Maternal psychological stress and distress as predictors of low birth ...

European Journal of Clinical Nutrition (2003) 57, 266–272 ß 2003 Nature Publishing Group All rights reserved 0954–3007/03 $25.00 www.nature.com/ejcn

ORIGINAL COMMUNICATION Maternal psychological stress and distress as predictors of low birth weight, prematurity and intrauterine growth retardation PHC Rondo´1*, RF Ferreira1, F Nogueira2, MCN Ribeiro1, H Lobert2 and R Artes2 1 Department of Nutrition, School of Public Health, University of Saõ Paulo, Saõ Paulo, Brazil; and 2Institute of Mathematics and ˜ o Paulo, Sa ˜ o Paulo, Brazil Statistics, University of Sa

Objectives: To evaluate: (1) the associations between maternal psychological stress, distress and low birth weight (LBW), prematurity and intrauterine growth retardation (IUGR); (2) the interactions between maternal stress, distress and smoking, alcohol and coffee intake; (3) the prevalences of stress and distress in pregnancy. Design: Longitudinal cohort study. Setting: Jundiaı´ city, Saõ Paulo state, Brazil. Subjects: A total of 865 pregnant women who attended antenatal care between September 1997 and August 2000. Methods: Measures of stress and distress were obtained, by interview, three times in pregnancy: at a gestational age (GA) lower than 16 weeks, from 20 to 26 weeks and from 30 to 36 weeks. Stress was investigated by the perceived stress scale, PSS, and distress by both the general health questionnaire, GHQ, and the State Trait Anxiety inventories, STAI. The outcomes were: LBW (birth weight < 2500 g), prematurity (gestational age (GA) at birth < 37 weeks) and IUGR (birth weight for GA 10th percentile of William’s curve). The associations between the outcomes and the psychological measures were assessed in multiple logistic regression models. Results: Maternal distress was associated with LBW (RR ¼ 1.97, P ¼ 0.019) and prematurity (RR ¼ 2.32, P ¼ 0.015), respectively. There was an interaction between distress and smoking in the second interview (P ¼ 0.05). The prevalences of stress and distress in the different interviews of pregnancy varied from 22.1 to 52.9%. Conclusions: The present study has confirmed that distress is associated with both birthweight and GA. Further research is needed to evaluate the effectiveness of psychological interventions that can improve maternal and foetal well-being. Sponsorship: Fundaçaõ de Amparo a` Pesquisa do Estado de Saõ Paulo - FAPESP (grant no. 1998=00321-0). European Journal of Clinical Nutrition (2003) 57, 266 – 272. doi:10.1038=sj.ejcn.1601526 Keywords: stress; distress; pregnancy; low birth weight; prematurity; intrauterine growth retardation

Introduction *Correspondence: PHC Rondo´, Department of Nutrition, School of Public õ Paulo, Avenida Dr Arnaldo 715, Sa õ Paulo, CEPHealth, University of Sa 01246-904, Brazil. E-mail: [email protected] Guarantor: PHC Rondo´. Contributors: PHCR designed the study protocol, secured funding, facilitated data collection, participated in the statistical analysis, interpreted data and did the main writing of the paper. RFF assisted in locating the study areas, coordinated all the data collection and participated in the interpretation of data. FN, HL and RA performed statistical data analysis, participated in the interpretation of the results and in discussion. MCNR participated in the interpretation of data and writing of the paper. Received 7 November 2001; revised 24 April 2002; accepted 7 May 2002

Low birth weight (LBW), prematurity and intrauterine growth retardation (IUGR) remain the leading causes of perinatal morbidity, mortality, neurodevelopmental impairments and disabilities among newborn babies (WHO, 1995). The belief that a mother’s psychological status may affect her foetus has been common in every society, since ancient times. Animal studies have demonstrated that exposing animals to stressors during pregnancy is associated with lower birth or foetal weights (Calhoun, 1962; Geber, 1966; Smith et al, 1975; Barlow et al, 1979; Beckhardt & Ward, 1983; Anderson et al, 1985; Pinto & Shetty, 1995). However, the impact of psychological factors in the human offspring has been examined in detail, relatively more recently. One

