Randomised controlled trial of effect of fish-oil supplementation on pregnancy duration

Effects of fish oil supplementation in late pregnancy on blood pressure: a randomised controlled trial

OBJECTIVE

To study the effect of fish oil supplementation on blood pressure during the third trimester of pregnancy.

DESIGN

In the 30th week of pregnancy 533 healthy women were randomly assigned in a ratio 2:1:1 to receive fish oil (2.7 g/day n-3 fatty acids (Pikasol)), or a control regimen of either olive oil or no oil supplementation.

MAIN OUTCOME MEASURES

Blood pressure measured with an automatic device (Dinamap 1846 SX, Criticon) at baseline and in weeks 33, 37, 39 and subsequently weekly until delivery.

RESULTS

Mean blood pressure increased during the third trimester, and this was not influenced by group assignment. No significant effects on either systolic or diastolic blood pressure were seen in the fish oil group compared to the control groups. The proportions of women with a systolic blood pressure above 140 mmHg or a diastolic blood pressure above 90 mmHg were not significantly different in the fish oil group compared with the control groups, although the proportion of women with diastolic above 90 mmHg tended to be lower in the fish oil group compared with the olive oil group. The corresponding relative risk was RR = 0.48 (95% CI 0.22-1.06; P = 0.07).

CONCLUSION

2.7 g/day of marine n-3 fatty acids provided in the third trimester of normal pregnancy showed no effect on blood pressure.

Increased docosahexaenoic acid levels in human newborn infants by administration of sardines and fish oil during pregnancy

In rhesus monkeys, maternal n-3 fatty acid deficiency during pregnancy produces infant monkeys deficient in n-3 fatty acids at birth. These results stimulated current experiments to find out if n-3 fatty acids from fish in the diets of pregnant women would influence the concentration of docosahexaenoic acid (DHA, 22:6 n-3) in the newborn human infant. Fifteen healthy pregnant women were enrolled to receive a 9-wk dietary supplementation of n-3 fatty acids from the 26th to the 35th wk of pregnancy. Sixteen pregnant women were not supplemented and served as controls. n-3 Fatty acid supplementation consisted of sardines and additional fish oil, which provided a total of 2.6 g of n-3 fatty acids per day (d) for the 9-wk period of supplementation. This included 1.01 g DHA. The end point of this study was the blood concentrations of DHA in the newborn infant. DHA in maternal red blood cells increased from 4.6% of total fatty acids to 7.15% at the end of the supplement period and at the time of delivery decreased (as expected) to 5.97% of total fatty acids. Maternal plasma showed a similar change from 2.12 to 3.51% of total fatty acids and then decreased to 2.35%. Levels of DHA in plasma and red blood cells of unsupplemented mothers did not change during the same time period. Levels of DHA in blood of newborn infants differed greatly in infants born from n-3-supplemented mothers compared with control infants. In red blood cells, DHA was 7.92% of total fatty acids compared with 5.86% (control infants). Plasma values showed a similar difference: 5.05% vs. 3.47% (controls). In n-3-supplemented infants, DHA concentrations were 35.2% higher than in control infants in red blood cells and 45.5% higher in plasma. These data indicate the importance of maternal dietary n-3 fatty acids and, in particular, maternal dietary DHA in promoting higher concentrations of DHA in the blood of the newborn infant.

Essential fatty acid status in neonates after fish-oil supplementation during late pregnancy

Healthy pregnant women (n 23) were supplemented with fish-oil capsules (2.7 g n-3 polyunsaturated fatty acids/d) from the 30th week of gestation until delivery. Subjects in a control group were either supplemented with olive-oil capsules (4 g/d, n 6) or received no supplementation (n 10). Fatty acid compositions of the phospholipids isolated from umbilical plasma and umbilical arterial and venous vessel walls were determined. Fatty acid compositions of maternal venous plasma phospholipids were determined as well. Maternal plasma phospholipids of the fish-oil-supplemented group contained more n-3 fatty acids and less n-6 fatty acids. Moreover, the amounts of the essential fatty acid deficiency markers Mead acid (20:3n-9) and Osbond acid (22:5n-6) were significantly lower. The extra amount of n-3 fatty acids consumed by the mothers resulted in higher contents of n-3 fatty acids, and of docosahexaenoic acid (22:6n-3) in particular, in the phospholipids of umbilical plasma and vessel walls. It is, indeed, possible to interfere with the docosahexaenoic acid status at birth: children born to mothers supplemented with fish oil in the last trimester of pregnancy start with a better docosahexaenoic acid status at birth, which may be beneficial to neonatal neurodevelopment.

