- Split View
-
Views
-
Cite
Cite
Linda Sharp, Zosia Miedzybrodzka, Amanda H. Cardy, Julie Inglis, Londale Madrigal, Simon Barker, David Chesney, Caroline Clark, Nicola Maffulli, The C677T Polymorphism in the Methylenetetrahydrofolate Reductase Gene (MTHFR), Maternal Use of Folic Acid Supplements, and Risk of Isolated Clubfoot: A Case-Parent-Triad Analysis, American Journal of Epidemiology, Volume 164, Issue 9, 1 November 2006, Pages 852–861, https://doi.org/10.1093/aje/kwj285
- Share Icon Share
Abstract
Worldwide, 1–4 per 1,000 births are affected by clubfoot. Clubfoot etiology is unclear, but both genetic and environmental factors are thought to be involved. Low folate status in pregnant women has been implicated in several congenital malformations, and folate metabolism may be affected by polymorphisms in the methylenetetrahydrofolate reductase gene (MTHFR). Using a case-parent-triad design, the authors investigated whether the MTHFR C677T polymorphism, and maternal periconceptional folic acid supplement use, influenced risk of isolated clubfoot. Three hundred seventy-five United Kingdom case-parent triads were recruited in 1998–1999. Among the children, there was a significant trend of decreasing clubfoot risk with increasing number of T alleles: relative risk for CT vs. CC = 0.75, 95% confidence interval: 0.57, 0.97; relative risk for TT vs. CC = 0.57, 95% confidence interval: 0.35, 0.91; p trend = 0.006. This association was not modified by maternal folic acid use. Maternal MTHFR genotype did not influence clubfoot risk for the offspring overall, although a possible interaction with folic acid use was found. This is the first known report of a specific genetic polymorphism associated with clubfoot. The direction of the association is intriguing and suggests that DNA synthesis may be relevant in clubfoot development. However, clubfoot mechanisms are poorly understood, and the folate metabolism pathway is complex. Further research is needed to elucidate these relations.
Congenital talipes equinovarus (clubfoot) is a common developmental disorder of the lower limb, with a prevalence of 1–4 per 1,000 births worldwide (1). It is a three-dimensional deformity immediately recognizable at birth; the ankle is in the plantar flexed (equinus) position, the heel is inverted (varus), and the midfoot and forefoot are inverted and adducted (varus) (2). Treatment can be protracted through childhood, and disability may persist into later life. Although some cases of clubfoot occur in association with other features or congenital anomalies as part of a genetic syndrome, the majority arise in isolation (3, 4). The etiology of isolated clubfoot has been relatively little studied and remains unclear. Several strands of evidence suggest environmental triggers acting in a genetically predisposed individual (4–7). Neither the environmental nor the genetic factors have been elucidated.
The B vitamin folate is crucially involved in several important metabolic processes, including DNA synthesis and repair and DNA methylation (8). Low folate status in pregnant women has been observed to affect risk of several congenital malformations (9, 10). Regarding clubfoot, a small reduction in the birth prevalence of isolated clubfoot has been observed in the United States after fortification of grains with folic acid was introduced (11). In addition, in an intervention study in Denmark, the rate of clubfoot-affected births was lower among women taking folic acid supplements than those not, although the number of clubfoot births was very small (12).
The mechanisms by which maternal folate status affects pregnancy outcome are not established. Supplementation with folic acid might overcome a genetically inherited block in folate metabolism (13). The enzyme 5,10-methylenetetrahydrofolate reductase (MTHFR) catalyzes the conversion of 5-10 methylenetetrahydrofolate into 5-methyltetrahydrofolate, thereby directing the folate pool toward either DNA synthesis and repair or DNA methylation. The C677T polymorphism in the MTHFR gene causes an alanine-to-valine change in the protein (14). Enzyme activity is reduced in a dose-response fashion in heterozygotes and homozygotes for the variant T allele (15). MTHFR has been implicated in the etiology of several congenital malformations—including neural tube defects (10) and orofacial clefts (16)—but, to our knowledge, has not previously been investigated in clubfoot.
We used a case-parent-triad design to investigate, firstly, whether the MTHFR C677T polymorphism affects risk of isolated clubfoot and, secondly, whether associations between genotype and clubfoot risk are modified by maternal periconceptional use of folic acid supplements.
MATERIALS AND METHODS
Recruitment of case-parent triads
We recruited two series of children affected by clubfoot and their parents. The source of one series was the United Kingdom support group for children with lower limb deformities, STEPS. All families registered with STEPS as having a child affected by clubfoot were invited to take part in the study. Parents were approached by mail and were asked whether they, and their affected child, would participate. Recruitment took place during 2001–2002. Participants provided a buccal DNA sample, collected by mouthwash or a cytocell cheek smear. Parents, and children who were old enough, provided their own sample. Samples from young children were collected by their parents by using a cytocell brush. Mothers completed a questionnaire on socioeconomic factors, ethnicity, the pregnancy and birth of the index child, the nature of the child's clubfoot (laterality, etc.), the child's other medical conditions (to enable assessment of syndromic status and exclusion of children with other foot deformities), family history of clubfoot, maternal reproductive history, and maternal use of supplements and consumption of alcohol during the index pregnancy. Fifty-two percent of eligible families took part.
The other series was identified via registers maintained by orthopedic surgeons treating clubfoot in Scotland and northeast England. Families of all children recorded on the register were approached and were invited to take part in the study. Recruitment took place during 1998–1999 (7, 17). Families were asked to attend a hospital clinic appointment; if they agreed, mouthwash samples were collected from the child and parents, and a questionnaire similar to the one described above was completed during the visit. If the family declined to attend the clinic, they were sent the mouthwash kits and questionnaire by mail and were asked to complete and return them. Forty-two percent of eligible families took part.
Only biologically related triads were included. Checks were made to ensure that no family was included in both series. Recruited were 289 triads from the first series and 98 from the second. None of the first series were excluded on the basis of having foot deformity that was not talipes equinovarus. Of the 387, 12 cases were assessed by a clinical geneticist (Z. M.) to be syndromic and were excluded, leaving 375 triads in the analysis.