Maternal stress and distress PHC Rondo´ et al

reason for the renewed interest in these factors is the lack of progress in reducing poor pregnancy outcomes, especially prematurity, which is still high not only in developing countries, but also in developed countries (Horta et al, 1996; Buekens & Klebanoff, 2001; Branun & Schoendorf, 2002). A direct relationship between maternal psychological stress=distress and LBW, prematurity and IUGR may be related to the release of catecholamines, which results in placental hypoperfusion and consequent restriction of oxygen and nutrients to the foetus, leading to foetal growth impairment and=or precipitation of preterm delivery (Omer, 1986; Copper et al, 1996). Higher amounts of a stressrelated corticosteroid or opiate increasing the vulnerability to infectious diseases (particularly chorioamnionitis), a higher degree of neuromuscular reactivity and increases in the secretion of oxytocin are other important direct mediators (Omer, 1986). The relationship may be indirect; stressed or distressed women are more likely to smoke cigarettes or use substances such as alcohol and caffeine (McAnarney & Stevens-Simons, 1990). Studies on the association between maternal stress= distress and birth weight have yielded mixed results (Henriksen et al, 1994; Jacobsen et al, 1997; Sable & Wilkinson, 2000). The first reason may be related to the terminology of stress and distress, which is not yet standardized. According to Hoffman and Hatch (1996), stress describes the recognition that an insult has occurred, and distress describes negative emotional states that may result from the perception of stress. A second reason is that most of the studies assessing stress, distress and LBW, prematurity and IUGR are retrospective, introducing a bias in the direction of the hypothesis. Mothers who have small babies might recognize their pregnancies to be more stressful in retrospect than they might have if their babies were of normal size (Istvan, 1986). A third reason is that only few studies have controlled the association for confounding variables such as parity, gestational age (GA), socio-economic status, age, morbidity and anthropometry of the mother. The findings in most studies, however, suggest that the contribution of maternal stress to LBW, either directly or mediated through health-risk behaviours, such as cigarette smoking, alcohol intake (Pagel et al, 1990; Pritchard & Teo, 1994; Copper et al, 1996; Nordentoft et al, 1996; Jacobsen et al, 1997; Paarlberg et al, 1999), and probably coffee intake, should be considered seriously (Aarst & Vingerhoets, 1993; Hedegaard et al, 1996). Therefore, we decided to carry out a large prospective study, evaluating: (1) the associations between maternal psychological stress, distress and LBW, prematurity, and IUGR; (2) the interactions between maternal stress, distress and smoking, alcohol and coffee intake; (3) the prevalences of stress and distress in pregnancy. Through a better knowledge of the intricate relationship between maternal stress, distress and foetal growth, we will be able to carry out prenatal programmes to modify the mother’s psychological status to the benefit of her foetus.

267

Subjects and methods From an original sample of 1182 women who attended antenatal care between September 1997 and August 2000 in 12 health units and five hospitals in Jundiaı´, Southeast Brazil, 865 were followed in a cohort study before the 16th week of pregnancy to the birth of their babies. The losses were due to multiple reasons such as failure to attend one of the scheduled appointments at the prenatal services (85), migration to another geographical area (29), miscarriage (57), GA higher than 16 weeks at the first interview (28), refusal to answer one of the three questionnaires (28), lack of reliable data on GA (71) and gemelarity and negative pregnancy tests (19). All participants were insured by the National Health Service (SUS) that assists low income families. Women with chronic infectious diseases, metabolic diseases, cardiopathy, mental diseases, hypertension=preeclampsia=eclampsia, vaginal bleeding and multiple deliveries were not included in the study. Psychological measures (stress and distress) and a range of risk factors for LBW, prematurity and IUGR were obtained by four psychologists who interviewed the women three times before delivery: at a GA lower than 16 weeks, from 20 to 26 weeks and from 30 to 36 weeks. Considering that a large percentage of women from this region do not attend antenatal care in the first trimester of pregnancy, we decided to involve in the first interview women who were already in the second trimester. Maternal stress was assessed by the 14-item Perceived Stress Scale, PSS (Cohen et al, 1983), which determines the degree to which situations during the last month had been appraised as stressful. The PSS was designed for use with community samples. The items are easy to understand and the response alternatives are simple to grasp. Moreover, the questions are quite general in nature and hence relatively free of content specific to any sub-population group. It is rated on a five-point scale ranging from ‘never’ to ‘very often’. Maternal psychological distress was assessed by the 12item version of the General Health Questionnaire, GHQ (Goldberg, 1972), a screening instrument designed for use in general population surveys. The scores indicate the severity of psychological disturbance on a continuum. The content and concurrent validity of this questionnaire have been tested in a Brazilian population (Mari & Williams, 1985). Pregnant women were classified into two groups, according to their GHQ scores (low, 0 – 3; high, 4 þ ). Anxiety, in particular, was assessed using the State – Trait Anxiety Inventories, STAI (Spielberger et al, 1970). The STAI is a wellstandardized, 40 item, self-report instrument designed to measure both state and trait anxiety. For state anxiety, subjects are asked how they feel at the time of being questioned, and for trait anxiety, subjects are asked how they feel generally. A cut-off point of scores greater than 40 was selected for both state anxiety and trait anxiety. Possible risk factors for LBW, prematurity and IUGR were assessed by a standardised questionnaire seeking