n-3 polyunsaturated fatty acids: update 1995

Epidemiologic and biochemical studies have suggested an anti-inflammatory effect of n-3 fatty acids. Beneficial therapeutic effects reported from small patient groups need to be confirmed in large-cohort controlled clinical trials. There is a growing number of clinical trials of n-3 fatty acid supplementation in disease. Clinical benefits have been moderate in patients with rheumatoid arthritis and with arterial hypertension. Clearly negative results have been reported during the past 2 years for patients with lupus nephritis and for patients with psoriasis or with atopic dermatitis. Such trials have now been completed. For patients with coronary artery disease following coronary angioplasty, earlier results of a large meta-analysis, could not be confirmed. For patients with IgA-nephropathy and for patients following kidney transplantation, a clear benefit was seen in patients receiving fish oil. These promising results are currently pursued in follow-up phase III clinical trials.

Comparison between the essential fatty acid status of preterm and full-term infants, measured in umbilical vessel walls

The essential fatty acid (EFA) composition of umbilical vessel walls is increasingly being studied as a longer-term reflection of the fetal EFA status. We evaluated the EFA content of umbilical artery and vein vessel walls in 43 preterm infants and compared it with that of 43 full-term cord vessels. In addition, relations among cord vessel wall fatty acid composition, gestational age (GA) at birth, and anthropometric parameters at birth (weight, head circumference, and length) were explored in the preterm infants. Generally, n-6 and n-3 EFA levels were lower, while levels of EFA deficiency markers were higher in preterm than in term cords, both in the walls of the draining arteries and the supplying vein. In preterm cords, significant correlations were observed between GA at birth and levels of n-6 and n-3 EFAs (positive) and EFA deficiency markers (negative). Birth weight showed significant (P < or = 0.01), positive correlations with n-6 and n-3 EFA levels in the cord artery walls of preterm infants, all after correction for GA at birth. In conclusion, substantial differences between the EFA profiles of preterm and full-term cord vessel walls indicate a lower biochemical EFA status of the preterm than of the term fetus. This lower preterm EFA status might be a reflection of a physiologically lower EFA demand for growth and development of the preterm fetus than of the term fetus.

Maternal essential fatty acid patterns during normal pregnancy and their relationship to the neonatal essential fatty acid status

Although essential fatty acids (EFA) and their longer chain, more unsaturated derivatives play a major role during pregnancy, hardly any information is available with respect to the course of the maternal EFA status during an uncomplicated pregnancy and its relationship to the neonatal EFA status. Therefore, a longitudinal study was started in which 110 pregnant women gave repeated blood samples from the 10th week of gestation until delivery. After birth a blood sample from the umbilical vein and a maternal venous blood sample were collected as well, and 6 months after delivery a final blood sample from the mother was taken. The absolute (mg/l) and relative (% total fatty acids) amounts of the fatty acids in plasma phospholipids were determined. The total amounts of fatty acids increased significantly during pregnancy. This pattern was similar for the individual fatty acids and fatty acid families. The relative amount of linoleic acid (18:2n-6) did not change during pregnancy, whereas the relative amount of arachidonic acid (20:4n-6) decreased. Despite maternal mobilization of docosahexaenoic acid (22:6n-3, DHA), suggested by a temporary increase in the DHA status until 18 weeks gestation, the DHA status steadily declined thereafter. This pattern was associated with a progressive increase in the DHA deficiency index in maternal blood throughout pregnancy and resulted in a sub-optimal neonatal DHA status. The overall maternal EFA status also declined steadily during pregnancy. Therefore, the question arises whether the mother, under the prevailing dietary conditions, is able to meet the high fetal requirement for EFA.

A clinico-pathological classification of perinatal deaths in the Faroe Islands

OBJECTIVE

To highlight the causes of perinatal mortality in the Faroe Islands where perimortality is high according to Nordic standards.

DESIGN

Two systems were employed to classify perinatal deaths on the basis of clinico-pathological findings, one focusing on obstetrical factors and the other on fetal-neonatal factors.

SETTING

Faroe Islands. Data from Iceland were used for comparison since the two communities have many similarities, including similar birthweight distributions.

SUBJECTS

Birth and death certificates and medical and midwife files were recovered for 98 of the 102 officially recorded perinatal deaths in the Faroes during 1977-1986. Icelandic data for 1976-1985 were available in the literature.

