Effect of iron content on the tolerability of prenatal multivitamins in pregnancy

Multivitamin-mineral supplementation is recommended before and during pregnancy to ensure adequate intake of several key nutrients including folate, iron, and more recently vitamin D – all of which have importance among pregnant women, women of childbearing age, and during fetal development [1–6]. Folic acid supplementation has been well documented to reduce the risk of neural tube defects (i.e. spina bifida) [7–9], and emerging evidence suggests that folic acid-containing multivitamin supplements are associated with reducing the risk of other malformations and certain pediatric cancers [10–13]. Furthermore, as pregnancy progresses, body size increases, alongside the nutritional requirements of the developing fetus, resulting in continuous depletion of vitamins and minerals that diet alone may not replenish [14, 15]. Thus, multivitamin-mineral supplementation is considered necessary.

Most commonly used prenatal multivitamins contain a high iron dose (i.e. 60 mg elemental iron) and are large in tablet size. Pregnant women may experience gastrointestinal (GI) symptoms, particularly 'morning sickness', aggravated by the iron content or swallowing a large tablet [16, 17]. Approximately 80% of pregnant women experience some degree of nausea and vomiting of pregnancy (NVP) which may improve towards the end of the first trimester; however, it is not uncommon for symptoms to continue beyond the first trimester or until the end of pregnancy [17, 18]. Pregnant women may also experience heartburn, acid reflux, indigestion, constipation, or diarrhea, and some women may experience exacerbation of GI conditions such as irritable bowel syndrome (IBS), Crohn's disease, or ulcerative colitis. The consequence of GI irritability can deter pregnant women from starting or continuing to take prenatal multivitamins.

In a previously published study, we determined that a twice-daily, small-tablet prenatal multivitamin with a low iron dose had better tolerability than a once-daily, large-tablet prenatal multivitamin with a high iron dose [19]. It would be of interest to assess the tolerability between 2 small-size prenatal multivitamins, differing in iron content. The objective of the present study was to compare rates of adherence and reported adverse events among pregnant women who were randomized to commence supplementation with a small-tablet prenatal multivitamin, containing either a low iron dose or high iron dose.

Between October 2004 and October 2006, 167 pregnant women were enrolled into the study through the Motherisk Program (Hospital for Sick Children), 92 women were randomized to the '35 mg' group (PregVit^®), and 75 women were randomized to the '60 mg' group (Orifer F^®). Although 27 women did not complete the study for various reasons (i.e. miscarriage, lost contact), some data were still collected and included in the analysis (Figure 1). Based on the inclusion criteria, 67% of subjects were enrolled because they had discontinued a previous multivitamin supplement in the current pregnancy, and the most frequently reported reasons for discontinuing or not starting to take any multivitamins were NVP (78%), tablet size or swallowing difficulties (32%), and various GI symptoms (Table 1). There was a noticeable difference between the 2 groups in terms of tablet size and constipation as the reasons reported for non-adherence to any previous multivitamins (Table 1). However, 66% of subjects in the '60 mg' group who reported issues with tablet size as a reason for prior non-adherence, still commenced supplementation with the assigned multivitamin, similarly to the 65% of subjects in the '35 mg' group who also reported prior issues with tablet size yet still commenced supplementation with the assigned multivitamin. As for the differences in previously reported constipation, it also had minimal deterrence considering that 12 of the 14 subjects (86%) in the '35 mg' group, who reported constipation as the reason for prior non-adherence, still commenced supplementation with the assigned multivitamin.

Eighty-four percent of subjects who were pregnant in the past had also supplemented with multivitamins in a previous pregnancy, and 52% of these multivitamin-takers had also discontinued the supplements in a previous pregnancy (Table 1).

A similar proportion of women in both treatment groups commenced supplementation with the assigned multivitamin (73% of '35 mg' group and 76% of '60 mg' group) (Table 2). There was no difference between the 2 groups in the gestational age at which subjects started taking the assigned multivitamins (gestational age as a mean with standard deviation, 15 ± 6 weeks for the '35 mg' group, 16 ± 7 weeks for the '60 mg' group, p = 0.57). Among those who started, only 37–38% of both groups were adherent at 80% or greater pill intake, and a slightly larger proportion of both groups had a pill intake of 50% or greater (Table 2). The range of pill intake for both groups was zero to 100%, and the mean pill intake for both groups was approximately 50%. Thus, no significant differences were detected, and any partial data collected from subjects who did not complete the study did not significantly impact results. Similarly, a large proportion of women with NVP in both groups commenced supplementation with the randomly assigned multivitamins; however, the proportions of NVP women who continued supplementation at 80% or greater pill intake was substantially lower (Table 3), with no significant difference detected between groups. Kaplan-Meier survival curves demonstrated no significant differences between the 2 groups in adherence over time, in terms of the proportion of women with standard adherence, and thus, a pill intake of at least 80% (Figure 2, p = 0.14) and in terms of the proportion of women who continued, after having commenced, supplementation with the randomly assigned multivitamin (overall adherence) (Figure 3, p = 0.59).