The study was approved by the Multi-Centre Research Ethics Committee for Scotland and the Grampian Research Ethics Committee.
Genotyping methods
DNA was extracted from the cheek smears and mouthwashes by using Instagene matrix (Bio-Rad, Hercules, California) and sodium hydroxide, respectively. The C677T polymorphism was detected by using restriction fragment length polymorphism methods devised by Frosst et al. (18) for mothers, fathers, and children from all 375 triads (i.e., 1,125 genotype determinations were undertaken). DNA was amplified by polymerase chain reaction using flanking primers. Polymerase chain reaction products were digested with HinfI, separated by 3 percent agarose gel electrophoresis, and visualized by ethidium bromide staining and ultraviolet transillumination.
Genotyping was undertaken without knowledge of exposure status (e.g., maternal folic acid use) of the triad. Gels were double-read (i.e., two people read each gel and assigned a genotype; discrepancies were discussed and resolved). Genotyping was repeated in a random sample of 130 samples (11.6 percent of 1,125), blind to the original genotyping result. No differences were found between the original and repeated genotypes.
Classification of folic acid, alcohol consumption, and family history
The questionnaires sought information on whether the mother had taken supplements containing folic acid in the 3 months prior to and/or in the first trimester of the index pregnancy. “Folic acid containing supplements” were defined as specific folic acid preparations intended for the periconceptional period, vitamin B-complex, or multivitamins. Period of use was classified as use during the 1) 3 months preconception and/or the first trimester, 2) 3 months preconception, or 3) first trimester. Data were missing on preconception use for seven triads (1.9 percent) and on first-trimester use for six (1.6 percent). Because alcohol has an adverse effect on folate metabolism (19), mothers were classified dichotomously according to whether they had consumed alcohol at any time during the index pregnancy. This information was missing for two triads (0.5 percent). A positive family history of clubfoot was defined as a report of clubfoot in any first-, second-, or third-degree relative of the index child. These data were missing for two triads (0.5 percent).
Statistical analysis
The analysis was conducted by using Stata 8 software (20). The Pearson χ2 test was used to determine whether genotype frequencies in mothers, fathers, and children were in Hardy-Weinberg equilibrium. Log-linear methods were used to calculate relative risks associated with maternal and child alleles (21, 22). These methods test for asymmetric distribution of the variant allele among affected offspring and their parents. The genotypes of cases and parents were stratified into 15 possible mating types, from which the relative risks for the child's and the mother's genotypes were computed by fitting a Poisson model. The initial analysis estimated separate relative risks for heterozygous (CT) and homozygous (TT) variant genotypes in both children and mothers, using homozygous wild-type (CC) as the reference category. Trend tests were applied to assess whether there was a “dose-response” pattern in risk with increasing number of variant alleles (23). Because it is not clear what model of inheritance might be postulated a priori, this analysis was repeated by assuming both dominant (i.e., CT/TT vs. CC) and recessive (i.e., TT vs. CC/CT) transmission.
Interactions between genotype and folic acid use were determined by stratifying on whether the mother used supplements during the index pregnancy and comparing the strength of the associations with genotype between the strata by using a χ2 -based test of heterogeneity (24). This assessment was done for the three categorizations of folic acid use. To increase statistical power in the interaction analysis, a dominant effect of the variant allele was assumed. Interactions between genotype and maternal alcohol consumption during the index pregnancy were investigated in the same way. To consider the possibility of etiologically distinct subgroups of clubfoot, the effects of maternal and child genotypes were compared in families with and without a family history of clubfoot.
The statistical models can be fitted with or without an assumption of Hardy-Weinberg equilibrium. The goodness of fit of both sets of models was compared; those without the assumption of Hardy-Weinberg equilibrium provided the better fit and were used in all analyses.
RESULTS
The characteristics of the 375 case-parent triads are summarized in table 1. The series included more male than female cases (male:female ratio 2.3:1). Sixty percent of the children had bilateral clubfoot, and, of those affected unilaterally, the right foot was involved slightly more often than the left. The median maternal and paternal ages at the birth of the affected child were 30 (range, 18–45) and 32 (range, 17–58) years, respectively. Seventeen percent of children had a positive family history of clubfoot.
. | No. . | %*,† . |
---|---|---|
Sex of the index child | ||
Male | 261 | 70 |
Female | 114 | 30 |
Affected foot | ||
Left only | 62 | 18 |
Right only | 80 | 23 |
Both | 211 | 60 |
Data missing | 22 | |
Period of birth of the index child | ||
1963–1979 | 15 | 4 |
1980–1989 | 52 | 14 |
1990–1994 | 131 | 35 |
1995–1999 | 152 | 40 |
2000–2002 | 24 | 7 |
Data missing | 1 | |
Age of the mother (years) at birth of the index child | ||
≤19 | 3 | 1 |
20–24 | 36 | 10 |
25–29 | 124 | 34 |
30–34 | 140 | 37 |
35–39 | 55 | 15 |
≥40 | 14 | 4 |
Data missing | 3 | |
Age of the father (years) at birth of the index child | ||
≤19 | 3 | 1 |
20–24 | 16 | 4 |
25–29 | 81 | 22 |
30–34 | 138 | 37 |
35–39 | 75 | 20 |
≥40 | 58 | 16 |
Data missing | 4 | |
Family history of clubfoot‡ | ||
Yes | 64 | 17 |
No | 309 | 83 |
Data missing | 2 | |
Maternal alcohol consumption during the index pregnancy | ||
Yes | 174 | 46 |
No | 199 | 54 |
Data missing | 2 | |
Maternal folic acid use in the 3 months prior to, or during the first trimester of, the index pregnancy | ||
Yes | 188 | 51 |
No | 182 | 49 |
Data missing | 5 |
. | No. . | %*,† . |
---|---|---|
Sex of the index child | ||
Male | 261 | 70 |
Female | 114 | 30 |
Affected foot | ||
Left only | 62 | 18 |
Right only | 80 | 23 |
Both | 211 | 60 |
Data missing | 22 | |
Period of birth of the index child | ||
1963–1979 | 15 | 4 |
1980–1989 | 52 | 14 |
1990–1994 | 131 | 35 |
1995–1999 | 152 | 40 |
2000–2002 | 24 | 7 |
Data missing | 1 | |
Age of the mother (years) at birth of the index child | ||
≤19 | 3 | 1 |
20–24 | 36 | 10 |
25–29 | 124 | 34 |
30–34 | 140 | 37 |
35–39 | 55 | 15 |
≥40 | 14 | 4 |
Data missing | 3 | |
Age of the father (years) at birth of the index child | ||
≤19 | 3 | 1 |
20–24 | 16 | 4 |
25–29 | 81 | 22 |
30–34 | 138 | 37 |
35–39 | 75 | 20 |
≥40 | 58 | 16 |
Data missing | 4 | |
Family history of clubfoot‡ | ||
Yes | 64 | 17 |
No | 309 | 83 |
Data missing | 2 | |
Maternal alcohol consumption during the index pregnancy | ||
Yes | 174 | 46 |
No | 199 | 54 |
Data missing | 2 | |
Maternal folic acid use in the 3 months prior to, or during the first trimester of, the index pregnancy | ||
Yes | 188 | 51 |
No | 182 | 49 |
Data missing | 5 |
Excluding triads for which the relevant data were missing.