RESULTS

The perinatal mortality rate was 13.7 per 1000 births. Obstetric factors were classified as unexplained, congenital anomaly, antepartum haemorrhage, pre-eclampsia, mechanical causes, and maternal disorders and accounted for 43%, 18%, 15%, 11%, 6% and 6% of the 98 cases, respectively. Fetal-neonatal factors were classified as antepartum asphyxia, congenital anomaly, intrapartum asphyxia, hyaline membrane disease, pulmonary immaturity, and other causes, and these factors accounted for 43%, 18%, 15%, 9%, 5%, and 8%, respectively. The excess perinatal mortality rate of 4.1 cases per 1000 births in the Faroes, relative to Iceland, could mainly be attributed to an excess of 2.9 cases per 1000 births in the group classified as unexplained, as defined according to the obstetrics classification.

CONCLUSIONS

No well defined cause was particularly common in the Faroes or could account for the excess perinatal mortality rate in the Faroes relative to Iceland.

Erythrocyte levels compared with reported dietary intake of marine n-3 fatty acids in pregnant women

It is well established that marine n-3 fatty acids measured in erythrocyte phospholipids of non-pregnant subjects reflect the subjects' intake of these fatty acids. In 135 pregnant women in the 30th week of gestation we compared intake of marine n-3 fatty acids and energy, estimated by a combined dietary self-administered questionnaire and interview, with fatty acids measured in erythrocyte phospholipids. Daily intake (g/d) and nutrient density of marine n-3 fatty acids (mg/MJ) correlated with the n-3 fatty acid: arachidonic acid ratio (FA-ratio) with correlation coefficients of 0.48 and 0.54 respectively. In a linear regression model with three frequency questions about marine sandwiches, marine cooked meals and fish oil as explanatory variables, and the FA-ratio as dependent variable, the multiple correlation coefficient was 0.46. Conclusions from the study were (1) levels of erythrocyte fatty acids in pregnant women may be employed as a qualitative method to rank subjects according to intake of marine n-3 fatty acids; (2) with respect to the power to explain FA-ratio variability, three simple marine food frequency questions were comparable with intake of marine n-3 fatty acids assessed by an elaborate semiquantitative dietary method involving an interview.

Gestation length and birth weight in relation to intake of marine n-3 fatty acids

It has been hypothesized that marine n-3 fatty acids ingested during pregnancy prolong duration of pregnancy and increase fetal growth rate in humans. By a combined self-administered questionnaire and interview applied in the 30th week of gestation we assessed dietary intake of marine n-3 fatty acids and energy in a population-based sample of 965 pregnant Danish women; in a random 14% subsample we also measured marine n-3 fatty acids relative to arachidonic acid (FA-ratio) in erythrocytes. Mean intake of marine n-3 fatty acids was 0.25 (95% range 0-0.75) g/d. We could detect no association between n-3 fatty acid intake and FA-ratio on the one hand, and gestation length, birth weight and birth length on the other. The analyses were adjusted for maternal height, prepregnant weight, parity and smoking. The conclusion from the study was that within the intake range of this population, marine n-3 fatty acids ingested in the weeks prior to the 30th week of pregnancy seem not to be a predictor of gestation length or fetal growth rate.

The effects of 3g eicosapentaenoic acid daily on recurrence of intrauterine growth retardation and pregnancy induced hypertension

OBJECTIVE

To study the effects of addition of 3 g eicosapentaenoic acid daily to the diet, on recurrence rate of intrauterine growth retardation and pregnancy induced hypertension in a high risk population.

DESIGN

Prospective, double blind, randomised multicentre study. Eicosapentaenoic acid or placebo were given from 12 to 14 weeks of gestation onwards.

SETTING

University Hospital and regional hospitals in the north of the Netherlands.

SUBJECTS

Sixty-three women with a history of intrauterine growth retardation (birthweight < 10th centile) with or without pregnancy induced hypertension in the previous pregnancy.

MAIN OUTCOME MEASURES

Birthweight centiles and signs of pregnancy induced hypertension in current pregnancy.

RESULTS

One-third of the women developed pregnancy induced hypertension and one-third of the infants had a birthweight below the 10th centile. There were no differences between eicosapentaenoic acid and placebo group.

CONCLUSION

Addition of 3 g eicosapentaenoic acid daily does not prevent recurrence of intrauterine growth retardation or pregnancy induced hypertension in a high risk population.