European Journal of Clinical Nutrition


268 information on women’s medical and obstetric history (parity, prior history of LBW), socio-economic and demographic characteristics (age, marital status, education, per capita income), toxic exposure (smoking, alcohol intake and coffee intake), and anthropometry (pre-pregnancy weight, height, body mass index-BMI). The anthropometric measurements were determined according to Cameron’s (1984) and Jelliffe’s and Jelliffe’s (1989) recommendations. Data obtained by interview were checked against medical records and antenatal care cards, especially for obstetric history. Smoking was estimated by the mean number of cigarettes smoked per day by the pregnant women, in the three interviews. Consumption of alcohol was assessed by the mean number of drinks of alcoholic beverages ingested per week by the pregnant women, in the three interviews (a drink is 12 ounces of beer, 4 ounces of wine or 112 ounces of liquor or ‘pinga’ (typical Brazilian spirit); Faden et al, 1997). Consumption of coffee was assessed by the mean number of cups of coffee ingested per day by the pregnant women, in the three interviews. A sample of 20 women were interviewed four times by all the psychologists, in one week intervals, and all the results were within 5% in all repeated interviews. Birth outcome was described by three variables: LBW (birth weight < 2500 g), prematurity (GA at birth < 37 weeks); IUGR (birth weight for GA 10th percentile of Williams’ curve; Williams et al, 1982). GA was determined by a combination of ultrasonography performed up to the 20th week of gestation, the Capurro method (determined between 12 and 48 h of birth; Capurro et al, 1978), and information on the date of the last menstrual period (LMP). When there was less than or a week discrepancy between at least two of the GA determinations, assessed by the three different methods, one of them was chosen, giving preference to the order of the methods cited above. Data on Apgar score, anthropometry and physical examination of the newborn babies were collected as a routine, from the medical records. For calculation of the sample size for this cohort study, data on prevalences of LBW (9.8%), prematurity (7.5%) and IUGR (17.5%) were used (Horta et al, 1996). The parameters adopted were: 80% power, 0.05 significance and 4.7 relative risk (RR) for LBW, 2.86 RR for prematurity and 1.84 RR for IUGR, considering minimum frequencies of 5% of LBW, 3% of prematurity and 9% of IUGR for the pregnant women not exposed to stress. The relative risks used for the calculations were obtained from a research on stress in pregnancy (Pritchard & Teo, 1994). According to the parameters cited above, considering a loss of 15% of women, it would be necessary to include a minimum of 830 pregnant women (Lwanga & Lemeshow, 1991). Range and consistency checks as well as visual comparison of the record print-outs with the questionnaires were conducted when entering data into EPI-INFO (USD Incorporated, Stone Mountain, GA, USA). European Journal of Clinical Nutrition

Statistics were calculated with EPI-INFO and STATA (STATA Corporation, College Station TX, USA). The associations between the outcomes (LBW, prematurity and IUGR) and the psychological measures were controlled for confounding factors (possible risk factors), and assessed separately in three multiple logistic regression models (Kleinbaum, 1996). The factors that were found to be statistically significant in the univariate analysis and the ones considered as important predictors of the outcomes (Kramer, 1987) were included in the models. The scores of the STAI and GHQ were introduced into the models as dichotomized variables while the PSS scores were divided in quartiles, considering that there are no specific cut-off points for the PSS scores. Results were expressed in relative risks (RR) and confidence intervals (CI), considering a probability value (P) < 0.05 as significant. Informed written consent was obtained from all subjects and the protocol was approved by the Ethical Committees of the School of Public Health, University of Saõ Paulo and the Health Secretariat of Jundiaı´, SP.

Results The characteristics of the 865 pregnant women included in the study are shown in Table 1. One-quarter of the women were teenagers (age 19 y; WHO, 1986) and more than half had from 4 to 8 y of formal education (57%) and were officially married (51.8%). The majority of the women had a low per capita income, considering that only 21.3% received more than 3 Brazilian minimum wages (73 US$) a month. Approximately 44% of them were primiparous and 2.5% had a prior history of LBW. According to their nutritional status, 23.8% had a low pre-pregnancy weight ( 50 kg), 10.9% were short (height 1.50 m) and 14.2% had a BMI < 20. The percentage of mothers who smoked cigarettes in the first trimester of pregnancy (14.2%) was higher than in the second (11%) and third (10.6%) trimesters. The mean number of cigarettes smoked per day in pregnancy was approximately seven. However, the majority of the women (58.5%) smoked less than six cigarettes=day. Comparing our results with previous data from the same region (Rondo´ et al, 1996; Paiva, 2000), smoking in pregnancy has decreased, probably as a consequence of a smoking prevention programme carried out by the Secretary of Health at antenatal care services. A large number of women ingested alcohol in pregnancy (21.8, 18.4 and 12.7% in the three trimesters respectively), but the mean number of drinks ingested was low, varying between 2.46 and 3.47 drinks=month. Only 10% of the women had a mean ingestion of alcohol higher than 12 drinks=month. Coffee was ingested by most of the women (72.4, 74.2 and 71% in the three trimesters, respectively), although the mean number of glasses ingested by these women was not high, varying between 0.66 and 0.68 glasses=day (1 glass ¼ 120 ml). The mean ( s.d.) birth weight of the babies was 3201 483 g (range: 1180 – 4700 g); the mean ( s.d.) GA