Percentages may not total 100 because of rounding.
Includes up to third-degree relatives of the index child.
. | No. . | %*,† . |
---|---|---|
Sex of the index child | ||
Male | 261 | 70 |
Female | 114 | 30 |
Affected foot | ||
Left only | 62 | 18 |
Right only | 80 | 23 |
Both | 211 | 60 |
Data missing | 22 | |
Period of birth of the index child | ||
1963–1979 | 15 | 4 |
1980–1989 | 52 | 14 |
1990–1994 | 131 | 35 |
1995–1999 | 152 | 40 |
2000–2002 | 24 | 7 |
Data missing | 1 | |
Age of the mother (years) at birth of the index child | ||
≤19 | 3 | 1 |
20–24 | 36 | 10 |
25–29 | 124 | 34 |
30–34 | 140 | 37 |
35–39 | 55 | 15 |
≥40 | 14 | 4 |
Data missing | 3 | |
Age of the father (years) at birth of the index child | ||
≤19 | 3 | 1 |
20–24 | 16 | 4 |
25–29 | 81 | 22 |
30–34 | 138 | 37 |
35–39 | 75 | 20 |
≥40 | 58 | 16 |
Data missing | 4 | |
Family history of clubfoot‡ | ||
Yes | 64 | 17 |
No | 309 | 83 |
Data missing | 2 | |
Maternal alcohol consumption during the index pregnancy | ||
Yes | 174 | 46 |
No | 199 | 54 |
Data missing | 2 | |
Maternal folic acid use in the 3 months prior to, or during the first trimester of, the index pregnancy | ||
Yes | 188 | 51 |
No | 182 | 49 |
Data missing | 5 |
. | No. . | %*,† . |
---|---|---|
Sex of the index child | ||
Male | 261 | 70 |
Female | 114 | 30 |
Affected foot | ||
Left only | 62 | 18 |
Right only | 80 | 23 |
Both | 211 | 60 |
Data missing | 22 | |
Period of birth of the index child | ||
1963–1979 | 15 | 4 |
1980–1989 | 52 | 14 |
1990–1994 | 131 | 35 |
1995–1999 | 152 | 40 |
2000–2002 | 24 | 7 |
Data missing | 1 | |
Age of the mother (years) at birth of the index child | ||
≤19 | 3 | 1 |
20–24 | 36 | 10 |
25–29 | 124 | 34 |
30–34 | 140 | 37 |
35–39 | 55 | 15 |
≥40 | 14 | 4 |
Data missing | 3 | |
Age of the father (years) at birth of the index child | ||
≤19 | 3 | 1 |
20–24 | 16 | 4 |
25–29 | 81 | 22 |
30–34 | 138 | 37 |
35–39 | 75 | 20 |
≥40 | 58 | 16 |
Data missing | 4 | |
Family history of clubfoot‡ | ||
Yes | 64 | 17 |
No | 309 | 83 |
Data missing | 2 | |
Maternal alcohol consumption during the index pregnancy | ||
Yes | 174 | 46 |
No | 199 | 54 |
Data missing | 2 | |
Maternal folic acid use in the 3 months prior to, or during the first trimester of, the index pregnancy | ||
Yes | 188 | 51 |
No | 182 | 49 |
Data missing | 5 |
Excluding triads for which the relevant data were missing.
Percentages may not total 100 because of rounding.
Includes up to third-degree relatives of the index child.
The homozygous variant genotype (TT) was present in 10 percent of children and in the same proportions of mothers and of fathers. Forty-one percent of children were heterozygous compared with 47 percent of mothers and 50 percent of fathers. Genotype frequencies in fathers, mothers, and children did not depart from Hardy-Weinberg equilibrium (p > 0.05). Table 2 shows the results of the log-linear analysis of maternal and child genotypes. For children, compared with CC individuals, heterozygotes (CT) had a modest, statistically significant, reduced risk of clubfoot (relative risk (RR) for CT vs. CC = 0.75, 95 percent confidence interval (CI): 0.57, 0.97). Risk was substantially, and significantly, decreased in homozygous variant children (RR for TT vs. CC = 0.57, 95 percent CI: 0.35,0.91), and there was a strong linear trend in risk with number of variant T alleles (p = 0.006). When the analysis was repeated by assuming either dominant or recessive models, the association was still evident, but the relative risks were somewhat ameliorated (as would be expected based on the observed trend with number of variant alleles). There was no strong association between maternal genotype and risk of clubfoot in the offspring. Compared with offspring of CC mothers, offspring of CT or TT mothers had a slight, nonsignificant, decreased risk of clubfoot.