A randomised double blind placebo controlled trial of fish oil in high risk pregnancy

OBJECTIVE

To determine whether n-3 fatty acid (EPA/DCHA) prophylaxis is beneficial in high risk pregnancies.

DESIGN

A randomised, double blind, placebo controlled trial.

SETTING

Antenatal clinic of St James's University Hospital, Leeds.

SUBJECTS

Two hundred and thirty-three pregnant women at high risk of developing proteinuric or nonproteinuric pregnancy induced hypertension or asymmetrical intrauterine growth retardation.

INTERVENTION

Active treatment was 2.7 g of MaxEpa daily (1.62 g of eicosapentaenoic acid and 1.08 g of docosahexaenoic acid). Placebo were matching air-filled capsules.

MAIN OUTCOME MEASURES

Occurrence of proteinuric, nonproteinuric pregnancy induced hypertension or birthweight < 3rd centile.

RESULTS

There was no difference in an intention to treat analysis between the placebo and active treatment groups for occurrence of proteinuric pregnancy induced hypertension (relative risk (RR) = 0.88; 95% CI 0.47-1.66), nonproteinuric pregnancy induced hypertension (RR = 0.89; 95% CI 0.48-1.64), birthweight < 3rd centile (RR = 0.89; 95% CI 0.48-1.64), or the duration of pregnancy (difference of mean durations = 0.1 days; 95% CI -4.8 to 4.9 days). Analyses stratified by use of tobacco, and analyses excluding known major protocol violators gave essentially identical results.

CONCLUSION

There is no evidence from this study for any useful effect of fish oil supplementation for women at high risk of adverse outcomes from a pregnancy, but a small protective effect remains a possibility.

Effects of new fish oil derivative on fatty acid phospholipid-membrane pattern in a group of Crohn's disease patients

Fish oil has been recently proposed as a possible effective treatment in inflammatory bowel disease (IBD); however, a lot of annoying side effects (ie, belching, halitosis, diarrhea, etc) affect patient compliance. We carried out a study of patient tolerance in a group of Crohn's disease (CD) patients with a new fish oil derivative consisting of 500-mg capsules of eicosapentaenoic-docosahexaenoic (EPA 40%-DHA 20%), a free fatty acid mixture (Purepa), and we also evaluated its incorporation into phospholipids, both in plasma and in red cell membranes. Five groups of 10 CD patients in remission received nine Purepa capsules daily in four different preparations (A: uncoated, B: coated, pH 5.5; C: coated, pH 5.5, 60 min time release; D: coated, pH 6.9) and 12 x 1-g capsules daily of a triglyceride preparation (Max-EPA, EPA 18%-DHA 10%), respectively. We coated three of the four Purepa preparations in order to delay the release of contents in an attempt to minimize the side effects. After six weeks of treatment, the group taking Purepa capsules, coated, pH 5.5, 60 min time release (group C) showed the best incorporation of EPA and DHA in red blood cell phospholipid membranes (EPA from 0.2 to 4.4%, DHA from 3.7 to 6.3%), and no side effects were registered, whereas in all other groups side effects were experienced in 50% or more of subjects. This new preparation will make it possible to treat patients for long periods.

The association between birth weight, placenta weight, pregnancy duration, subfecundity, and child development

It is well known that very low birth weight and preterm birth are risk indicators for delayed child development. It is the purpose of this study to estimate the association between birth weight, placenta weight, and gestational age in consecutive pregnancies which survived till after 28th week of gestation. The association between fecundity and child development is also studied. Data stem from pregnant women in a well defined regional area in Denmark (Odense) who participated in a concerted action on moderate alcohol consumption in pregnancy (EuroMac). All pregnant women with an alcohol consumption of 5 drinks or more per week or more in the first trimester were selected for the study in 1988 to 1989. A one to one match of pregnant woman was selected among the remaining pregnant woman based upon expected time of delivery and age. Altogether 326 women were selected for the study and the two groups are combined since alcohol intake in the measured dose range had no association with child development. The newborn went through two psychological tests at 18 month (the Bayley test) and again at 42 months of age (the Griffiths' test). Two hundred fiftynine pairs of mothers and children participated in all parts of data collection. Birth weight and gestational age was associated with the psychological scoring in the test performed at 18 and 42 months of age, especially the psychomotor index. Especially newborns with low birth weight and high placenta weight had low score values on mental development indices. No association was seen between a measure of fecundity (waiting time to pregnancy) and reduced child development.