269 Table 1

Characteristics of the 865 pregnant women

Characteristics

Table 1 cont. n

Age (y) 19 19 – 24 24 – 30 30 – 34 > 34 Education (y) 8 Marital status Single (without partner) Single (with partner) Married Separated=divorced or widow Per capita income (Brazilian minimum wagea) 0–1 1–2 2–3 >3 Parity 0 1–4 5 Prior history of low birth weight (LBW) Yes No Pre-pregnancy weight (kg) 50 50 – 60 60 – 70 > 70 Height (m) 1.50 1.51 – 1.56 1.57

%

n

%

123 479 207 56

14.2 55.4 23.9 6.5

123 742

14.2 85.8

95 770

11.0 89.0

92 773

10.6 89.4

189 676

21.8 78.2

159 706

18.4 81.6

110 755

12.7 87.3

626 239

72.4 27.6

642 223

74.2 25.8

614 251

71.0 29.0

Characteristics 2 b

218 266 204 119 58

25.2 30.7 23.6 13.8 6.7

77 493 295

8.9 57.0 34.1

163 233 448 21

18.9 26.9 51.8 2.4

208 300 173 184

24.0 34.7 20.0 21.3

380 465 20

43.9 53.8 2.3

22 843

2.5 97.5

206 369 202 88

23.8 42.7 23.3 10.2

94 248 523

10.9 28.7 60.4

Body mass index — BMI (weight=heigh ) 20 20 – 25 25 – 30 > 30 Smoking at first interview Yes No Smoking at second interview Yes No Smoking at third interview Yes No Alcohol intake at first interview Yes No Alcohol intake at second interview Yes No Alcohol intake at third interview Yes No Coffee intake at first interview Yes No Coffee intake at second interview Yes No Coffee intake at third interview Yes No a

Minimum wage ¼ 73 US dollar. Based on the pre-pregnancy weight.

b

Table 1 continued.

was 39.1 1.4 weeks (range: 32.2 – 42.6 weeks). Approximately 6.5% of the babies were classified as having LBW, 4.2% were delivered prematurely and 10.8% were intrauterine growth retarded (IUGR). Table 2 presents the results of logistic regression analyses of the maternal factors that predict LBW, prematurity and JUGR (controlling for the confounding factors specified in

Table 1). According to these results, LBW was associated with a GHQ score > 3 in the second interview (RR ¼ 1.97; P ¼ 0.019), parity 5 (RR ¼ 5.88; P ¼ 0.002) and age 19 y (RR ¼ 2.31; P ¼ 0.005); prematurity was associated with a GHQ score > 3 in the third interview (RR ¼ 2.32; P ¼ 0.015), a prior history of LBW (RR ¼ 4.96; P ¼ 0.017) and age 19 y (RR ¼ 2.48; P ¼ 0.011); IUGR was associated

Table 2 Results of the logistic regressions of low birth weight (LBW), prematurity and intrauterine growth retardation (IUGR) on predictor variables (n ¼ 865) Outcomes LBW

Prematurity

IUGR

b

Variables

Regression coefficient

Standard error

GHQa > 3 (second interview) Parity 5 Age 19 y GHQ > 3 (third interview) Prior history of LBW Age 19 y Parity 5 Primiparity Height 1.50 m

0.67 1.77 0.83 0.84 1.60 0.91 1.58 0.95 0.68

0.29 0.56 0.29 0.35 0.67 0.36 0.55 0.23 0.32

Relative risk (95% CI) 1.97 5.88 2.31 2.32 4.96 2.48 4.88 2.60 1.97

(1.12 – 3.47) (1.97 – 17.60) (1.30 – 4.12) (1.18 – 4.60) (1.33 – 18.46) (1.23 – 5.00) (1.65 – 14.48) (1.64 – 4.14) (1.04 – 3.73)

P-value 0.019 0.002 0.005 0.015 0.017 0.011 0.004 < 0.001 0.037

GHQ, general health questionnaire.