Model and genotype . | No. . | % . | RR* . | 95% CI* . | LRT*,† . |
---|---|---|---|---|---|
Primary analysis | |||||
Mother | |||||
CC | 160 | 42.7 | 1.0 | ||
CT | 177 | 47.2 | 0.88 | 0.65, 1.20 | |
TT | 38 | 10.1 | 0.94 | 0.55, 1.62 | \(\mathrm{{\chi}}_{2}^{2}\) = 0.61, p = 0.736 |
Child | |||||
CC | 182 | 48.5 | 1.0 | ||
CT | 154 | 41.1 | 0.75 | 0.57, 0.97 | |
TT | 39 | 10.4 | 0.57‡ | 0.35, 0.91 | \(\mathrm{{\chi}}_{2}^{2}\) = 7.56, p = 0.024 |
Assumed recessive model | |||||
Mother | |||||
CC/CT | 337 | 89.9 | 1.0 | ||
TT | 38 | 10.1 | 1.0 | 0.59, 1.69 | \(\mathrm{{\chi}}_{1}^{2}\) = 0.00, p = 1.00 |
Child | |||||
CC/CT | 336 | 89.6 | 1.0 | ||
TT | 39 | 10.4 | 0.70 | 0.46, 1.08 | \(\mathrm{{\chi}}_{1}^{2}\) = 2.75, p = 0.097 |
Assumed dominant model | |||||
Mother | |||||
CC | 160 | 42.7 | 1.0 | ||
CT/TT | 215 | 57.3 | 0.89 | 0.67, 1.20 | \(\mathrm{{\chi}}_{1}^{2}\) = 0.56, p = 0.453 |
Child | |||||
CC | 182 | 48.5 | 1.0 | ||
CT/TT | 193 | 51.5 | 0.72 | 0.56, 0.94 | \(\mathrm{{\chi}}_{1}^{2}\) = 5.82, p = 0.016 |
Model and genotype . | No. . | % . | RR* . | 95% CI* . | LRT*,† . |
---|---|---|---|---|---|
Primary analysis | |||||
Mother | |||||
CC | 160 | 42.7 | 1.0 | ||
CT | 177 | 47.2 | 0.88 | 0.65, 1.20 | |
TT | 38 | 10.1 | 0.94 | 0.55, 1.62 | \(\mathrm{{\chi}}_{2}^{2}\) = 0.61, p = 0.736 |
Child | |||||
CC | 182 | 48.5 | 1.0 | ||
CT | 154 | 41.1 | 0.75 | 0.57, 0.97 | |
TT | 39 | 10.4 | 0.57‡ | 0.35, 0.91 | \(\mathrm{{\chi}}_{2}^{2}\) = 7.56, p = 0.024 |
Assumed recessive model | |||||
Mother | |||||
CC/CT | 337 | 89.9 | 1.0 | ||
TT | 38 | 10.1 | 1.0 | 0.59, 1.69 | \(\mathrm{{\chi}}_{1}^{2}\) = 0.00, p = 1.00 |
Child | |||||
CC/CT | 336 | 89.6 | 1.0 | ||
TT | 39 | 10.4 | 0.70 | 0.46, 1.08 | \(\mathrm{{\chi}}_{1}^{2}\) = 2.75, p = 0.097 |
Assumed dominant model | |||||
Mother | |||||
CC | 160 | 42.7 | 1.0 | ||
CT/TT | 215 | 57.3 | 0.89 | 0.67, 1.20 | \(\mathrm{{\chi}}_{1}^{2}\) = 0.56, p = 0.453 |
Child | |||||
CC | 182 | 48.5 | 1.0 | ||
CT/TT | 193 | 51.5 | 0.72 | 0.56, 0.94 | \(\mathrm{{\chi}}_{1}^{2}\) = 5.82, p = 0.016 |
RR, relative risk; CI, confidence interval; LRT, likelihood ratio test.
Chi-squared statistic and p value from likelihood ratio test for the overall contribution of the mother or child to the statistical model.
Likelihood ratio test for linear trend in risk with number of variant alleles,
Model and genotype . | No. . | % . | RR* . | 95% CI* . | LRT*,† . |
---|---|---|---|---|---|
Primary analysis | |||||
Mother | |||||
CC | 160 | 42.7 | 1.0 | ||
CT | 177 | 47.2 | 0.88 | 0.65, 1.20 | |
TT | 38 | 10.1 | 0.94 | 0.55, 1.62 | \(\mathrm{{\chi}}_{2}^{2}\) = 0.61, p = 0.736 |
Child | |||||
CC | 182 | 48.5 | 1.0 | ||
CT | 154 | 41.1 | 0.75 | 0.57, 0.97 | |
TT | 39 | 10.4 | 0.57‡ | 0.35, 0.91 | \(\mathrm{{\chi}}_{2}^{2}\) = 7.56, p = 0.024 |
Assumed recessive model | |||||
Mother | |||||
CC/CT | 337 | 89.9 | 1.0 | ||
TT | 38 | 10.1 | 1.0 | 0.59, 1.69 | \(\mathrm{{\chi}}_{1}^{2}\) = 0.00, p = 1.00 |
Child | |||||
CC/CT | 336 | 89.6 | 1.0 | ||
TT | 39 | 10.4 | 0.70 | 0.46, 1.08 | \(\mathrm{{\chi}}_{1}^{2}\) = 2.75, p = 0.097 |
Assumed dominant model | |||||
Mother | |||||
CC | 160 | 42.7 | 1.0 | ||
CT/TT | 215 | 57.3 | 0.89 | 0.67, 1.20 | \(\mathrm{{\chi}}_{1}^{2}\) = 0.56, p = 0.453 |
Child | |||||
CC | 182 | 48.5 | 1.0 | ||
CT/TT | 193 | 51.5 | 0.72 | 0.56, 0.94 | \(\mathrm{{\chi}}_{1}^{2}\) = 5.82, p = 0.016 |
Model and genotype . | No. . | % . | RR* . | 95% CI* . | LRT*,† . |
---|---|---|---|---|---|
Primary analysis | |||||
Mother | |||||
CC | 160 | 42.7 | 1.0 | ||
CT | 177 | 47.2 | 0.88 | 0.65, 1.20 | |
TT | 38 | 10.1 | 0.94 | 0.55, 1.62 | \(\mathrm{{\chi}}_{2}^{2}\) = 0.61, p = 0.736 |
Child | |||||
CC | 182 | 48.5 | 1.0 | ||
CT | 154 | 41.1 | 0.75 | 0.57, 0.97 | |
TT | 39 | 10.4 | 0.57‡ | 0.35, 0.91 | \(\mathrm{{\chi}}_{2}^{2}\) = 7.56, p = 0.024 |
Assumed recessive model | |||||
Mother | |||||
CC/CT | 337 | 89.9 | 1.0 | ||
TT | 38 | 10.1 | 1.0 | 0.59, 1.69 | \(\mathrm{{\chi}}_{1}^{2}\) = 0.00, p = 1.00 |
Child | |||||
CC/CT | 336 | 89.6 | 1.0 | ||
TT | 39 | 10.4 | 0.70 | 0.46, 1.08 | \(\mathrm{{\chi}}_{1}^{2}\) = 2.75, p = 0.097 |
Assumed dominant model | |||||
Mother | |||||
CC | 160 | 42.7 | 1.0 | ||
CT/TT | 215 | 57.3 | 0.89 | 0.67, 1.20 | \(\mathrm{{\chi}}_{1}^{2}\) = 0.56, p = 0.453 |
Child | |||||
CC | 182 | 48.5 | 1.0 | ||
CT/TT | 193 | 51.5 | 0.72 | 0.56, 0.94 | \(\mathrm{{\chi}}_{1}^{2}\) = 5.82, p = 0.016 |
RR, relative risk; CI, confidence interval; LRT, likelihood ratio test.
Chi-squared statistic and p value from likelihood ratio test for the overall contribution of the mother or child to the statistical model.
Likelihood ratio test for linear trend in risk with number of variant alleles,
