Risk Factors of Cleft Lip and / or Palate in Japan

This paper described a case-control study on cleft lip and/or palate which aimed to disclose demographic and epidemiologic risk factors. Cases in this study were defined as the infants born with cleft lip and/or palate during the study period from April, 1978 to September, 1981, who visited two specified hospitals and lived in one of five cities. Controls were randomly selected from normal infants who were born in a major hospital in each city during the study period. Controls were matched to cases, with an allocation ratio of 1 : 1, for sex, birth order, residential area, and maternal age at birth to within one year. Routine demographic and numerous epidemiologic information were collected by direct interview of the study subject's mothers at home by one interviewer. Routine statistical analysis by odds ratio with 95% confidence interval was carried out on 194 cases, who comprized 55 cases with cleft lip alone, 87 with cleft lip and palate, and 52 with cleft palate alone, in comparison with 194 matched controls. Major findings are as follows: 1) A significantly elevated risk was associated with parental lower educational attainment, parental occupation of unskilled/ service workers, previous episode of artificial abortion, family history of cleft lip and/or palate, and maternal smoking habits at the first trimester, and maternal episodes at the first trimester of suffering from any diseases (common cold in particular), and of ingesting any drugs (cold remedies in particular). 2) A significantly reduced risk was linked with frequent maternal consumption at the first trimester of such animal proteins as meats, fishes and shells, eggs and milk. 3) Maternal episodes of spontaneous abortion/stillbirth, radiation exposure and frequent maternal consumption at the first trimester of fresh vegetables, fruits, Japanese tea, and black tea were not significantly associated. 4) Frequent maternal consumption of coffee at the first trimester significantly elevated the risk, but turned to be insignificant after statistical adjustment of maternal smoking habits. Causal implications were discussed on these epidemiologic factors which significantly enhanced or reduced the risk.Cleft lip and/or palate, Case-control study, Risk factors, Epidemiology, Japan


INTRODUCTION
Cleft lip and/or palate is a relatively infrequent disease, but relatively common among congenital malformations.
In Japan, the incidence rate is reported (1)(2)(3) to be approximately 0.2% among live-and stillborns, with an unaltered trend in 1952-1985 and with minor geographical variations (2) .Cleft lip and/or palate have not been epidemiologically investigated over years in Japan because of its rarity and the lack of epidemiology-oriented clinicians.About ten years ago, however, we undertook the first typical epidemiological investigation to disclose etiological clues to and epidemiological risk factors for this particular disease.Since then, no other epidemiological studies to explore risk factors for cleft lip and/or palate have been reported, to our best knowledge, and, therefore, we will here introduce our study on this special occasion of IEA scientific meeting in Asia-Pacific region, as one of typical epidemiological investigations in Japan.

