- Email: [email protected]
Paternal age and the risk of birth defects in Norway
Paternal Age and the Risk of Birth Defects in Norway METHOD KAZAURA, MSC, ROLV T. LIE, PHD, AND ROLV SKJÆRVEN, PHD
PURPOSE: We studied 1,869,388 births from The Medical Birth Registry of Norway to assess the effect of father’s age on risks of birth defects in offspring. METHOD: Thirteen separate categories were studied including pooled categories of neural tube defects and any type of defect. We used logistic regression models to adjust for maternal age, year of birth, maternity institution, parity, and correlation between siblings. RESULTS: There was little evidence of increased risk by high paternal age for any category of defects, except for a category of “other central nervous system” where risk estimates were 2.5-fold (95% CI: 1.2– 5.5) for fathers aged between 45 and 49 years compared with the reference age group (25–29 years). The risk for neural tube defects was 1.3-fold (95% CI: 1.1–1.5) when the father was aged between 20 and 24 years relative to the reference. A pattern of moderately higher risks for younger fathers was consistent for anencephaly and spina bifida. Increased risk of heart defects was also estimated among children of young fathers. CONCLUSIONS: This study does not show consistent evidence that paternal ageing is a risk for birth defects among offspring. Low paternal age, or factors associated with younger parents, may however be associated with increased risk of neural tube defects in their offspring. Ann Epidemiol 2004;14:566–570. 쑕 2004 Elsevier Inc. All rights reserved. KEY WORDS:
Birth Defects, Paternal Age, Risk, Norway.
INTRODUCTION Epidemiological studies of birth defects have primarily focused on maternal factors including environmental exposures within the first trimester of pregnancy. There are, however, theoretical reasons to also study the effects of including high paternal age on the risk of birth defects (1). With age, mutations may accumulate in the spermatogonia during the continuous process of cell division (2). An increased number of mutations in the sperm of older fathers may subsequently increase the number of birth defects in their offspring. One study has indicated that fathers aged 40 years and above have a two-fold risk relative to young fathers of having children with hydrocephaly (3). High paternal age was also found to contribute to cleft lip and cleft palate and to defects
From the Centre for International Health, University of Bergen, Bergen, Norway (M.K.); Section for Medical Statistics, Department of Public Health and Primary Health Care, University of Bergen, Bergen, Norway (M.K., R.T.L., R.S.); and Medical Birth Registry of Norway, Locus of Registry Based Epidemiology, University of Bergen, Bergen, Norway (R.T.L., R.S.). Address correspondence to: Method R. Kazaura, Centre for International Health, Armauer Hansen Building, N-5021, Bergen, Norway. Tel: ⫹475558-8518; Fax: ⫹47-5597-4979. E-mail: [email protected] Received April 30, 2003; accepted October 13, 2003. 쑕 2004 Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010
of nails. Other studies have indicated an effect of father’s age for other birth defects (4–6). A recent study found an elevated risk among older fathers for the combined category of neural tube defects, for hydrocephaly and spina bifida, congenital cataracts, and reduction of upper limb defects (7). The same study reported increased risks among offspring of young fathers (⬍20 years) for neural tube defects, anencephaly, and hydrocephaly. Studies from New York and Beijing indicated an association of young paternal age with congenital heart defects (6, 8). The age of the father and the mother are highly correlated. Careful adjustment is therefore needed to separate the effects of the two if there are effects of maternal age (9). There are several examples of such adjustments (10, 11). Mother’s age is a well-known risk factor for chromosomal aberrations. We previously studied the effect of paternal age on Down’s syndrome and gastroschisis (including omphalocele) separately (12, 13). We did not find conclusive evidence of an association between paternal age and Down’s syndrome. However, there were some indications that young paternal age increased the risk of gastroschisis. Using a large population-based data set we analyze here a broad spectrum of other defects categories to look for effects of paternal age and possible consistencies with previous studies. We adjusted the analyses for maternal age, year of birth of the child, parity, and place of birth. 1047-2797/04/$–see front matter doi:10.1016/j.annepidem.2003.10.003
AEP Vol. 14, No. 8 September 2004: 566–570
Selected Abbreviations and Acronyms CI ⫽ confidence interval CNS ⫽ central nervous systems OR ⫽ odds ratio