Prenatal multivitamin groups, differing in iron content
	35 mg iron group (n = 78)	60 mg iron group (n = 62)	p-value (chi-squared test)
No. who started taking assigned prenatal multivitamin	57 (73%)	47 (76%)	0.86
Proportion who were ≥ 80% adherent†.	21/57 (37%)	18/47 (38%)	0.96
Proportion who were ≥ 50% adherent¶.	32/57 (56%)	28/47 (60%)	0.88

^β No significant difference was detected when partial data from subjects who did not complete the study were included.

^† Based on at least 80% pill intake.

^¶ Based on at least 50% pill intake

Prenatal multivitamin groups, differing in iron content
	35 mg iron group (n = 78)	60 mg iron group (n = 62)	p-value (chi-squared test)
No. of subjects with NVP (at time of enrolment)	67 (86%)	56 (90%)	0.59
Proportion who started taking assigned prenatal multivitamin	49/67 (73%)	44/56 (79%)	0.63
Proportion who were ≥ 80% adherent.ψ	17/49 (35%)	18/44 (41%)	0.69

^βNo significant difference was detected when partial data from subjects who did not complete the study were included.

^ψ Based on at least 80% pill intake

Pregnant women enrolled in the present study were those who contacted the Motherisk counseling program (Toronto, Canada), thus they were probably self-selected by heightened awareness of pregnancy health issues. Sixty-seven percent of the enrolled women had also discontinued previous multivitamin supplements in the current pregnancy due to various GI irritabilities, thus poor adherence and GI adverse events were already established among most women in the study. Although we examined a selective group of pregnant women, the findings from this present study may still be applied to a large proportion of pregnant women in the general population, considering that approximately 80% of pregnant women experience some degree of NVP or GI irritability and that prenatal multivitamin supplementation is commonly recommended in pregnancy despite these potential challenges.

The limited sample size and unequal randomization of the study is a reflection of recruiting and randomizing subjects belonging to a vulnerable population. However, the proportion of subjects who did not complete the study (i.e. drop-out) did not differ between the 2 treatment groups (15% for '35 mg' group, 17% for '60 mg' group).

Results from this randomized, prospective study of pregnant women who had not started or had discontinued previous multivitamins due to adverse events suggested that a prenatal multivitamin with a low iron dose did not improve adherence as there was no significant difference between the '35 mg' and '60 mg' groups. Although a large proportion of women in both groups commenced supplementation with the assigned multivitamins, a substantially reduced proportion of women in both groups were adherent throughout pregnancy (Tables 2). A major finding of our study was that under ideal conditions of high level of motivation and continued supervision, adherence ranged from zero to 100%, and on average women consumed only half of their pills in both groups.

Although recall of pill intake may be associated with biased reporting as it generally results in over-estimation of adherence; the interviewer encouraged subjects to be honest and emphasized that no personal judgment was placed on responses. Self-report has been commonly utilized in several studies examining intervention adherence [20–22].

Our results suggested that the low iron dose did not reduce rates of GI adverse events; however, it is important to acknowledge that the 2 multivitamin tablets differed in a number of ways – volume, shape, iron compound, and regimen – in addition to iron dose which were also considered when comparing the 2 multivitamins in terms of both adherence and GI adverse events. Both prenatal multivitamins were small tablets (relative to other common multivitamin tablet sizes), and although the shapes and volumes of the tablets differed, it did not substantially impact the issue of swallowing difficulties between the 2 groups (Table 4), and thus, no further significant differences were detected in adherence or other adverse events between the 2 groups. These results suggest that tablet size tolerability may reflect a relative perception of small tablets versus large tablets, regardless of exact volume and shape. As for the difference in iron preparations (35 mg as ferrous fumarate and 60 mg as ferrous sulphate), it also had a negligible impact as no significant differences were detected in adherence or GI adverse events between the 2 groups, which is consistent with several studies that examined GI intolerance with respect to iron content (i.e. doses, compounds) [16, 23–25].

It is of interest that with the '35 mg' group, the prenatal multivitamin is taken as 2 tablets per day, separating the iron from calcium and hence preventing their interaction, and allowing a lower dose of elemental iron (35 mg) to yield similar systemic exposure as 60 mg taken with calcium [26]. Despite having to take 2 tablets per day, overall adherence with the '35 mg' group was not lower compared to the once-per-day '60 mg' group. This finding is consistent with the results of a systematic review showing no differences in adherence between once versus twice daily administration, with a significant decrease in adherence only with three doses per day [27].

Multivitamin tolerability depends not only on the supplement itself, but also on whether or not pregnant women suffer from and properly manage NVP symptoms. Similar to the general population of pregnant women, most of the pregnant women in both treatment groups experienced NVP and this may well explain why overall adherence under optimal conditions was not predominantly higher than 50% pill intake.

These new findings must be considered when planning public health strategies to ensure folate and other vitamin and mineral supplementation throughout pregnancy. If prenatal multivitamin formulation remains in the present state, adherence among pregnant women, even with optimal motivation and guidance, may not significantly improve above 50% pill intake; however, this present study demonstrated that potential improvements may be associated with tablet size.