270 with primiparity (RR ¼ 2.61; P < 0.001), parity 5 (RR ¼ 4.88; P ¼ 0.004) and height 1.50 m (RR ¼ 1.97; P ¼ 0.037). It is important to emphasize that it was easier for mothers to remember the weight than the GA of their children. This is the explanation for having a prior history of LBW as a predictor of prematurity in the model, and not a prior history of prematurity. Figure 1 shows that mean scores of the GHQ in the three interviews were significantly higher (P < 0.001) for women who smoked cigarettes in pregnancy than for non-smokers, showing an association between GHQ scores and smoking. There was also an interaction between GHQ scores and smoking in the second interview (P ¼ 0.05). However, there were no associations or interactions between GHQ scores, alcohol and coffee intake. There were 22.1 – 24.6% pregnant women in the highest quartile of the PSS (mean score ¼ 30.21, s.d. ¼ 3.59 (min ¼ 28; max ¼ 53)) in the three trimesters of pregnancy. Comparing the scores of these women with the mean scores of the whole population (22.5, s.d. ¼ 6.4), and the mean scores of the populations studied by Cohen et al (1983), we consider that these women had stress. The prevalences of maternal distress, assessed by the GHQ and STAI, varied between 25.9 and 52.9% in the three trimesters of pregnancy.

Discussion The present study has confirmed that psychological factors are associated with both birth weight and GA. It was found that LBW and preterm delivery are associated with distress in the second interview (20 – 26 weeks gestation) and in the third interview (30 – 36 weeks gestation), respectively, even after controlling for other risk factors. There are only few prospective studies which have focused on the association between psychological factors and preterm labour, but most of them have reported a positive association, showing that preterm deliveries have often been preceded by stressful situations (Austin & Leader,

Figure 1 Comparison of the mean scores of the GHQ in the three interviews according to smoking by pregnant women (n ¼ 865).


2000)(Austin & Leader, 2000). Hedegaard et al (1993) carried out a research evaluating the importance of psychological distress in pregnancy (estimated by the GHQ questionnaire) on preterm labour, and found a dose – response relationship between distress assessed in the 30th week of pregnancy and risk of preterm delivery. The results of this research are similar to ours, considering that the association of maternal distress with prematurity was observed only in the third interview, but not in the first and second interviews. Copper et al (1996) also reported that women with a GA between the 25th and 29th weeks of pregnancy, who perceived their lives as being stressful, are at increased risk for spontaneous preterm birth and LBW. Wadhwa et al (1993) found that, independent of biomedical risk, each unit increase of prenatal life event stress (from a possible sample range of 14.7 units) was associated with a 55.03 g decrease in infant birth weight and with a significant increase in the likelihood of LBW (odds ratio ¼ 1.32); each unit increase of prenatal pregnancy anxiety (from a possible sample range of 5 units) was associated with a 3 day decrease in GA at birth. Two reviews of the literature on stress and birth outcomes have concluded that life-event stress is most consistently correlated with prenatal events, particularly with preterm delivery or GA, rather than with complications during delivery (Levin & DeFrank, 1988; Lederman, 1995). According to Paarlberg et al (1995) there is a tendency to a higher muscular tone in stressing conditions, which can probably potentiate early contractions, triggering preterm labour, especially in the later phases of pregnancy. A cohort study carried out by Teixeira et al (1999) found an association between scores of the STAI (Spielberger et al, 1970) in pregnancy and higher mean uterine artery resistance index (r ¼ 0.31, P < 0.002; r ¼ 0.28, P < 0.005), respectively. This study suggests a mechanism by which the psychological state of the mother may affect foetal development, and may explain epidemiological associations between maternal anxiety and LBW. A large study from Denmark, evaluated the influence of psychosocial factors on intrauterine growth retardation and preterm delivery. Psychosocial stressors and duration of schooling appeared to influence preterm delivery (Nordentoft et al, 1996). A study involving women with depressive symptoms late in pregnancy showed that these women had a 3.97 risk of delivering a LBW, a 3.39 risk for having a preterm delivery, and a 3.02 risk for IUGR (Steer et al, 1992). Exposure to stressful conditions might also influence GA or birth weight by promoting specific behaviours in human beings such as smoking, alcohol and coffee intake, which may be independently associated with poor pregnancy outcomes. Studies of women’s smoking habits have suggested that women smoke to relieve stress, unhappiness and depression (Floyd et al, 1993). A study involving low-income pregnant adolescents reported that they smoked to counter anxiety (Lawson et al, 1994). Grimstadt et al (1999) referred that mothers with a