A birth weight adjusted comparison of perinatal mortality in the Faroe Islands and Denmark

The objectives were to compare perinatal mortality (PNM) in the Faroes and Denmark while accounting for the high birth weights in the Faroes, and to discuss methodological aspects related to this task. We applied conventional methods employing absolute birth weight standards, and the Wilcox-Russell way of comparing relative birth weights. During 1977-85 perinatal mortality (PNM) in the Faroes was 14.7 (98 cases) per 1,000 births, and 1.57 times higher than that in Denmark. Conventional method: birth weight-standardised risk ratio for PNM in the Faroes v Denmark was 1.95; the risk ratio declined with increasing birth weight. Wilcox-Russell model: the risk tended to be more uniformly increased across the birth weight distribution when babies with same relative birth weights were compared; the residual component of the birth weight distribution (i.e. the excess of observed births in the lower tail beyond what could be predicted by a Gaussian distribution) was 2.1% in the Faroes and 3.6% in Denmark, which does not fit with the model assumption that the size of the residual component is a strong determinant of a population's PNM.

Distribution of unsaturated fatty acids in phospholipids of arteries from nonpregnant, pregnant and fetal sheep

Normal ovine pregnancy is associated with elevated levels of circulating vasodilator prostaglandins (PGs) and increases in PG production by uterine and systemic arteries. We hypothesized that the availability of fatty acid substrate may regulate PG production in vasculature from nonpregnant, pregnant and fetal sheep. In pregnant sheep, levels of arachidonic acid (20:4 omega 6), the immediate PG precursor, were significantly lower in phospholipids from uterine versus systemic (renal) arteries. Although linoleic acid (18:2 omega 6), the primary arachidonic acid precursor, was elevated (p < 0.001) in uterine arteries during pregnancy, levels in systemic arteries were decreased. Levels of omega 3 polyunsaturated fatty acids (PUFAs) also were increased (p < 0.001) in both uterine and systemic arteries during pregnancy. In fetal-placental arteries, the levels of arachidonic and linoleic acid were 50 and 90% less, respectively, than that in maternal arteries. We conclude that during ovine pregnancy vascular prostanoid production may be regulated, in part, by the availability of fatty acid precursors.

Frequency of seafood intake in pregnancy as a determinant of birth weight: evidence for a dose dependent relationship

STUDY OBJECTIVE

The aim was to explore whether maternal consumption of seafood is a determinant of birth weight in a dose dependent manner.

DESIGN

A population based survey of lifestyle factors in pregnancy was linked with information from antenatal and obstetric records.

SETTING AND PARTICIPANTS

Seventy five per cent of all 1362 women who delivered in the Faroe Islands during the study period 1986-87 who gave a structured post partum interview on lifestyle factors.

MAIN RESULTS

Altogether, 2, 6, 16, 33, 26, 14, and 3% of women had consumed approximately 0, 1, 2, 3, 4, 5, and 6+ respectively seafood (fish or whale) dinner meals per week during pregnancy. The average birth weight (p = 0.02) and birth length (p = 0.002) varied significantly between the seven groups, and increased by about 0.2 kg and 1 cm, respectively between women who ate 0 and those who consumed 3 seafood meals per week. Mean birth weight and length tended to level off with further fish consumption: when fitting a second degree polynomial, the quadratic terms were negative and significant for both birth weight (p = 0.005) and length (p = 0.001). Analogous analyses for pregnancy duration were not significant, but exhibited similar trends. All analyses were adjusted for maternal height, weight, parity, age, marital status, and smoking.

CONCLUSIONS

The weight and length of the newborn increased with the frequency of seafood dinner meals consumed in pregnancy but only up to a consumption level of about 3 meals per week.

Fish oil derivatives as a prophylaxis of recurrent miscarriage associated with antiphospholipid antibodies (APL): a pilot study

Since 1989, 22 patients with persistent antiphospholipid syndrome (PAPS) associated with recurrent miscarriage (defined as three or more miscarriages) were treated with fish oil, equivalent to 5.1 g eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) at a ratio of 1.5 EPA to DHA. Twenty-two patients had 23 pregnancies (one patient had two pregnancies) over a period of 3 years. There was only one intrauterine fetal death at the 27th week associated with pre-eclampsia. Twenty-one pregnancies, 19 of which ended after the 37th week, produced a baby. Two pregnancies ended with cesarean section for pre-eclampsia at 30th and 35th week of gestation and one is ongoing at 32nd week. All babies are well. The weight at birth of babies delivered at term was always > 2500 g. These encouraging results favour a therapeutic role, without any adverse reaction, of fish oil to prevent recurrent miscarriage in PAPS.

