Characteristics of women in supplement-taking and nonsupplement-taking groups
The total number of participants was 1274. All had information on supplement intake in the first and second trimester; 425 women had information on supplement intake in the third trimester. The proportion of pregnant women taking any form of daily supplements was 82%, 22% and 33% for the first, second and third trimesters, respectively (). Women who reported taking supplements at any stage of pregnancy were more likely to have a university degree and be vegetarian, and less likely to be smokers. They were less likely to be living in an area with an IMD score in the most deprived quartile. Women who reported taking daily supplements in the first and second trimester were more likely to be primiparous. However, there was no difference between primiparous and multiparous women in their use of supplements in the third trimester. There were also no differences between women who reported taking daily supplements at any stage in pregnancy from those who did not with regards to prepregnancy weight, ethnic origin or history of long-term illness. Out of the women who took daily supplements in the third trimester (n = 139), 94% (n = 131) also reported taking daily supplements in the first trimester of their pregnancy and 66% (n = 91) took daily supplements in their second trimester. Only five women who reported taking daily supplements in the third trimester had not taken supplements in the first or second trimester.
Characteristics of women by whether they have reported taking any daily supplements in the first, second and third trimester, Leeds, UK, 2003–06
Based on midwife-administered 24-hour recall dietary assessment at 8–12 weeks of gestation, women in our cohort had average dietary intakes from food above the reference nutrient intake values for most vitamins and minerals21
except vitamin D, iron, folate, selenium and iodine (). The table shows the mean intake in our cohort, the nutrient requirements for adult women plus the additional requirement recommended for consumption during pregnancy, and the proportion of the women in our cohort with dietary intakes above the recommended reference nutrient intake in pregnancy. The mean total energy intake was 2125 kcal/day (95% CI 2084, 2166).
Average daily intakes of vitamins and minerals (from diet alone) based on 24-hour dietary recall at 8–12 weeks of pregnancy, Leeds, UK, 2003–06 (n = 1257)
Type of supplements
Women reported taking 22 different types of supplements including folic acid, iron, combined folic acid–iron preparations, multivitamin–mineral preparations (six brands), evening primrose, cod liver oil, omega 3, vitamin C, vitamin B, vitamin D, vitamin E, vitamin A, calcium, zinc, magnesium and selenium preparations (). Folic acid was the most frequently reported daily supplement in the first trimester. Multivitamin–mineral supplements were the most frequently reported daily supplements in the third trimester.
Number of women taking different types of supplements during pregnancy
Birthweight was known for 1259 babies. The mean birthweight was 3439 g (95% CI 3397, 3461). 4.4% weighed <2500 g (n = 55). 13% (n = 166) weighed less than the tenth centile, 8% (n = 99) less than the fifth centile and 5% (n = 65) less than the third centile. 9% (n = 118) weighed more than the 90th centile. Out of the 1234 pregnancies with information on gestational age, 55 (4.5%) delivered before 37 weeks of gestation.
Relationship between supplement taking and birthweight
Using a multiple linear regression model, taking any type of daily supplement during the first, second or third trimester of pregnancy was not associated with the customised birth centile as a measure of birth size (adjusted difference 2.7, 95% CI: 2.5, 7.8, P= 0.3 for the first trimester; 3.2, 95% CI: 0.9, 7.4, P= 0.1 for the second trimester; and 0.5, 95% CI: 6.0, 7.0, P= 0.9 for the third trimester) when adjusting for cotinine levels, self-reported alcohol intake, IMD group, having a university degree (39%), mother being a vegetarian (9%), history of long-term chronic illness (13%) and past history of miscarriage (24%).
Using birthweight in grams as an outcome, and adjusting for the above factors as well as maternal age, height, ethnicity, prepregnancy weight, parity, gestational age and baby’s sex, there was also no relationship between taking daily supplements at any stage in pregnancy and birthweight (adjusted difference 6 g, 95% CI: 70, 82, P= 0.9 for the first trimester, 24 g, 95% CI: 36, 83, P= 0.4 for the second trimester, and −7 g, 95% CI: 106, 91, P= 0.9 for the third trimester).
When we looked at taking particular types of supplements, taking a daily multivitamin–mineral preparation at any stage in pregnancy was not associated with size at birth using the continuous outcomes of birthweight in grams and customised birth centile, as well as the binary outcome of small-for-gestational-age (less than tenth centile) (). It was not associated with having a baby weighing less than the third centile (adjusted OR = 1.5, 95% CI 0.8, 2.7, P= 0.3 for the first trimester, 1.2, 95% CI 0.5, 2.6, P= 0.7 for the second trimester, 1.6, 95% CI 0.7, 3.7, P= 0.3 for the third trimester). There were no associations with having a baby weighing less than the fifth centile or more than the 90th centile. In addition, taking iron-containing supplements at any stage in pregnancy was not associated with size at birth (data not shown).
The relationship between maternal multivitamin-mineral supplement use during pregnancy and birth outcomes, Leeds, UK, 2003–06
Relationship between supplement taking and preterm birth
We used a logistic regression model to examine the relationship between the risk of preterm birth and patterns of supplement-taking during pregnancy adjusting for salivary cotinine levels, self-reported alcohol intake, vegetarian diet, ethnicity, maternal age, baby’s sex, parity, IMD score, having a university degree, past history of miscarriage and long-term chronic illness. Any type of daily supplement-taking during the third trimester was associated with an increase in the risk of preterm birth (adjusted OR 3.0, 95% CI 1.2, 7.4, P= 0.02). This relationship was not statistically significant for supplement-taking in the second trimester (adjusted OR 1.6, 95% CI 0.8, 3.2, P= 0.2) and was marginally significant in the first trimester, although confidence intervals were wide (adjusted OR 4.3, 95% CI 1.0, 18.2, P= 0.05).
Taking multivitamin–mineral supplement preparations during the third trimester was also associated with an increased risk of preterm birth (adjusted OR 3.4, 95% CI 1.2, 9.6, P= 0.02). This relationship was not statistically significant in the first or second trimester (). When looking at any iron-containing supplement, the relationship remained significant only for supplement-taking in the third trimester (adjusted OR 3.0, 95% CI 1.2, 7.6, P= 0.02).
In addition, we adjusted for the clinical diagnosis of IUGR detected by ultrasound scan during pregnancy and documented in the maternity notes, in the relationship between taking a multivitamin–mineral supplement preparation and both birthweight and preterm birth. The risk of preterm birth when taking supplements in the third trimester (adjusted OR 3.5, 95% CI 1.2, 10.0, P= 0.02) remained broadly unchanged.
To take into account the possibility that the pattern of multivitamin–mineral supplement use is influenced by previous adverse birth outcomes, we also performed the same analysis separately by parity. In primiparous women, the adjusted OR for the relationship between taking multivitamin–mineral supplement in the third trimester and preterm birth was 5.4 (95% CI 1.3, 22.7, P= 0.02). In multiparous women, the adjusted OR was 3.7 (95% CI 0.5, 29.4, P= 0.2). However, numbers were small with resulting wide confidence intervals.