271 history of physical abuse had higher anxiety scores and more often smoked in pregnancy. Pritchard and Teo (1994) studied the relationship between persistent stressful aspects of daily life and birth outcomes in 393 women from Scotland. He found that psychosocial difficulty in the household role represents a risk factor for preterm birth and LBW, and suggested that smokers who experience psychosocial problems may be at particular risk. A review by Hoffman and Hatch (1996), covering the last 10 y of research on stress, social support and pregnancy outcome, concluded that acute life stressors have direct effects on birth outcome, whereas strong partner or family support as a promoter of foetal growth can with almost equal certainty be ruled in. In this study, IUGR was not associated with stress=distress, but only with few risk factors — primiparity, a parity 5 and maternal height 1.50 m. One of the possible explanations is the high cut-off points, corresponding to the 10th percentile of the Williams’ curve (Williams et al, 1982). An epidemiological study evaluating risk factors for IUGR (not including stress) in the same region of the country, but using another curve for classification of IUGR babies, found associations with several factors (Ronao´ et al, 1997). However, according to some studies from the literature, the impact of psychosocial factors on IUGR seems to be not so important as their impact on LBW and prematurity. Aarts & Vingerhoets (1993) in Netherlands, investigating the role of maternal chronic stress and well-being in the three different trimesters of pregnancy on foetal growth, reported that smoking, family income and maternal weight were the only predictors of this outcome. In a large study from Norway (Jacobsen et al, 1997), parental education, maternal body proportion and lifestyle were the risk factors for IUGR. Stress, distress and physical strain did not influence birth outcome. A large study carried out by Brooke et al (1989) investigating the effects of smoking, alcohol and caffeine consumption, socio-economic factors and psychosocial stress on birth weight, did not find any effect of social and psychological factors on foetal growth. According to the results of a large study carried out by Edwards et al (1994) women with a positive self attitude and higher self esteem were more likely to deliver infants at term. The number of persons in the women’s social support network was also directly correlated with her infant’s GA. Brownell and Shumaker (1985) emphasize the importance of preventing or alleviating sources of stress in pregnancy, whenever possible, even considering the difficulty of reducing life stress among women from low income groups. The implication of these findings is that a keener assessment of stress=distress experienced by women during pregnancy should be made, considering the impact of psychological problems in this population. Further research is also needed to evaluate the effectiveness of psychological interventions and the possible value of social support systems which will improve maternal and foetal general well-being.

Acknowledgements We thank the pregnant women whose participation made this study possible, and the field workers Veranilce F Oliveira, Ma´ rcia R Souza, Sandra Vergı´lio, Rosemary F Ferreira, Janaina de Carvalho, Ka´ tia K Valverde and Vale´ ria Sttopa for their assistance. Scholarships from Conselho Nacional de Desenvolvimento Cientı´fico e Tecnolo´ gico (CNPq) supported three of the field workers. The authors gratefully acknowledge Dr Ronaldo Moise´ s, Ms Vera M Saccheto, Dr Mauri Senise and Dr Gustavo Albuquerque from the Hospital Maternidade de Jundiaı´; Dr Maria Emilia Ganito, Dr Alessandro D’Angieri, Dr Paulo Pontes, Dr Nelson O Maia, Dr Maria Emı´lia R Miglioli and Dr Renato Tardelli from Hospital de Caridade Saõ Vicente de Paulo; Dr Silvia Borgonovi, Dr Silvia Vilas Boas de Oliveira and Dr Jose´ Paulo May from Hospital Dr Paulo Sacramento; Dr Carim Chahine Yossef from Hospital Santa Rita de Ca´ ssia; Dr Jonas Zulpo, Dr Komei Samejima and Dr Cecı´lia M Cordeiro from Casa de Sau´ de Dr Domingos Anasta´ cio; Dr Rui Otanara and Dr Moacyr RF Paula from Centro Me´ dico Pitangueiras; Dr Rubens A Dutra Alves and Dr Ju´ lio C Lamarca Health Secretaries of Jundiaı´, for allowing us to carry out our study in the hospitals and health centres of this city. Dr Laura C Rodrigues from the Department of Infectious and Tropical Diseases, London School of Hygiene and Tropical Medicine and Professor Andrew Tomkins from the Centre for International Child Health, University of London, are acknowledged for fruitful discussions concerning the analysis and results of the study.