MATERIALS AND METHODS The Medical Birth Registry of Norway is a populationbased registry covering all births since 1967. Norwegian law requires notification to the registry of all pregnancies exceeding 16 weeks within the first week after the birth. By 1998, there were a total of 1,869,388 births recorded. In this study, the following categories of birth defects, classified according to The International Classification of Diseases (ICD-8) were included: Anencephaly, spina bifida, hydrocephaly, other CNS, ear/face/neck, heart, circulatory, respiratory, total cleft lip, anal, renal, limb defects, and cleft palate defects. We also included a category for “any birth defect” that accounts for a child with any of the above categories of birth defects. We used major categories of malformations as described in previous studies (14–16). There is considerable variation in the ascertainment of birth defects in the Medical Birth Registry of Norway. By comparing with independently collected data, it has been estimated that as much as 90% of cases with neural tube defects and 80% with cleft lip are ascertained (17). Ascertainment by the Medical Birth Registry of Norway for internal defects like heart defects is probably more incomplete. For 7% (130,536) of the children, father’s identity was unknown and paternal age therefore unknown. These births were excluded from the analysis. Logistic regression models were used to estimate the effect of paternal age measured by odds ratios (ORs). Separate analyses were performed for the different categories of birth defects. Both maternal and paternal ages were computed as age in days but scaled to completed years. Paternal age was first treated as a continuous variable in the analyses to assess a linear trend in risk of birth defects with advancing paternal age and then also categorized into 5-year intervals as shown in the tables. Adjustment was made in the logistic models for several potential confounders, most importantly for a possible effect of maternal age (18). To minimize the possibility of residual confounding within broader categories of maternal age, maternal age was used as a continuous regression variable in the models. Since an effect of maternal age could have a non-linear relationship with risk, we also entered secondand third-degree terms of maternal age in the models for each category of birth defects. We also fitted models with independent first- and second-order terms of maternal age
Kazaura et al. PATERNAL AGE AND BIRTH DEFECTS
567
above and below 30 years as an alternative mode of adjustment for all categories of birth defects (12). Since maternity institutions across Norway have slightly different procedures of diagnosis and reporting of birth defects, place of birth was also adjusted for. Five categories of institutions were distinguished in the adjustment: 1) Ulleval and Aker hospitals; 2) Haukeland Hospital; 3) Trondheim regional hospital; 4) Stavanger central hospital; and 5) all of the remaining institutions. These specific hospitals cover the largest regional hospitals in Norway. It was also expected that the prevalence of various birth defects has also changed over time partly because of diagnostic improvements (e.g., ultrasound). We also adjusted for child’s year of birth. These birth periods were categorized into 5-year intervals, with the exception of the first (1967– 1970) and the last category (1996–1998). Previous studies also controlled for paternal race (3, 6, 7). Racial disparity was, however, considered negligible in Norway. Many children born to older fathers are higher parity births. Adjustment for parity was therefore also performed. All sibships can be identified in our registry, and we adjusted for the correlation between siblings using a method of robust calculation of confidence intervals and p-values in logistic regression models. We applied each of the categories of birth defects as an outcome variable in the binary logistic regression model with all analyses performed using the statistical software STATA (19).
RESULTS The overall birth prevalence of ascertained serious birth defects (excluding Down’s syndrome and all known chromosomal anomalies) in Norway was 2.5%. There was no clear overall pattern of increased risk by increasing paternal age. No effects were observed for maternal age (Table 1). TABLE 1. Prevalence of birth defectsa at birth by parental age groups in Norway, 1967–1998 Prevalence by fathers age
Prevalence by mothers age
Age group (years)
Cases
Percent
Cases
Percent
⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹ Total
253 5862 14392 12675 6287 2336 753 255 42813
2.44 2.29 2.45 2.54 2.51 2.47 2.45 2.33 2.46
3026 13522 16502 9705 3414 628 27 46824
2.34 2.43 2.55 2.60 2.55 2.47 2.02 2.51
a
Excluding Down’s syndrome and other known chromosomal aberrations.