MATERIALS AND METHODS
Our epidemiological investigation undertaken is a case-control study with the following study design.
Identification and eligibility of cases and controls: Cases are defined as the infants born with cleft lip and/or palate for the period from April 1978 to September 1981, who visited either Department of Oral Surgery, Nagoya University Hospital or Department of Plastic Surgery, Shizuoka Children Hospital, and whose residence is located in one of five cities (Nagoya, Kasugai, Kariya, and Toyohashi in Aichi Prefecture, and Shizuoka in Shizuoka Prefecture).Controls are the normal infants born in the same period as did the cases, who are matched to cases with an allocation ratio of 1 : 1 for sex, birth order, residential area, and maternal age at birth to within one year.Controls are randomly selected from normal infants born in a major hospital in each area (Nagoya University Hospital, Kasugai Municipal Hospital, Kariya General Hospital, National Toyohashi Hospital,and Shizuoka Saiseikai Hospital).
Collection of epidemiological information: Routine demographic and epidemiological information were collected by direct interview of the study subject's mothers at home, using a questionnaire specifically designed for this study.Interviewer was a doctor of Oral Surgery (I.T.), and interviewees were the subject's mother, as described above, without any proxies.Information obtained by direct interview were 1) demographic and obstetrical information on infants: sex, date and place of birth, gestational age, fetal presentation at delivery, type of delivery, abnormalities of amniotic fluid, premature rupture of the membrane, height and weight at birth, and circumference of the head and the chest at birth, 2) maternal information at the first trimester of the current pregnancy: gestational week at being aware of the present pregnancy, diseases suffered and drugs ingested, radiation exposures, smoking habits, drinking habits of alcoholic beverages, and dietary practices of selected foodstuffs, 3) parental information: school attended as the highest educational attainment, occupation and jobs, age at marriage, age at the current pregnancy, difference in parental age, places of birth, height and weight before the current pregnancy, consanguineous marriage in parent, family history of cleft lip and/or palate, and maternal episodes of spontaneous and artificial abortions, stillbirths and live-births, and 4) other relevant informations: numbers of family members, type of living quarters, pets breeding, kind of drinking waters, and so forth.Some of these epidemiological information and certain obstetrical information were also abstracted from in/out patient medical records and Maternal and Child Health Handbook (Boshi-techo in Japanese).
Subjects to be analysed: During the study period of 3.5 years, we could identify 252 infants with cleft lip and/or palate in the two hospitals.Among them, however, 194 infants were successfully matched to control infants, and comprized 55 infants with cleft lip, 87 with cleft lip and palate, and 52 with cleft palate.
Statistical analysis: An association of each epidemiological variable with development of cleft lip and/or palate was examined by odds ratio routinely calculated in a basic table construction (2 2 table), and statistical significance of odds ratio was evaluated by its 95% confidence interval; being significant when the interval excludes 1.

