Article Text
Abstract
Objectives The number of new oral cleft patients has fallen in the Netherlands. This may be explained by two hypotheses: (1) greater prenatal detection of congenital anomalies has led to more pregnancy terminations and (2) increased folic acid use has reduced the oral cleft risk. Both hypotheses would mainly apply to the category cleft lip/alveolus ± cleft palate (CL±P), since, unlike cleft palate only (CP), CL±P can be detected prenatally by two-dimensional (2D) ultrasound and develops during the period recommended for folic acid use. The authors aimed to determine trends in prevalence over 1997–2006 and to evaluate the hypotheses by stratifying trends by cleft category.
Methods This study was a time–trend analysis of infants born alive with oral clefts in the Netherlands during 1997–2006 and registered in the national oral cleft registry. The authors calculated prevalence rates and the estimated annual percentage change (EAPC) for all oral clefts and the two categories.
Results In 1997–2006, 3308 infants out of 1 970 872 live births had oral clefts, an overall prevalence per 10 000 live births of 16.8 (CL±P: 11.3; CP: 5.5). Time–trend analysis showed that the prevalence of all oral clefts decreased (EAPC −1.8%; 95% CI: –3.0% to −0.6%), as did the CL±P prevalence (EAPC −2.3%; 95% CI: −3.8% to −0.9%). No significant trends were found for the CP prevalence.
Conclusions Because the live-birth prevalence of CL±P decreased, that of all oral clefts decreased. These findings are in line with both hypotheses and may therefore have implications for prenatal counselling and folic acid policy.
Statistics from Altmetric.com
Introduction
Oral clefts – one of the most common birth defects in humans – range from mild types to complete clefts affecting the lip/alveolus/palate. They may either be isolated or be associated with other congenital anomalies, syndromes or chromosomal defects. Although the aetiopathogenesis of non-syndromic clefts has been widely studied, it is still poorly understood. Many genetic and environmental factors, such as nutrition and medication, have been suggested to contribute to their development.1,–,3
Since 1997, the 15 cleft palate teams in the Netherlands have reported their new live-born presurgery patients with oral clefts to the national oral cleft registry, which is maintained by the Dutch Association for Cleft Palate and Craniofacial Anomalies (NVSCA). These teams treat virtually all surviving children with clefts who reside in the Netherlands.4 5 The annual reports of this registry show that the number of cleft patients referred to the teams has fallen since 2003.6 This decline may be explained by two hypotheses. Firstly, the performance of prenatal 2D ultrasound examinations has increased since the 1990s, resulting in greater prenatal detection of congenital anomalies, including oral clefts,7 which has led to more terminations of affected pregnancies. Secondly, the government-sponsored mass media-campaign in 1995 and the proactive intervention by Dutch pharmacies in 2004 to promote the use of folic acid have increased the periconceptional use of this supplement,8 thereby reducing the risk of oral clefts. To investigate these hypotheses, oral clefts have to be divided into two categories: cleft lip/alveolus with or without cleft palate (CL±P) and cleft palate only (CP). These categories differ embryologically and epidemiologically2 and, unlike CP, the category CL±P can be detected prenatally using 2D ultrasound.9 This means that if pregnancies were terminated because of the presence of an oral cleft with or without associated anomalies, the CL±P prevalence would have been affected most. Additionally, most types of CL±P develop during 4–7 weeks postconception,3 10 which is during the recommended period for folic acid use (4 weeks before conception to 8 weeks after it). CP, however, develops during 8–12 weeks postconception,3 10 and may therefore be less influenced by a higher use of folic acid during the recommended gestational period in the population.
What is already known on this topic
▶ Prevalence data on oral clefts in the Netherlands have been incomplete for live births and stillbirths, and national time trends have not been established.
▶ Several studies suggest that decreases in oral cleft prevalence may be associated with the possible preventive effect of periconceptional folic acid use.
▶ The impact of prenatal diagnosis of oral clefts and/or associated anomalies leading to TOP might, however, be underestimated.
What this study adds
▶ During 1997–2006, the live-birth prevalence of oral clefts in the Netherlands decreased significantly, due to a decrease in CL±P.