Folic acid–containing supplements were taken by 24 percent of mothers in the 3 months prior to the index pregnancy, by 49 percent in the first trimester, and by 51 percent in either time period. The effects of genotype stratified by maternal folic acid use in either time period are shown in figure 1. For mothers, carrying the T allele was associated with a modest reduced risk of clubfoot in the offspring among those who had used folic acid (RR for CT/TT vs. CC = 0.72, 95 percent CI: 0.47, 1.09) but not among those who had not used folic acid (RR = 1.01, 95 percent CI: 0.68, 1.51). However, the test for interaction was not statistically significant (p interaction = 0.218). Among children, the relative risk of clubfoot for CT/TT versus CC genotypes was reduced irrespective of maternal folic acid use (folic acid not used: RR = 0.69, 95 percent CI: 0.48, 0.99; folic acid used: RR = 0.74, 95 percent CI: 0.52, 1.05; p interaction = 0.778). When the analysis was repeated for preconception and first-trimester use of folic acid separately, the results were very similar (data not shown).
Forty-six percent of mothers reported consuming alcoholic drinks at some point during the index pregnancy. We found no interaction between maternal alcohol consumption and child's genotype (p = 0.883; figure 2). Regarding maternal genotype, the risk of clubfoot in the offspring was reduced for only those women with the T allele who consumed alcohol (RR for CT/TT vs. CC = 0.65, 95 percent CI: 0.42, 1.00); the test for interaction was borderline significant (p = 0.060).
There was no evidence of an interaction between family history of clubfoot and either maternal (p interaction = 0.522) or child (p = 0.479) genotype (figure 3).
DISCUSSION
Although MTHFR has been investigated in several congenital malformations (10, 16, 25), to our knowledge this is the first study of genetic variation in folate metabolism in clubfoot. For children, carrying the variant 677T allele was associated with a significantly reduced risk of isolated clubfoot, and risk decreased with increasing number of variant alleles. This association was not affected by whether the mother took supplemental folic acid in the periconceptional period. There was a suggestion of an interaction between maternal genotype and folic acid use and risk of clubfoot in the offspring, but the tests of interaction were not statistically significant.
Strengths and limitations of the study
The mains strengths of our study are its size and ability to investigate combinations of maternal and child genotype and maternal folic acid use. We included 375 complete triads, which provided 80 percent power to detect a relative risk of 0.67 associated with a genotype occurring in 10 percent, assuming dominant transmission and a type I error of 0.05. A further strength was the case-parental-control design, which is efficient and has greater power than the case-unrelated-control approach for investigating associations between polymorphic genes and disease (26, 27). Moreover, it avoids the possibility of population stratification (i.e., that case-control differences are due to selection of controls whose genetic background differs systematically from that of cases (28)).
A priori, it was not clear whether the child's or the mother's genotype, or both, were etiologically relevant. Confounding occurs if the child's genotype is measured when the mother's genotype was actually the relevant one (29). To avoid this problem, we used the log-linear approach developed by Wilcox et al. (21) and Weinberg et al. (22), which permits investigation of the independent effects of both maternal and child genotypes.
Several assumptions underpin the case-parent design. Firstly, there must be mendelian transmission of alleles in the population. It seems unlikely that this assumption was violated. Secondly, survival of the index child from fertilization must not be associated with genotype. Studies of MTHFR and fetal loss are inconsistent (30–33), and there is no evidence on MTHFR and survival after birth. Finally, in respect to interactions, it is essential that the genetic and environmental factors are independent. There is no evidence that periconceptional folic acid use and MTHFR genotype are related.
The main limitation of the study is that the cases were not drawn from a population-based sampling frame. However, for the results to be seriously biased, the probability of participation would need to have been associated with maternal or child genotype. This seems unlikely, particularly because there is no known MTHFR-clubfoot association, and participants were not aware of the study hypotheses when they agreed to take part. Ten percent of mothers and fathers in the study had the TT genotype, which is consistent with other United Kingdom series (10). The male excess and dominance of bilateral disease in our study population mirrors patterns in other European and US studies of children with clubfoot (34–41). The frequency of children with a family history of clubfoot (17 percent) falls between estimates from two US investigators (14 percent and 24 percent) (41, 42). Altogether, this finding suggests that our series is unlikely to be seriously biased.