568 Kazaura et al.
AEP Vol. 14, No. 8 September 2004: 566–570
PATERNAL AGE AND BIRTH DEFECTS
Using logistic regression models, we estimated the effect of father’s age adjusting for mother’s age, parity, maternity institution, and birth year of the child. Only minor differences were seen across all defect categories in estimates of the effect of father’s age depending on mode of adjustment for mother’s age. We have therefore tabulated effects estimated from models adjusting for first- and second-degree terms of mother’s age in Tables 2, 3, and 4. We found the highest risk (OR ⫽ 2.5; 95% CI: 1.2–5.5) among fathers aged between 45 and 49 years compared with those aged 25 to 29 years, for a category of “other central nervous systems” (CNS) defects (Table 2). Although there was no clear monotonic pattern of risks for “other CNS” defects with advancing paternal age, a test for linear trend across the whole range of paternal age was significant (p ⬍ 0.05). We found similar results of elevated risks for respiratory, renal, and ear/face/neck defects with advancing father’s age (Tables 2, 3, and 4). A test for linear trend showed that only the ear/face/neck and renal defects categories were positively associated with paternal age (p ⬍ 0.05). However, the average risk of these defects within paternal age groups was weak. There were indications of an association between low paternal age (less than 25 years) and risks of neural tube defects and heart birth defects. For example, we estimated that fathers aged between 20 and 24 years had a 1.4-fold (95% CI: 1.1–1.8) risk of fathering a child with anencephaly as compared with fathers aged between 25 to 29 years. Although not significant, a similar pattern of risks was seen for spina bifida.
DISCUSSION The current study was aimed at assessing the effect of paternal age on the risk of selected categories of birth defects. We found little or no evidences for such effect. Unlike with one previous study (7) that found increasing risks for a combined category neural tube (and weak effects for anencephaly and spina bifida categories) with advancing paternal age, such patterns were not consistent in our study. We found an indication of an effect of advanced paternal age on broad category of birth defects categorized as “other CNS” whose major subcategories include encephalocele, microcephaly, and other defects of the brain, spine, or CNS. This association has never been reported in other similar studies. The association would therefore need confirmation from subsequent studies. In addition, the observed positive linear trend of paternal age with categories of ear/ face/neck and renal defects would require further attention from other similar studies. We did not find paternal age
TABLE 2. Relative risks for selected birth defects by paternal age in Norway, 1967–1998 Paternal age (years) and categories of birth defects Neural tube defects ⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹ Anencephaly ⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹ Spina bifida ⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹ Hydrocephaly ⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹ Other CNS ⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹
Number of cases
Risk (per 10,000 births)
Adjusted odds ratioa
95% confidence intervalb
12 222 447 347 166 78 19 12
11.6 8.7 7.6 7.0 6.6 8.3 6.2 11.0
1.4 1.3 1.0 0.9 0.8 0.9 0.8 0.9
0.9–2.5 1.1–1.5 Reference 0.8–1.0 0.6–1.0 0.7–1.2 0.5–1.3 0.4–1.9
4 93 183 133 63 27 5 4
3.9 3.6 3.1 2.7 2.5 2.9 1.6 3.7
1.4 1.4 1.0 0.8 0.8 0.8 0.3 1.3
0.6–3.4 1.1–1.8 Reference 0.6–1.0 0.6–1.1 0.5–1.3 0.1–1.0 0.4–3.7
8 129 264 214 103 51 14 8
7.7 5.0 4.5 4.3 4.1 5.4 4.5 7.3
1.4 1.2 1.0 0.9 0.8 0.9 1.0 0.7
0.8–2.8 1.0–1.5 Reference 0.8–1.1 0.6–1.0 0.6–1.4 0.6–1.8 0.2–2.0
5 65 185 183 96 33 12 4
4.8 2.5 3.2 3.7 3.8 3.5 3.9 3.6
1.2 0.8 1.0 1.0 1.1 0.8 1.2 0.5
0.6–2.6 0.6–1.0 Reference 0.8–1.2 0.8–1.5 0.5–1.3 0.6–2.2 0.1–2.0
3 52 98 75 63 16 11 4
2.9 2.0 1.7 1.5 2.5 1.7 3.6 3.7
1.8 1.4 1.0 0.9 1.4 1.1 2.5 2.6
0.6–4.9 1.0–2.0 Reference 0.7–1.3 0.9–2.0 0.6–1.9 1.2–5.5 1.0–7.1
a
Adjusted for mother’s age, parity, maternity institution, and birth year of the child. Using robust estimation of variances accounting for correlation between siblings.