▶ This preliminary analysis supports that this may be explained by more pregnancy terminations after prenatal diagnosis and/or reduced oral cleft risk after periconceptional folic acid.
▶ This would have implications for healthcare and policy makers, as evidence-based guidelines can optimise prenatal counselling, and increases in folic-acid exposure may produce further prevalence reductions.
To investigate whether the prevalence of oral clefts among live births in the Netherlands decreased over the 1997–2006 period, we used data from the NVSCA to establish the rates of oral clefts. By stratifying trends by cleft category, we were able to investigate whether the oral cleft prevalence may have been affected by the greater prenatal detection of CL±P with or without associated anomalies, and/or by the higher periconceptional use of folic acid.
Methods
The methodology by which the NVSCA registry was established is described elsewhere4 11 and is summarised here.
The NVSCA registry includes anonymous presurgery patients with oral clefts (no age limit) reported by the Dutch cleft palate teams. Using a unique recording system, detailed information on the topography and morphology of each anatomical structure of the head and neck area (eg, lip, alveolus and hard/soft palate) is recorded. Additionally, general information concerning the infant/parents and diagnoses of associated anomalies are included. Note that the form is completed during the first visit of the patient to the team, which is usually in the immediate postnatal period, and that there is no active follow-up of patients. To optimise data quality, data are verified on a case-by-case basis and the cleft palate teams perform case-ascertainment activities. Furthermore, the registry has been systematically validated.11
The study population included all infants born alive in the Netherlands from 1997 to 2006 who had been registered in the NVSCA database with an oral cleft (with or without associated anomalies). We excluded median and atypical facial clefts because of their different pathogenesis.12 13
The descriptive data of the study population (infant/maternal characteristics and information on associated anomalies) were presented as percentages, medians with IQR or means ± 1 SD.
We performed time–trend analyses to estimate the change in live-birth prevalence of all clefts and of the two categories (CL±P and CP) over the period 1997–2006. The prevalence was determined annually as the number of registered live-born infants with a cleft per 10 000 live births in the Netherlands for the same year. The annual numbers of live births in the Netherlands were retrieved from Statistics Netherlands.14
We used Poisson regression, modelling counts against year, to calculate the estimated annual percentage change (EAPC) in prevalence; this method has been previously described by De Vries and colleagues.15 The EAPC, presented with its 95% CI, represents an estimate of the year-to-year change in rate; negative values represent declining rates and positive values represent increasing rates. Statistics were performed using the SPSS v 17.0 software package.
Results
During the study period, 3308 infants were born alive with an oral cleft and were registered in the national database. The descriptive data of the study population are presented in table 1. The median age at cleft palate team admission was 0.6 months. Most infants were born to Caucasian parents (approximately 88%), after a normal gestational age of 39 weeks, and the mean birth weight was 3299 g. Almost 28% of the infants had a family history of congenital anomalies, about 11% had congenital anomalies of the head and neck area other than clefts and 13% had accompanying defects of other organ systems. In 10%, the oral cleft was part of a syndrome and/or chromosomal defect.
Between 1997 and 2006, there were 1 970 872 live births in the Netherlands; the average oral cleft prevalence was 16.8 per 10 000 live births. Figure 1 displays the prevalence over time. The annual prevalence of all clefts ranged from 14.5 to 18.6 per 10 000 live births (IQR: 15.8–17.7). Time–trend analysis showed that over the 1997–2006 period, the oral cleft prevalence decreased significantly by –1.8% per year (95% CI: –3.0% to –0.6%).
Infants with CL±P accounted for 67% of the study population (2218 infants), corresponding to an average prevalence of 11.3 per 10 000 live births. The prevalence of CP was 5.5 per 10 000 live births. The annual CL±P prevalence ranged from 9.6 to 12.9 per 10 000 live births (IQR: 10.5–12.0), and the annual CP prevalence ranged from 4.7 to 6.3 per 10 000 live births (IQR: 5.0–6.2). Time-trend analysis for CL±P showed a trend similar to that observed for all clefts; per year, the prevalence decreased by a significant –2.3% (95% CI: –3.8% to –0.9%). By contrast, for CP there was no evidence of a significant trend over time (EAPC –0.8%; 95% CI: –2.9% to 1.2%).