A further limitation of the study is that we did not have detailed information on the dose, frequency, and duration of maternal use of either folic acid supplements or alcohol during the index pregnancy. In the United Kingdom, the proportions of pregnant women taking folic acid supplements varies by time period, maternal age, and social class (43), and it is probable that exposure levels among those who used supplements vary considerably. This variation is likely to be reflected in our study population, making it possible that our results are biased, most likely toward the null.
Interpretation of interaction analyses
When all triads were analyzed together, we found no association between maternal genotype and the risk of having a clubfoot-affected child. However, when mothers were stratified by folic acid use in the periconceptional period, the T allele was associated with reduced risk of clubfoot in the offspring for mothers who had used folic acid (RR = 0.72) but not for mothers who had not used folic acid (RR = 1.01). In the normal embryo, the foot is formed in the vertical plane, and rotation into the plantar (sole on the ground) position commences between 9 and 13 weeks and continues throughout gestation (44), suggesting that first-trimester exposures—such as folic acid supplementation—may influence the process of foot development. When the maternal genotype–folic acid use analysis was repeated by limiting folic acid use to the first trimester, the difference in the stratum-specific relative risks was slightly more pronounced (nonusers: RR for CT/TT vs. CC = 1.06, 95 percent CI: 0.72,1.57; users: RR for CT/TT vs. CC = 0.67, 95 percent CI: 0.44, 1.03). Although not statistically significant, these results suggest an interaction between maternal folic acid supplement use and maternal genotype in relation to clubfoot in the offspring. Caution is needed in interpretation, however, because the case-parent design does not enable comparison of absolute levels of risk, only comparison of relative effects of the genetic variant in each exposure stratum (45). Hence, we cannot discriminate between the situation in which clubfoot risk is reduced by folic acid use among mothers with the T allele and that in which clubfoot risk is increased by folic acid use among mothers without the T allele.
It is well established that alcohol interferes with folate absorption and utilization (19, 46), and this knowledge provided the rationale for investigating interactions between maternal alcohol consumption and maternal and child genotype. Although we observed a borderline significant interaction between maternal alcohol consumption and maternal genotype, interestingly, this interaction was in the same direction as the maternal genotype–folic acid association (i.e., the risk reduction associated with the T allele was observed among only those mothers who reported drinking alcohol). This finding may reflect a positive association between alcohol consumption and folic acid use in our study population; 44 percent of mothers who did not have any alcoholic drinks reported taking folic acid supplements compared with 58 percent who did drink alcohol (
Honein et al. (47) reported an interaction between family history and maternal smoking during pregnancy, suggesting that the etiology of clubfoot in cases with and without a family history of the disease might differ. We found no notable difference in the effects of genotype according to whether the child had a positive or negative family history. However, only 64 triads had a positive family history, so the possibility of an MTHFR–family history interaction cannot be excluded.
Possible mechanisms
Our observation of a reduced risk of clubfoot in children with the variant T allele is intriguing, particularly in view of the fact that the T allele is associated with reduced enzyme activity (15). The product of the MTHFR reaction, 5-methyltetrahydrofolate, is a major methyl donor in the remethylation of homocysteine to methionine. Homocysteine, or its derivatives, in high doses may have toxic effects on developing tissues (48), and two studies have reported an association between increased maternal homocysteine levels and clubfoot (49, 50). In several (although not all) studies (51–53), carriers of the T allele have been found to have raised homocysteine levels. At first glance, these observations would appear to be inconsistent with the results of the current study. However, the nature of the homocysteine-MTHFR relation has not yet been fully clarified. Several investigators have found that the relation holds only when folate status is low (54–59). Others have suggested that the relation might be present in some age groups but not others (60, 61) and/or might vary by gender (57, 59), smoking status (57), weight (62), alcohol intake (63), or riboflavin status (64). In addition, the two studies of homocysteine and clubfoot were flawed; both included relatively small numbers of clubfoot-affected births (10 in one study and 62 in the other) and measured maternal homocysteine levels sometime after the relevant pregnancies (49, 50). Thus, it is possible that the effects of MTHFR in clubfoot may not operate through the homocysteine pathway (although further evidence is needed before this pathway can be definitely excluded).
The substrate for the MTHFR enzyme, 5,10-methylenetetrahydrofolate, is required as a methyl donor for the conversion of deoxyuridine monophosphate to thymidine monophosphate, which is then used for DNA synthesis and repair. When folate is depleted, the conversion is blocked, deoxyuridine monophosphate accumulates, and uracil may be misincorporated into DNA in place of thymine, leading to catastrophic DNA repair, DNA double strand breakage, and chromosome damage (65, 66). Decreased MTHFR activity provides increased methylenetetrahydrofolate for DNA synthesis, reducing uracil misincorporation, DNA instability, and chromosome breaks. The decreased risk of clubfoot with the “low-activity” T allele suggests that this pathway could be relevant in clubfoot development, such that the presence of the MTHFR T allele in the fetus provides increased levels of circulating methylenetetrahydrofolate, thus maintaining (or raising) DNA repair capacity and protecting against clubfoot. We acknowledge, however, that this potential mechanism is speculative, and further investigation is needed to confirm (or refute) these speculations. Many of the genes involved in repair of strand breaks are essential for normal embryo development, and DNA repair capacity is thought to be involved in the response of the conceptus to genotoxic agents that could induce malformations (67). However, current knowledge regarding DNA repair systems in organogenesis is incomplete (67), and human evidence on the roles of MTHFR and folate in maintaining DNA integrity is limited and inconclusive (68–73).