References Aarst MCG & Vingerhoets AJJM (1993): Psychosocial factors and intrauterine fetal growth — a prospective study. J. Psychosom. Obstet. Gynecol. 14, 249 – 258. Anderson DK, Rhees RW & Fleming DE (1985): Effects of prenatal stress on differentiation of the sexually dimorphic nucleus of the preoptic area (SDN-POA) of the rat brain. Brain Res. 332, 113 – 118. Austin MP & Leader L (2000): Maternal stress and obstetric and infant outcomes: epidemiological findings and neuroendocrine mechanisms. Aust. NZ J. Obstet. Gynaecol. 40, 331 – 337. Barlow SM, Knight AF & Sullivan FM (1979): Prevention by diazepam of adverse effects of maternal restraint stress on postnatal development and learning in the rat. Teratology 19, 105 – 110. Beckhardt S & Ward IL (1983): Reproductive functioning in the prenatally stressed female rat. Dev. Psychobiol. 16, 111 – 118. Branun AM & Schoendorf KC (2002): Changing patterns of low birth weight in the United States. Paediatr. Perinat. Epidemiol. 16, 8 – 15. Brooke OG, Anderson HR, Bland JM, Peacock JL & Steward CM (1989): Effects on birth weight of smoking, alcohol, caffeine, socioeconomic factors, and psychosocial stress. Br. Med. J. 298, 795 – 801. Brownell A & Shumaker SA (1985): Where do we go from here: the policy implications of social support. J. Soc. Issues 41, 111 – 121. Buekens P & Klebanoff M (2001): Preterm birth research: from disillusion to the search for new mechanisms. Paediatr. Perinat. Epidemiol. 15(Suppl 2), 159 – 161. Calhoun JB (1962): Population density and social pathology. Sci. Am. 206, 139 – 148. Cameron N (1984): The Measurement of Human Growth, 1st edn. London: Croom Helm.



272

Capurro H, Konishezky S, Fonseca D & Caldeyro-Barcia (1978): A simplified method for diagnosis of gestational age in the newborn infant. J. Pediatr. 93, 120 – 122. Cohen S, Kamarck T & Mermelstein R (1983): A global measure of perceived stress. J. Health. Soc. Behav. 24, 385 – 396. Copper RL, Goldenberg RL, Das A, Elder N, Swain M, Norman G, Ramsey R, Cotroneo P, Collins BA, Johnson F et al (1996): The preterm prediction study: maternal stress is associated with spontaneous preterm birth at less than 35 weeks’ gestation. Am. J. Obstet. Gynecol. 175, 1286 – 1292. Edwards CH, Cole OJ, Oyemade UJ, Knight EM, Johnson AA, Westney OE, Laryea H, West W, Jones S & Westney LS (1994): Maternal stress and pregnancy outcomes in a prenatal clinic population. J. Nutr. 124(Suppl 6), S1006 – S1021. Faden VB, Graubard BI & Dufour M (1997): The relationship of drinking and birth outcome in a US national sample of expectant mothers. Paediatr. Perinat. Epidemiol. 11, 167 – 180. Floyd RL, Rimer BK, Giovino GA, Mullen PD & Sullivan SE (1993): A review of smoking in pregnancy: effects on pregnancy outcomes and cessation efforts. Ann. Rev. Public Health 14, 379 – 411. Geber WF (1966): Developmental effects of chronic maternal audiovisual stress on the rat fetus. J. Embriol. Exp. Morphol. 16, 1 – 16. Goldberg DP (1972): The Detection of Psychiatric Illness by Questionnaire. London: Oxford University Press. Grimstadt H, Schei B, Backe B & Jacobsen G (1999): Anxiety, physical abuse, and low birth weight. Scand. J. Public Health 27, 296 – 300. Hedegaard M, Henriksen TB, Sabroe S & Secher NJ (1993): Psychological distress in pregnancy and preterm delivery. Br. Med. J. 307, 234 – 238. Hedegaard M, Henriksen TB, Scher NJ, Hatch MC & Sabroe S (1996): Do stressful life events affect duration of gestation and risk of preterm delivery? Epidemiology 7, 339 – 345. Henriksen TB, Hedegaard M & Secher NJ (1994): The relation between psychosocial job strain, and preterm delivery and low birthweight for gestational age. Int. J. Epidemiol. 23, 764 – 774. Hoffman S & Hatch C (1996): Stress, social support and pregnancy outcome: a reassessment based on recent research. Paediatr. Perinat. Epidemiol. 10, 380 – 405. Horta BL, Barros FC, Halpern R & Victora C (1996): Baixo peso ao nascer em duas coortes de base populacional no Sul do Brasil. Cad Sau´de Pu´bl. 12, 27 – 31. Istvan J (1986): Stress, anxiety, and birth outcomes: a critical review of the evidence. Psychol. Bull. 100, 331 – 348. Jacobsen G, Schei B & Hoffman HJ (1997): Psychosocial factors and small-for-gestational-age infants among parous Scandinavian women. Acta Obstet. Gynecol. Scand. 76(Suppl 165), S14 – S18. Jelliffe DB & Jelliffe EFP (1989): Community Nutritional Assessment, with Special Reference to Less Technically Developed Countries, 2nd edn. London: Oxford University Press. Kleinbaum DG (1996): Logistic Regression — A Self-Learning Text. New York: Springer. Kramer MS (1987): Determinants of low birth weight: methodological assessment and meta-analysis. Bull. WHO 65, 663 – 737. Lawson EJ (1994): The role of smoking in the lives of low-income pregnant adolescents: a field study. Adolescence 29, 61 – 79. Lederman RP (1995): Relationship of anxiety, stress, and psychosocial development to reproductive health. Behav. Med. 21, 101 – 112. Levin JS & DeFrank RS (1988): Maternal stress and pregnancy outcomes: a review of the psychosocial literature. J. Psychosom. Obstet. Gynecol. 9, 3 – 16. Lwanga SK & Lemeshow S (1991): Sample Size Determination in Health Studies — a Practical Manual. Geneva: WHO. Mari JJ & Williams P (1985): A comparison of the validity of two psychiatric screening questionnaires (GHQ-12 and SRQ-20) in Brazil, using relative operating characteristic (ROC) analysis. Psychol. Med. 15, 651 – 659.