Discussion
Our study shows that the live-birth prevalence of oral clefts in the Netherlands decreased significantly during 1997–2006. By stratifying this trend by cleft category, we found a trend for CL±P similar to that observed for all clefts, while there was no significant trend for CP. This specific decrease in CL±P supports the hypothesis that the live-birth prevalence of oral clefts was affected by the greater prenatal detection of CL±P (with or without associated anomalies), and/or by the higher periconceptional use of folic acid.
Strengths and weaknesses
While the use of the NVSCA database was the main strength of our study – it allowed us to collect detailed national data on oral clefts among live births – it also had some limitations. The first is that registry databases are known to be prone to underreporting and misclassification. Our findings are nonetheless unlikely to be explained by underreporting or misclassification, because case-ascertainment is performed annually, and extra control activities were performed after the decline in new recorded patients. Furthermore, quality studies showed that the NVSCA registry has a high case-ascertainment and contains high quality data for both CL±P and CP.5 11
Another limitation is that the national database records only patients who are treated by the cleft palate teams. Stillbirths are therefore not included, and infants with severe associated anomalies who die during the first weeks of life might also not be captured. A change in perinatal/neonatal mortality could thus have affected our rates. Although we cannot rule out the impact of this factor, we suspect it is of minor importance given that the Dutch perinatal/neonatal mortality decreased during the study period.14 Besides, a change in these mortalities should have mainly affected the CP prevalence since further anomalies are more frequently associated with CP than with CL±P.1
Finally, the database does not provide complete and reliable data on associated anomalies (Rozendaal et al 2010; Unpublished data). The NVSCA has no active follow-up of patients, and therefore, associated anomalies detected later in infancy are often not included.16 17 For this reason, trends according to whether clefts were isolated or associated with other anomalies could not be given to further support that the increased prenatal detection accounted for the decline in prevalence.
Another source for Dutch oral cleft data is the EUROCAT registry of the Northern Netherlands.18 However, cleft rates of the Northern Netherlands are significantly higher than those of the Netherlands and can therefore not be used in a national context.19 20 Additionally, international registries show large variations in cleft prevalence without consistent time trends;1 21 22 the worldwide prevalence comes to 15.2 per 10 000 births.21 Unfortunately, comparison of our findings with those of other studies is restricted particularly due to the great differences between data sources, times of diagnosis, classifications, inclusion/exclusion criteria, time scales, sample sizes and population characteristics.1 16 23
Possible explanations
Our first hypothesis – that greater prenatal detection of congenital anomalies, including oral clefts, has led to more terminations of affected pregnancies – is a plausible explanation for the decline in oral clefts among live births in the Netherlands. This hypothesis is supported by our findings and additional national data.24
Since 2007, 2D ultrasound screening in the Netherlands has been routinely performed at 18–20 weeks gestation. However, the use of prenatal ultrasounds started to increase as early as the 1990s and, before 2007, over 90% of pregnant women residing in the Netherlands underwent one or more ultrasound scan.7 As it did elsewhere,9 22 25 26 the subsequent rise in prenatally detected anomalies, including oral clefts,7 may have led more affected pregnancies to be terminated in the Netherlands. The rates reported internationally for termination of pregnancy (TOP) on the basis of an isolated cleft range from 0% to 92%,26 but TOPs are performed more frequently when the cleft is associated with other anomalies.9 Unlike CP, the category CL±P can be detected prenatally using 2D ultrasound,9 23 which may explain why we found a significant decreasing trend for CL±P and no trend for CP.