Maternal characteristics and expected birth weight

Fetal growth charts currently used aggregate birth weights of infants with various natural histories from 1931 until 1967. In order to modernize these charts, avoiding deviation from the natural history of fetal development, we report data from infants born after spontaneous onset of labour in 'normal' pregnancy from a gestational age of 267 to 295 days between 1972 and 1982 (n = 14,113). The relationship between birth weight and gestational age in days was studied by multiple regression analysis, containing dummy variables for parity and gender. The estimated proportion of the variance in the model, attributed to these characteristics, was 15%. This could be improved to 22% by supplementing the model with maternal characteristics such as age, height, mid-pregnancy weight and ethnic origin. According to this extended model, in the Dutch section of the population 511 (4.6%) babies had a birth weight below the 5th percentile, whereas 412 (3.7%) babies would be labeled as such according to the conventional birth weight tables. Moreover, 93 babies would be wrongly considered too small, corresponding with a sensitivity of 62.4%, and 192 babies would be wrongly considered normal, corresponding with a specificity of 99.3%. Integration of the four currently used tables into one, and adjustment for easily available maternal characteristics, could substantially improve classification methods.

Effects of fish oil supplementation in the third trimester of pregnancy on prostacyclin and thromboxane production

OBJECTIVE

Disturbance in thromboxane and prostacyclin biosynthesis has been observed in preeclampsia. We studied whether fish oil supplementation in late pregnancy interferes with maternal and fetal production of thromboxane A2 and prostacyclin I2.

STUDY DESIGN

Forty-seven women in the thirtieth week of pregnancy were randomly assigned in a ratio of 2:1:1 to receive fish oil (2.7 gm of n-3 fatty acid per day [Pikasol], or either olive oil or no oil supplementation as controls. Metabolites of thromboxane A2 and A3 and of prostacyclin I2 and I3 were quantified by mass spectrometry methods in serum and urine, respectively. Maternal serum and urine were sampled at baseline, in the thirty-third and thirty-seventh weeks of pregnancy. Fetal serum was sampled at delivery.

RESULTS

At the thirty-seventh week the mean concentrations of the eicosapentaenoic-derived metabolites, thromboxane B3 and prostacyclin I3, was twofold to threefold higher (p < 0.001) in the group receiving fish oil compared with combined control groups. There were no significant effects of fish oil on the prostacyclin I2 metabolite, although there was a trend toward a reduction in thromboxane B2 in this group. In umbilical cord blood the mean concentration of thromboxane B2 was lowest in the group receiving fish oil (p = 0.03).

CONCLUSIONS

Fish oil was metabolized to the eicosapentaenoic acid-derived eicosanoids thromboxane A3 and prostacyclin I3 in pregnant women. Correspondingly, analog products of arachidonic acid tended to be depressed. It remains to be established whether these biochemical effects will prove beneficial in the prevention or treatment of preeclampsia and intrauterine growth retardation.

Fish consumption and reproductive outcomes in Green Bay, Wisconsin

Prenatal exposure to polychlorinated biphenyls (PCBs) and reproductive outcome was determined in a population of 1112 women during 1987-1989. The women studied were from the Green Bay, Wisconsin area, thereby providing a population with potential PCB exposure from Lake Michigan sport fish consumption. All women with positive pregnancy tests from two Green Bay prenatal clinics were asked to complete a self-administered questionnaire on fish consumption, health and reproductive history, and other relevant issues and to provide blood samples for PCB analysis. A positive correlation was found between the amount of Lake Michigan fish mothers claimed to consume and their PCB serum levels. After the pregnancy period, reproductive outcome measures (fetal wastage; stillbirths; and birth weight, birth length, head circumference, ponderal index, and birth weight percentiles for live births) were abstracted from hospital labor reports. Typical negative associations between birth size measures and consumption of caffeine, smoking, and alcohol were found. Birth size was positively associated with gestational age, birth order, weight gain during pregnancy, male babies, and rural residence. Birth size was also associated with PCB exposure; however contrary to expectations, a positive association was found (P < 0.044) for most mothers (the exception being those mothers who gained more than 34 lb during their pregnancy). PCB exposures were lower than the other studies that found that birth size was negatively associated with PCB exposure.