McAnarney ER & Stevens-Simons C (1990): Maternal psychological stress=depression and low birth weight. Am. J. Dis. Child. 144, 789 – 792. McCormick MC, Gortmaker SL & Sobol AM (1990). Very low birth weight children: behavior problems and school difficulty in a national sample. J. Pediatr. 117, 687 – 693. Nordentoft M, Lou HC, Hansen D, Nim J, Pryds O, Rubin P & Hemmingsen R (1996): Intrauterine growth retardation and premature delivery: the influence of maternal smoking and psychological factors. Am. J. Public Health 86, 347 – 354. Omer H (1986): Possible psychophysiologic mechanisms in premature labor. Psychosomatics 27, 580 – 584. Paarlberg KM, Vingerhoets AJJM, Passchier J, Dekker GA & Van Geijn HP (1995): Psychosocial factors and pregnancy outcome: a review with emphasis on methodological issues. J. Psychosom. Res. 39, 563 – 595. Paarlberg KM, Vingerhoets AJJM, Passchier J, Dekker GA, Heinen AGJJ & van Geijn HP (1999): Psychosocial predictors of low birthweight: a prospective study. Br. J. Obstet. Gynaecol. 106, 834 – 841. Pagel MD, Smilkstein G, Regen H & Montano D (1990): Psychosocial influences on new born outcomes: a controlled prospective study. Soc. Sci. Med. 30, 597 – 604. Paiva AA (2000): Ana´ lise da relaç a˜ o entre os nı´veis de ferro de parturientes e rece´m-nascidos a termo. MPH thesis. Saõ Paulo: University of Saõ Paulo. Pinto ML & Shetty PS (1995): Influence of exercise-induced maternal stress on fetal outcome in Wistar rats: inter-generational effects. Br. J. Nutr. 73, 645 – 653. Pritchard CW & Teo PYK (1994): Preterm birth, low birthweight and the stressfulness of the household role for pregnant women. Soc. Sci. Med. 38, 89 – 96. Rondo´ PHC, Rodrigues LC & Tomkins AM (1996): Coffee consumption and intrauterine growth retardation in Brazil. Eur. J. Clin. Nutr. 50, 705 – 709. Rondo´ PHC, Abbott R, Rodrigues LC & Tomkins AM (1997): The influence of maternal nutritional factors on intrauterine growth retardation. Pediatr. Perinat. Epidemiol. 11, 152 – 166. Sable MR & Wilkinson DS (2000): Impact of perceived stress, major life events and pregnancy attitudes on low birth weight. Fam. Plann. Perspect. 32, 288 – 294. Smith DJ, Joffe JM & Heseltine GFD (1975): Modification of prenatal stress effects in rats by adrenalectomy, dexamethasone, and chlorpromazine. Physiol. Behav. 15, 461 – 469. Spielberger CD, Gorsuch RI & Lushene RE (1970): STAI Manual for the State – Trait Anxiety Inventory. Palo Alto, CA: Consulting Psychologist Press. Steer RA, Scholl TO, Hediger MI & Fischer RL (1992): Self-reported depression and negative pregnancy outcomes. J. Clin. Epidemiol. 45, 1093 – 1099. Teixeira JMA, Nicholas MF & Glover V (1999): Association between maternal anxiety in pregnancy and increased uterine artery resistance index: cohort based study. Br. Med. J. 318, 153 – 157. Wadhwa PD, Sandman CA, Porto M, Dunkel-Schetter C & Garite TJ (1993): The association between prenatal stress and infant birth weight and gestational age at birth: a prospective investigation. Am. J. Obstet. Gynecol. 169, 858 – 865. WHO (1995): Physical status: the use and interpretation of anthropometry. Technical Report Series 854. Geneva: World Health Organization. WHO (1986): Young people’s health — a challenge for society. Report of a WHO Study Group on Young People and ‘Health for all by the Year 2000’. Technical Report Series 731. Geneva: World Health Organization. Williams RL, Creasy RK, Cunningham MD, Hawes WE, Norris FD & Tashiro M (1982): Fetal growth and perinatal viability in California. Obstet. Gynecol. 59, 624 – 632.