41 wceks gcstation.

Weight gain in pregnancy (Peckham & Christiansen, 1971; Dawes & Grudzinskas 1991). It is likely that the results of maternal weight gain in detecting SGA infants wcre mostly due to the effect of booking weight on weight gain and we have shown that this factor was independently a better predictor of SGA infants than maternal weight gain. The predictive values of booking weight for SGA infants were low compared with two ultrasound scans which have a negative predictive value of 99.6% and a positive predictive value of 28% (Neilson et al. 1984). Maternal weight gain has been incorporated into multifactorial risk scores as a screening test for SGA infants with encouraging results (Cnattingius et al. 1985; Wennergren 1982). The most important factors for primigravidae were maternal smoking followed by symphysis-fundal hcight, and maternal weight before pregnancy. The weighted value of weight gain in these scoring systems is small. Futhcrmore as the detection of oedema is no longer recommended as a diagnostic feature of pre-eclampsia (Nelson 1955) there is no argument for using weight gain to detect fluid increase. Since the early 1940s body weight has achieved a prominent place on the antcnatal record, next to blood pressure, but with little evidence to support its importance. The theory that maternal weight gain might predict infant weight is not supported by thc evidence presented. Weighing generates anxiety in patients and requires as much cxplanation as any other screening procedure. The false assurance given to women who gain a ‘normal’ amount of wcight is as wrong as the worry caused by ‘low’ or ‘high’ weight gain. The woman who has a high or low weight at antenatal booking nccds investigation for a possible organic or psychiatric cause. In the absence of either of these. referral to a dietician should be considered. The routine recording of maternal weight gain in pregnancy in women with a normal weight at booking should cease.

Acknowledgments This study was supported by a grant from the Royal College of General Practitioners. I thank Mr M. D. Gillmcr, Dr M. Pike and staff at the ICRF, Dr K. McPhcrson, and Mrs S. Smith for their help and advice.

193

References Abrams B. F. & Laros R. K. (1986) Preprcgnancy weight, weight gain, and birth weight. A m J Obstet Gynecoll54,503-509. Beilly J. S . & Kurland I. I. (1945) Relationship of matcrnal weight gain and weight of the newborn infant. A m J Obstet Gynecol50,202-226. Cnattingius S . , Axellson O., Eklund G. & Lindmark G. (1985) Scrccning for intra uterine growth retardation in late pregnancy. Early Hum Dev 10, 225-235. Dawes M. D. & Grudzinskas J. G. (1991) Patterns of maternal wcight gain in pregnancy. Br J Obstet Gynaecol98, 195-201. Elder M. G., Burton E. R., Gordon H., Hawkins, D. F. & McClure-Browne J. C. (1970) Maternal weight and girth changes in late pregnancy and the diagnosis of placental insufficiency. J Obsfef Gynaecol Br Commwlrh 77,481-491. Gordon R. R. (1977) Neonatal and ‘perinatal’ mortality rates by birth wcight. Br Med J 2,1202-1204. Gordon Y . B.. Lewis J. D., Pendelbury D. J., Lcighton M. & Gold J. (1978) Is mcasurement of placental function and maternal weight worthwhile? Lancer i, 1001-1003. Grant A. & Mohidc P. (1982) Screening and diagnostic test in antcnatal care. In Effectiveness andSarisfaction in Antenatal Care. (Enkin M., Chalmers I., eds), International Medical Publications, London, pp. 22-27. Harrison G. G., Udall J. N. & Morrow G. (1980) Maternal obesity, weight gain in pregnancy, and infant weight. A m J Obstet Gynecol 136,411-412. Lawton F. G., Mason G. C., Kelly K. A,, Ramsay 1. N. & Morcwood G. A. (1988) Poor maternal weight gain bctween 28 and 32 wccks gestation may predict small-for-gestational agc infants. Br J Obstet Gynaecol95,884-887. Lukc B., Dickinson C. & Petrie R. H. (1981) Intrauterine growth correlations of maternal nutritional status and rate of gestational wcight gain. Eur J Obstef Gyn Reprod Bioll2, 113-121. Miller H. C . . Hassancin K. & Henslcigh P. A. (1976) Fctal growth rctardation in rclation to maternal smoking and weight gain during pregnancy. A m J Obstet Gynecol125,55-60. Ministry of Health. (1930) Memorandum on Antenatal Clinics: Their Conduct and Scope (Grcat Britain) MOH Appendix F. Neilson J., Munjanja S. P. & Whitfield C. R. (1984) Screening for small for dates fetuses. A controlled trial. Br Med J 289, 1179-1182. Nelson T. R . (1955) A clinical study of pre-cclampsia. J Obstet Gynaecol Br Emp 62,48-57. Niswandcr K. R . , Singcr J., Westphal M. & Weiss W. (1969)Weight gain during pregnancy and prcprcgnancy weight. Obstet Gynecol33, 482-491. Nyirjcsy I., Lonergan W. M. & Kane J. J. (1968) Clinical significancc of total wcight gain in prcgnancy Obstet Gynecol32, 391-3%. Ounsted M. & Scott A. (1984) Associations between

194

M. G. Dawes & J. G. Grudzinskas

maternal weight, height, weight for height, weight gain and birth weight. In Maternal Nutrition in Pregnancy. Eating for 2. (Dobbing J . ed). Acadcmic Press, London, pp. 113-125. Peckham C. H. & Christiansen R. E. (1971) The relationship between prepregnancy weight and certain obstetric factors. A m J Obstet Gynecol 111, 1-7. Rosso P. (1985) A ncw chart to monitor weight gain during pregnancy. Am J Clin Nun 41,644-652. Rosso P. & Luke B. (1978)The influence of maternal weight gain on the incidence of fetal growth retardation. Am J Clin Nutr 31, 696. Scott J . A. & Benjamin 9. (1948) Weight changes in pregnancy. Lancet i, 550-551. Simpson J. W., Lawless R. W. &Mitchell A. C. (1975) Responsibility of the obstetrician to the fetus. I1 Influence of prcpregnancy weight and pregnancy weight gain on thc birth wcight. Obstet Gynecol45, 481-487.

Singer J. E., Westphall M. & Niswander K. R. (1y68) Relationship of weight gain during pregnancy to birth weight and infant growth and devclopment in the first year of life. Obstet Gynecol31,417-422. Thomson A. M. & Billewicz W . 2. (1957) Clinical significance of wcight trends during pregnancy. Br Med J i, 243-247. Wennergren M., Karlsson K. & Olsson T.(1982) A scoring systcm for antenatal detection of fetal growth. Br J Obstet Gynaecol89,520-524. Yudkin P.L., Aboualfa M., Eyre J. A,, Redman C. W. G. & Wilkinson A. R. (1987) New birthweight and hcad circumfercnce centilcs for gestational ages 24 to 42 weeks. Early Hum Devel15,45-52.

Received 22 August 1989 Resubmitted 11 April 1990 Accepted 30 August 1990