RESULTS AND COMMENTS
The results obtained from this study were reported in details elsewhere(4-6) with full discussion.Here we will briefly describe epidemiologic risk factors, chiefly based on the previous paper (3) .Table 1 demonstrates the frequency distribution of study subjects by matching variable.Residential area, sex, and birth order were perfectly matched.Maternal age at birth was also satisfactorily matched with no significant difference in mean age between cases and controls.
Table 2 presents odds ratios with 95% confidence interval by parental education and occupation.A significantly increased risk of developing CL/P is associated with the lower paternal educational attainment; an odds ratio being significantly 2.32.Type-specific risk is significant for CLP (odds ratio = 2.54) and CP as well (2.75), but is not for CL.This pattern of association is exactly the same by maternal educational attainment with the similar magnitudes of risk; odds ratios being 2.56 for CLP, 3.31 for CP, and 2.47 for CL/P.Paternal occupation of category III. is at greater risk of CL, CLP, and CL/P, when compared to all other occupations; odds ratios being 2.59, 2.75, and 2.17, correspondingly.This is also exactly the case by maternal occupational category, to which all mothers, who was continuously working and who quitted their jobs when they became aware of being pregnant as well, were classified.When compared to mothers with all other occupations plus mothers with no jobs, the risk is significant for CL (odds ratio = 3.85) and almost so for CL/P.In short, the risk is significantly associated with lower parental educational attainment and lower-waged occupations, i.e., an association with lower socioeconomic status.This finding, however, is not deemed to mean the direct risk modification by themselves, since education and occupation are related each other and are the major determinants of such life-styles as dietary practices, personal hygiene, smoking habits, drinking habits,and so forth, which are the actual and direct determinants of disease risks.Table 3 indicates odds ratios with 95% confidence interval by maternal episodes of previous pregnancy and family history of cleft lip and/or palate.Artificial abortion is significantly associated with the risk of developing CLP (odds ratio = 2.30) and CL/P (1.96), but not the episodes of spontaneous abortion or stillbirth.This association is also not presumed to mean the direct risk modification by the episode of artificial abortion itself, since whether artificial abortion is to be performed or not is mostly decided by the economic reasons of the households or some other reasons for family planning in Japan.Family history of CL/P demonstrated significantly large odds ratios of 12.00 for CL, 5.37 for CLP, and 4.57 for CL/P, though 95% confidence intervals were quite wide.The wide confidence intervals were simply resulted from the fact that a few study subjects had the history and most study subjects had not.This association undoubtedly indicates the genetic relevance in the development of CL/P.  4 shows odds ratios and 95% confidence interval by maternal smoking habits and radiation exposure at the first trimester.Mothers with smoking habits in this table included those who were smoking at the first trimester and those who quitted smoking when they became aware of being pregnant.Overall frequency of smoking habits were 43 out of 194 mothers with cases (22.2%) and 21 out of 194 control mothers (10.8%), which rendered the odds ratio of 2.35 (95% confidence interval: 1.35-4.09)for CL/P.Type-specific odds ratios are 8.20 and 2.95 for CL and CLP, respectively, with statistical a significance.Maternal smoking is a well known risk factor for various unfavourable pregnancy outcomes and some congenital malformations.
Our finding is presumed to confirm this relationship.Maternal drinking habits at the first trimester were also examined in this study.Mothers with drinking habits (those who drink alcoholic beverages either regularly or occasionally at the first trimester and those who quitted drinking when they became aware of being pregnant) were accounted for 8.9% in cases and 9.2% in controls; resulting in an insignificant odds ratio.Maternal episode of radiation exposure to any parts of body at the first trimester was found in 9.8% among cases (19 out of 194) and 6.2% in controls (12 out of 194), which gave an insignificant odds ratio for CL/P, CL, CLP, and CP.
Table 5 summarizes odds ratios with 95% confidence intervals by maternal episode of disease suffered at the first trimester.Mothers who suffered from any diseases at the first trimester are at significantly greater risk of developing CLP, CP, and CLIP; the magnitude of the risk being more than two fold, respectively.Among four major diseases which are most likely to be suffered from at the first trimester, "common cold" alone demonstrated significantly larger odds ratios of 2.88, 4.08, and 2.70 for CLP, CP, and CL/P, correspondingly.
Suffering from any diseases, particularly common cold, may possibly indicate somewhat direct relevance to the development of CL/P through infectious agents, but more plausible explanations may probably be the relevance of drugs ingested for treatment and/or the indirect relevance of some maternal physiological conditions at the first trimester which led to the diseases.This finding, however, could not be presumed to provide any possible causal relationships between a specific disease and CL/P, since the present indicator of "any diseases or common cold" is too crude to discuss any specific risk modification.Table 6 describes the risk modification by maternal ingestion of drugs at the first trimester.Maternal ingestion of any drugs was found to increase significantly the risks for CLP, CP, and CL/P; the magnitude of risk being 2.4-2.8.Among drugs ingested cold remedies are only drugs which significantly increased the risk by 3.7-6.7.Hematopoietics and gastrointestinal agents are two other drugs which suggestively increased the risk of CL/P.Causal implications of ingesting any drugs, cold remedies in particular, are as same as the implications discussed above on suffering from any diseases, common cold in particular.
Table 7 and 8 summarizes the odds ratios by some foodstuffs and usual beverages consumed by mothers at the first trimester.Frequent maternal intakes of meats, and fishes and shells significantly reduced the risk of developing CL/P (odds ratio = 0.38, 0.40) and CL (0.15, 0.08).Risk reduction is shown to be more substantial for CL than for CL/P.Frequent maternal consumption of eggs also significantly reduced the risk of CL/P (odds ratio = 0.33) and CLP (0.21).Of interest in Table 7 is that other insignificant odds ratios by meats, fishes and shells and eggs are all less than unity.Fresh vegetables and fruits were not linked with the risk of developing CL/P.
In Table 8, Japanese tea and black tea were not at all associated with the risk, but frequent consumption of milk significantly reduced the risk of CLP and CL/P; odds ratios being 0.37 and 0.45, respectively.Frequent maternal drinking of coffee was found to be at greater risk of developing CL and CL/P, but turned to be not significantly associated with when maternal smoking habits were adjusted.In short, from Table 7 and 8, significant risk reduction is apparent by frequent maternal consumption of such animal proteins as meats, fishes and shells, eggs, and milk.

Table 1
Distribution of Study Subjects by Matching Variable

Table 4
Odds Ratio and Its 95% Confidence Interval by Maternal Habits or Radiation Exposure at First Trimester

Table 6
Odds Ratio and Its 95% Confidence