This hypothesis is also supported by national data on pregnancy terminations.24 In the Netherlands, TOP is allowed only under the provision of the Termination of Pregnancy Act; after 24 weeks gestation it is prohibited.27 Under the terms of this act, all Dutch hospitals or clinics licensed to perform TOPs are required to report information relevant to these TOPs. Indications are not included. The annual reports submitted under this act show that the number of second-trimester terminations, especially those performed by the hospitals, have increased since 2003. This implies that there has been a rise in the TOPs affected with congenital anomalies, since these are performed mainly by the hospitals.24
Our second hypothesis – that higher periconceptional use of folic acid has reduced the risk of oral clefts in the Netherlands – is also plausible. Several studies have reported a significant protective association between periconceptional folic acid and oral cleft risk,28,–,32 but the evidence on the role of this supplement in cleft aetiology is still inconclusive.31 32 In the Netherlands, women are recommended to take 400 µg folic acid/day from 4 weeks before conception until 8 weeks thereafter, and the frequency of expecting mothers correctly using additional folic acid increased gradually over the past decade.8 33 Our study shows a gradual decline in CL±P prevalence, but not in CP prevalence, during approximately the same time frame. These findings are consistent with two other studies that investigated the same supplementation period,28 30 while countries with compulsory fortification (United States and Canada) showed a decline in both CL±P and CP.31
A possible explanation for these findings is that the recommended period for folic acid supplementation does not cover the aetiologically relevant time period for CP (8–12 weeks postconception).3 This theory is supported by the study of Bakker and colleagues,34 who showed that, after discontinuation of folic acid supplementation, the folate concentration (a general term for this B-vitamin) in serum immediately decreases and the plasma total homocysteine immediately increases. Given that there may be a dose–response relationship between folic acid and oral clefts, and that folate may be indirectly associated with clefts through its effects on homocysteine metabolism,28 31 35 folic acid supplementation until 8 weeks postconception might be too short to prevent CP.
Finally, other environmental or lifestyle factors (eg, dietary patterns) changing over time may also account for the decrease in cleft prevalence.1 Specific data on these factors are not available for the oral cleft population, but data based on the general Dutch population show that maternal smoking and alcohol consumption decreased during the study period.36 Since these factors are suggested to be associated with CL±P and CP risk,1 they may play contributory roles regarding the decrease in prevalence.
Possible implications
Our findings may have several implications for healthcare and policy makers. Firstly, TOP after prenatal diagnosis of congenital anomalies raise moral and ethical dilemmas, since most of these anomalies are nonlethal (eg, non-syndromic clefts). If oral clefts are identified prenatally, future parents should be counselled by a multidisciplinary cleft palate team that focuses on psychosocial support, genetic counselling, education on the management of clefts and parents' options, TOP being one of them.25 26 37 38 The Netherlands does not yet have a uniform strategy, but is developing an evidence-based guideline to optimise prenatal counselling on oral clefts.
Secondly, if folic acid protects against oral clefts, increases in pregnant women's exposure to folic acid and extension of the recommended period to at least 12 weeks postconception ought to produce further reductions in prevalence. Since folic acid consumption has been increased by food fortification,31 33 the possible effect on oral clefts is also relevant to the ongoing discussions about fortification.
Population-based studies including live births and stillbirths could give more insight into the causes of the decrease in oral clefts, especially if prenatally diagnosed cases are included. Since national data on prenatally diagnosed anomalies and indications of TOP are still lacking, uniform registration is needed to evaluate the epidemiological impact of prenatal ultrasound screening. Additionally, to further investigate the possible preventive effect of folic acid, future studies should focus on timing, duration, dose and intensity of use of folic acid as well as of folic-acid containing supplements.32 Further studies should also differentiate between the various cleft types within CL±P and CP, since these are related to different time frames and cell biological mechanisms in embryonic development.3 10 12 Therefore, a unique case–control study has been recently started in the Northern Netherlands, based on detailed data regarding the various cleft types and periconceptional folic acid/multivitamins from the NVSCA and EUROCAT registries.
Acknowledgments
We would like to express our gratitude to the cleft palate teams in the Netherlands, and the NVSCA and its board in particular. Without their enthusiasm and efforts, the NVSCA registry would not have succeeded.
References
Footnotes
-
Competing interests None.
-
Ethical approval Studies of anonymous data from health registries do not require ethical approval in the Netherlands, and the principles outlined in the Declaration of Helsinki were followed.
-
Provenance and peer review Not commissioned; externally peer reviewed.