b
effects for some previously reported birth defects like hydrocephaly, spina bifida, and limb defects (3, 7). A study conducted in British Columbia also reported association of young paternal age with the risk of anencephaly, spina bifida, and the combined category of neural tube defects (7). This is consistent with indications in our study that these birth defects may indeed have a weak association
AEP Vol. 14, No. 8 September 2004: 566–570
Kazaura et al. PATERNAL AGE AND BIRTH DEFECTS
569
TABLE 3. Relative risks for selected birth defects by paternal age in Norway, 1967–1998
TABLE 4. Relative risks for selected birth defects by paternal age in Norway, 1967–1998
Paternal age (years) and categories of birth defects
Paternal age (years) and categories of birth defects
Ear/Face/Neck ⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹ Heart ⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹ Circulatory ⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹ Respiratory ⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹ Isolated cleft palate ⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹
Number of cases
Risk (per 10,000 births)
Adjusted odds ratioa
95% confidence intervalb
4 127 292 271 168 63 19 9
3.9 5.0 5.0 5.4 6.7 6.7 6.2 8.2
0.7 0.9 1.0 1.2 1.3 1.2 1.4 1.1
0.3–1.5 0.7–1.2 Reference 1.0–1.4 1.0–1.6 0.9–1.7 0.9–2.2 0.5–2.5
23 434 1093 1143 601 258 76 28
22.2 16.9 18.6 22.9 24.0 27.3 24.7 25.6
1.5 1.0 1.0 1.1 1.0 1.1 1.0 1.1
1.0–2.0 0.9–1.2 Reference 1.0–1.2 0.9–1.2 0.9–1.3 0.8–1.3 0.8–1.7
5 71 218 273 132 44 12 9
4.8 2.8 3.7 5.5 5.3 4.7 3.9 8.2
1.7 1.0 1.0 1.2 0.9 1.0 0.9 1.8
0.8–3.6 0.7–1.3 Reference 1.0–1.5 0.7–1.2 0.7–1.4 0.5–1.7 0.9–3.8
3 109 320 345 164 58 21 5
2.9 4.3 5.5 6.9 6.6 6.1 6.8 4.6
0.9 1.1 1.0 1.1 1.0 1.2 1.2 0.7
0.4–2.1 0.9–1.3 Reference 0.9–1.3 0.8–1.3 0.9–1.6 0.7–2.0 0.3–2.0
9 107 298 251 129 59 22 8
8.7 4.2 5.1 5.0 5.6 6.2 7.2 7.3
1.4 0.8 1.0 1.0 0.8 1.1 1.1 1.1
0.6–2.5 0.6–1.0 Reference 0.8–1.2 0.7–1.1 0.8–1.6 0.6–1.8 0.5–2.5
Total cleft lip ⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹ Anal ⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹ Renal ⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹ Limb ⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹ Any birth defectc ⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹
a
a
b
b
Adjusted for mother’s age, parity, maternity institution, and birth year of the child. Using robust estimation of variances accounting for correlation between siblings.
with young paternal age. We also found a similar effect for heart defects. For the combined category of neural tube defects, because older fathers (age 30 and above) in age groups of 5-year intervals had almost equal average risks (Table 2), we pooled them to make a reference group. The increased risk in this
Risk (per 10000 births)
Adjusted odds ratioa
95% confidence intervalb
14 348 815 635 327 143 57 15
13.5 13.6 13.9 12.7 13.1 15.1 18.5 13.7
1.2 1.0 1.0 0.9 0.9 1.0 1.2 0.9
0.8–1.8 0.9–1.1 Reference 0.8–1.0 0.8–1.1 0.8–1.3 0.9–1.6 0.5–1.5
7 108 283 225 119 59 16 5
6.7 4.2 4.8 4.5 4.8 6.2 5.2 4.6
1.4 0.9 1.0 0.8 0.8 1.0 0.6 0.6
0.7–2.6 0.7–1.1 Reference 0.7–1.0 0.6–1.1 0.7–1.4 0.3–1.2 0.2–1.6
4 111 296 340 201 79 20 10
3.9 4.3 5.0 6.8 8.0 8.4 6.5 9.1
1.1 1.1 1.0 1.1 1.2 1.4 1.3 1.6
0.5–2.4 0.9–1.4 Reference 0.9–1.3 0.9–1.4 1.0–1.9 0.8–2.0 0.8–3.2
61 1463 3726 3400 1714 620 210 61
58.7 57.1 63.5 68.2 68.4 65.6 68.2 55.7
0.9 1.0 1.0 1.0 1.0 1.0 1.0 0.9
0.7–1.1 0.9–1.0 Reference 1.0–1.1 1.0–1.1 0.9–1.2 0.9–1.2 0.6–1.1
253 5862 14392 12675 6287 2336 753 255
2.4 2.3 2.5 2-5 2.5 2.5 2.5 2.3
1.0 1.0 1.0 1.0 1.0 1.1 1.1 0.9
0.9–1.2 0.9–1.0 Reference 1.0–1.0 1.0–1.1 1.0–1.1 1.0–1.2 0.8–1.0
Number of cases
Adjusted for mother’s age, parity, maternity institution, and birth year of the child. Using robust estimation of variances accounting for correlation between siblings. Excluding Down’s syndrome and other known chromosomal aberrations.