Interestingly, our results are broadly compatible with what has been reported for colorectal cancer, with regard to both the main effect of C677T and interactions with folate (reviewed by Sharp and Little (74)). In most of the available studies, colorectal cancer risk is reduced in those with the TT genotype, and the group at lowest disease risk are T allele carriers with high folate levels (refer, for example, to Ma et al. (51), Chen et al. (75), and Le Marchand et al. (76)). These relations were not what might have been expected a priori (on the basis of the reported inverse relation between folate and colorectal cancer) and suggest that the roles of folate and folate metabolism genes in the disease etiology are complex (74). By analogy, this conclusion is also likely to apply to clubfoot.
Conclusions
We found that children who carry the 677T variant of the MTHFR gene are at decreased risk of clubfoot. Moreover, risk may be affected by an interaction between maternal genotype and supplement use during the index pregnancy. These findings suggest that folate status could be relevant in clubfoot etiology. The importance of replication of findings of genetic association studies has been emphasized in recent years (75, 77), and we recognize that our novel results require confirmation.
Folate metabolism is complex, and several other nutrients and polymorphic genes are involved (74, 78). In addition, despite numerous hypotheses, the pathogenesis of clubfoot is very poorly understood (4). Further etiologic and mechanistic research is warranted to elucidate the role of the folate pathway in this common developmental disorder.
The first two authors contributed equally to this work, which was performed at the University of Aberdeen.
The research was funded by Sports Action Research for Kids (SPARKs). L. M. was funded by BBSRC.
The authors are grateful to STEPS (National Association for Children with Lower Limb Abnormalities) for facilitating the recruitment of families affected by clubfoot, to Martine Barnes and Hazel Hailey for administrative support, and to Ewan Stronach for assistance with DNA extraction. They also thank the orthopedic surgeons who agreed to let the authors approach their patients.
Conflict of interest: none declared.
References
Carey M, Bower C, Mylvaganam A, et al. Talipes equinovarus in Western Australia.
Kelsey J. Epidemiology of musculoskeletal disorders. In: Kelsey J, ed. Monographs in epidemiology and biostatistics. New York, NY: Oxford University Press,
Chung CS, Nemechek RW, Larsen IJ, et al. Genetic and epidemiological studies of club foot in Hawaii. General and medical considerations.
Miedzybrodzka Z. Congenital talipes equinovarus (clubfoot): a disorder of the foot but not the hand.
Chapman C, Stott NS, Port RM, et al. Genetics of club foot in Maori and Pacific people.
Miedzybrodzka Z, Chesney D, Barker S, et al. The genetic basis of idiopathic congenital talipes equinovarus. (Abstract).
Duthie S. Folic acid deficiency and cancer: mechanisms of DNA instability.
Prevention of neural tube defects: results of the Medical Research Council Vitamin Study. MRC Vitamin Study Research Group.
Botto LD, Yang Q. 5,10-Methylenetetrahydrofolate reductase gene variants and congenital anomalies: a HuGE review.
Moorthi RN, Hashmi SS, Langois P, et al. Idiopathic talipes equinovarus (ITEV) (clubfeet) in Texas.
Ulrich M, Kristoffersen K, Rolschau J, et al. The influence of folic acid supplement on the outcome of pregnancies in the county of Funen in Denmark. Part II: congenital anomalies: a randomised study.
Goyette P, Sumner JS, Milos R, et al. Human methylenetetrahydrofolate reductase: isolation of cDNA, mapping and mutation identification.
Rozen R. Genetic predisposition to hyperhomocysteinemia: deficiency of methylenetetrahydrofolate reductase (MTHFR).
Jugessur A, Wilcox AJ, Lie RT, et al. Exploring the effects of methylenetetrahydrofolate reductase gene variants C677T and A1298C on the risk of orofacial clefts in 261 Norwegian case-parent triads.
Barker S, Chesney D, Sharp L, et al. Epidemiology of idiopathic congenital talipes equinovarus in the UK.
Frosst P, Blom HJ, Milos R, et al. A candidate genetic risk factor for vascular disease: a common mutation in methylenetetrahydrofolate reductase.
StataCorp. Stata statistical software, release 8.0. College Station, TX: Stata Corporation,
Wilcox AJ, Weinberg CR, Lie RT. Distinguishing the effects of maternal and offspring genes through studies of “case-parent triads.”
Weinberg CR, Wilcox AJ, Lie RT. A log-linear approach to case-parent-triad data: assessing effects of disease genes that act either directly or through maternal effects and that may be subject to parental imprinting.
Breslow NE, Day NE eds. Statistical methods in cancer research. Vol 1. The analysis of case-control studies. Lyon, France: International Agency for Research on Cancer,
Umbach DM, Weinberg CR. The use of case-parent triads to study joint effects of genotype and exposure.
Hobbs CA, Sherman SL, Yi P, et al. Polymorphisms in genes involved in folate metabolism as maternal risk factors for Down syndrome.
Risch N, Merikangas K. The future of genetic studies of complex human diseases.
Starr JR, Hsu L, Schwartz SM. Assessing maternal genetic associations. A comparison of the log-linear approach to case-parent triad data and a case-control approach.
Flanders WD, Khoury MJ. Analysis of case-parental control studies: method for the study of associations between disease and genetic markers.
Umbach DM. Invited commentary: on studying the joint effects of candidate genes and exposures.
Lissak A, Sharon A, Fruchter O, et al. Polymorphism for mutation of cytosine to thymine at location 677 in the methylenetetrahydrofolate reductase gene is associated with recurrent early fetal loss.
Murphy RP, Donoghue C, Nallen RJ, et al. Prospective evaluation of the risk conferred by factor V leiden and thermolabile methylenetetrahydrofolate reductase polymorphisms in pregnancy.
Dilley A, Benito C, Hooper WC, et al. Mutations in the factor V, prothrombin and MTHFR genes are not risk factors for recurrent fetal loss.
Unfried G, Griesmacher A, Weismuller W, et al. The C677T polymorphism of the methylenetetrahydrofolate reductase gene and idiopathic recurrent miscarriage.
Palmer RM, Conneally PM, Yu PL. Studies of the inheritance of idiopathic talipes equinovarus.