c
category of birth defects was more evident. For example, very young fathers (less than 20 years) had a 1.7-fold (95% CI: 1.0–2.9) increased risk of a child with neural tube defect compared with at least 30 years old. One of the main advantages of this study is that we used a large data set that produced reasonably stable estimates
570 Kazaura et al. PATERNAL AGE AND BIRTH DEFECTS
of effects even for rare categories of defects. The Medical Birth Registry of Norway, like other population-based registries, does not have complete ascertainments for key categories of defects (for example, neural tube defects). Some cases of birth defects are diagnosed too late and therefore not recorded and some diagnoses are simply missed (17). A general misclassification of defects should therefore produce too low estimates of effects. It is however not likely that under-reporting of birth defects would depend on father’s age. We had little opportunity to assess whether effects of father’s age were missed because of such misclassification of cases. We excluded 7% of births because the age the father was unknown. For a small proportion of births the information on father’s identity and subsequently his age is likely to be wrong. The effect of such errors is likely to be small. A series of previously considered biological, socio-economical, and environmental confounding factors like smoking and the use of medications were not linked in the present data. Smoking has been repeatedly reported as a risk factor for several malformations, for example limb defects, clubfoot, cleft lip and palate (20–24). This could potentially explain some differences between our results and of the previous studies. Our observed effects of young paternal age could, for example, be due to confounding from specific lifestyle factors. In conclusion, our data do not show consistent evidence that the risk of birth defects in general increases with increased paternal age. Except for a mixed category of “other CNS” defects, we found no indications of increased risks by high paternal age for any specific category of defects. Slightly increased risks for neural tube defects and heart defects among offspring of young fathers may, if real, be caused by lifestyle factors that are more prevalent among young couples.
REFERENCES 1. Trasler JM, Doerksen T. Paternal exposures–reproductive risks. Teratology. 1999;60:161–172. 2. Penrose LS. Parental age and mutation. Lancet. 1955;2:312–313. 3. Lian ZH, Zack MM, Erickson JD. Paternal age and the occurrence of birth defects. Am J Hum Genet. 1986;9:648–660.
AEP Vol. 14, No. 8 September 2004: 566–570
4. Harlap S, Paltiel O, Deutsch L, Knaanie A, Masalha S, Tiram E, et al. Paternal age and preeclampsia. Epidemiology. 2002;13:660–667. 5. Olshan AF, Schnitzer PG, Baird PA. Paternal age and the risk of congenital heart defects. Teratology. 1994;50:80–84. 6. Polednak AP. Paternal age in relation to selected birth defects. Hum Biol. 1976;48:727–739. 7. McIntosh GC, Olshan AF, Baird PA. Paternal age and the risk of birth defects in offspring. Epidemiology. 1995;6:282–288. 8. Zhan SY, Lian ZH, Zheng DZ, Gao L. Effect of fathers’ age and birth order on occurrence of congenital heart disease. J Epidemiol Community Health. 1991;45:299–301. 9. Stene J, Stene E. Statistical methods for detecting a moderate paternal age effect on incidence of disorder when a maternal one is present. Ann Hum Genet. 1977;40:343–353. 10. Erickson JD, Bjerkedal TO. Down syndrome associated with father’s age in Norway. J Med Genet. 1981;18:22–28. 11. Martinez FML, Rodriguez PE, Prieto L. Prenatal exposure to salicylates and gastroschisis: A case–control study. Teratology. 1997;56:241–243. 12. Kazaura MR, Lie RT. Down’s syndrome and paternal age in Norway. Paediatr Perinat Epidemiol. 2002;16:314–319. 13. Kazaura MR, Lie RT, Irgens LM, Didriksen A, Kapstad M, Egenaes J, et al. Increasing risk of gastroschisis in Norway: An age-period-cohort analysis. Am J Epidemiol. Submitted. 14. Melve KK, Skjærven R. Families with birth defects: Is birth weight of nonmalformed siblings affected? Am J Epidemiol. 2002;155:932–940. 