Bellyei A, Czeizel A. A higher incidence of congenital structural talipes equinovarus in gypsies.
Cartlidge I. Observations on the epidemiology of club foot in Polynesian and Caucasian populations.
Somppi E. Clubfoot. Review of the literature and an analysis of a series of 135 treated clubfeet.
Pryor GA, Villar RN, Ronen A, et al. Seasonal variation in the incidence of congenital talipes equinovarus.
Lochmiller C, Johnston D, Scott A, et al. Genetic epidemiology study of idiopathic talipes equinovarus.
Rebbeck TR, Dietz FR, Murray JC, et al. A single-gene explanation for the probability of having idiopathic talipes equinovarus.
Mathews F, Yudkin P, Neil A. Folate in the periconceptional period: are women getting enough?
Weinberg CR, Umbach DM. Choosing a retrospective design to assess joint genetic and environmental contributions to risk.
Halsted CH, Villanueva JA, Devlin AM, et al. Metabolic interactions of alcohol and folate.
Honein MA, Paulozzi LJ, Moore CA. Family history, maternal smoking, and clubfoot: an indication of a gene-environment interaction.
Rosenquist TH, Finnell RH. Genes, folate and homocysteine in embryonic development.
Vollset SE, Refsum H, Irgens LM, et al. Plasma total homocysteine, pregnancy complications, and adverse pregnancy outcomes: the Hordaland Homocysteine Study.
Karakurt L, Yilmaz E, Serin E, et al. Plasma total homocysteine level in mothers of children with clubfoot.
Ma J, Stampfer MJ, Giovannucci E, et al. Methylenetetrahydrofolate reductase polymorphism, dietary interactions and risk of colorectal cancer.
Ma J, Stampfer MJ, Christensen B, et al. A polymorphism of the methionine synthase gene: association with plasma folate, vitamin B12, homocyst(e)ine, and colorectal cancer risk.
Casas JP, Bautista LE, Smeeth L, et al. Homocysteine and stroke: evidence on a causal link from mendelian randomisation.
Jacques PF, Bostom AG, Williams RR, et al. Relation between folate status, a common mutation in methylenetetrahydrofolate reductase, and plasma homocysteine concentrations.
Girelli D, Friso S, Trabetti E, et al. Methylenetetrahydrofolate reductase C677T mutation, plasma homocysteine, and folate in subjects from northern Italy with or without angiographically documented severe coronary atherosclerotic disease: evidence for an important genetic-environmental interaction.
Kim KN, Kim YJ, Chang N. Effects of the interaction between the C677T 5,10-methylenetetrahydrofolate reductase polymorphism and serum B vitamins on homocysteine levels in pregnant women.
Brown KS, Kluijtmans LA, Young IS, et al. The 5,10-methylenetetrahydrofolate reductase C677T polymorphism interacts with smoking to increase homocysteine.
Papoutsakis C, Yiannakouris N, Manios Y, et al. Plasma homocysteine concentrations in Greek children are influenced by an interaction between the methylenetetrahydrofolate reductase C677T genotype and folate status.
Papoutsakis C, Yiannakouris N, Manios Y, et al. The effect of MTHFR(C677T) genotype on plasma homocysteine concentrations in healthy children is influenced by gender.
Spotila LD, Jacques PF, Berger PB, et al. Age dependence of the influence of methylenetetrahydrofolate reductase genotype on plasma homocysteine level.
van Beynum IM, den Heijer M, Thomas CM, et al. Total homocysteine and its predictors in Dutch children.
Thawnashom K, Tungtrongchitr R, Petmitr S, et al. Methylenetetrahydrofolate reductase (MTHFR) polymorphism (C677T) in relation to homocysteine concentration in overweight and obese Thais.
Chiuve SE, Giovannucci EL, Hankinson SE, et al. Alcohol intake and methylenetetrahydrofolate reductase polymorphism modify the relation of folate intake to plasma homocysteine.
McNulty H, McKinley MC, Wilson B, et al. Impaired functioning of thermolabile methylenetetrahydrofolate reductase is dependent on riboflavin status: implications for riboflavin requirements.
Luzzatto L, Falusi AO, Joju EA. Uracil in DNA in megaloclastic anemia.
Ames BN. DNA damage from micronutrient deficiencies is likely to be a major cause of cancer.
Blount BC, Mack MM, Wehr CM, et al. Folate deficiency causes uracil misincorporation into human DNA and chromosome breakage: implications for cancer and neuronal damage.
Jacob RA, Gretz DM, Taylor PC, et al. Moderate folate depletion increases plasma homocysteine and decreases lymphocyte DNA methylation in postmenopausal women.
Crott JW, Mashiyama ST, Ames BN, et al. Methylenetetrahydrofolate reductase C677T polymorphism does not alter folic acid deficiency-induced uracil incorporation into primary human lymphocyte DNA in vitro.
Andreassi MG, Botto N, Cocci F, et al. Methylenetetrahydrofolate reductase gene C677T polymorphism, homocysteine, vitamin B12, and DNA damage in coronary artery disease.
Zijno A, Andreoli C, Leopardi P, et al. Folate status, metabolic genotype, and biomarkers of genotoxicity in healthy subjects.
Narayanan S, McConnell J, Little J, et al. Associations between two common variants C677T and A1298C in the methylenetetrahydrofolate reductase gene and measures of folate metabolism and DNA stability (strand breaks, misincorporated uracil and DNA methylation status) in human lymphocytes in vivo.
Sharp L, Little J. Polymorphisms in genes involved in folate metabolism and colorectal neoplasia: a HuGE review.
Chen J, Giovannucci E, Kelsy K, et al. A methylenetetrahydrofolate reductase polymorphism and the risk of colorectal cancer.
Le Marchand L, Donlon T, Hankin JH, et al. B-vitamin intake, metabolic genes, and colorectal cancer risk (United States).
Ioannidis JP, Ntzani EE, Trikalinos TA, et al. Replication validity of genetic association studies.