15. Lie RT, Wilcox AJ, Skjærven R. Survival and reproduction among males with birth defects and risk of recurrence in their children. JAMA. 2001;285:755–760. 16. Lie RT, Wilcox AJ, Skjaerven R. A population-based study of the risk of recurrence of birth defects. N Engl J Med. 1994;331:1–4. 17. Lie RT, Heuch I, Irgens LM. Maximum likelihood estimation of the proportion of congenital malformations using double registration system. Biometrics. 1994;50:433–444. 18. Erickson DJ. Paternal age and Down syndrome. Am J Hum Genet. 1979; 31:489–497. 19. STATA Corporation. STATA User’s Guide. Texas: Stata Press; 2001. 20. Wasserman CR, Shaw GM, O’Malley CD, Tolarova MM, Lammer EJ. Parental cigarette smoking and risk for congenital anomalies of the heart, neural tube, or limb. Teratology. 1996;53:261–267. 21. Chung KC, Kowalski CP, Kim HM, Buchman SR. Maternal cigarette smoking during pregnancy and the risk of having a child with cleft lip/ palate. Plast Reconstr Surg. 2000;105:485–491. 22. Kallen K. Maternal smoking and orofacial clefts. Cleft Palate Craniofac J. 1997;34:11–16. 23. Van den Eeden SK, Karagas MR, Daling JR, Vaughan TL. A case–control study of maternal smoking and congenital malformations. Paediatr Perinat Epidemiol. 1990;4:147–155. 24. Zhang J, Savitz DA, Schwingl PJ, Cai WW. A case–control study of paternal smoking and birth defects. Int J Epidemiol. 1992;21:273–278.
PURPOSE: We studied 1,869,388 births from The Medical Birth Registry of Norway to assess the effect of father’s age on risks of birth defects in offspring. METHOD: Thirteen separate categories were studied including pooled categories of neural tube defects and any type of defect. We used logistic regression models to adjust for maternal age, year of birth, maternity institution, parity, and correlation between siblings. RESULTS: There was little evidence of increased risk by high paternal age for any category of defects, except for a category of “other central nervous system” where risk estimates were 2.5-fold (95% CI: 1.2– 5.5) for fathers aged between 45 and 49 years compared with the reference age group (25–29 years). The risk for neural tube defects was 1.3-fold (95% CI: 1.1–1.5) when the father was aged between 20 and 24 years relative to the reference. A pattern of moderately higher risks for younger fathers was consistent for anencephaly and spina bifida. Increased risk of heart defects was also estimated among children of young fathers. CONCLUSIONS: This study does not show consistent evidence that paternal ageing is a risk for birth defects among offspring. Low paternal age, or factors associated with younger parents, may however be associated with increased risk of neural tube defects in their offspring. Ann Epidemiol 2004;14:566–570. 쑕 2004 Elsevier Inc. All rights reserved. KEY WORDS:
Birth Defects, Paternal Age, Risk, Norway.
INTRODUCTION Epidemiological studies of birth defects have primarily focused on maternal factors including environmental exposures within the first trimester of pregnancy. There are, however, theoretical reasons to also study the effects of including high paternal age on the risk of birth defects (1). With age, mutations may accumulate in the spermatogonia during the continuous process of cell division (2). An increased number of mutations in the sperm of older fathers may subsequently increase the number of birth defects in their offspring. One study has indicated that fathers aged 40 years and above have a two-fold risk relative to young fathers of having children with hydrocephaly (3). High paternal age was also found to contribute to cleft lip and cleft palate and to defects
From the Centre for International Health, University of Bergen, Bergen, Norway (M.K.); Section for Medical Statistics, Department of Public Health and Primary Health Care, University of Bergen, Bergen, Norway (M.K., R.T.L., R.S.); and Medical Birth Registry of Norway, Locus of Registry Based Epidemiology, University of Bergen, Bergen, Norway (R.T.L., R.S.). Address correspondence to: Method R. Kazaura, Centre for International Health, Armauer Hansen Building, N-5021, Bergen, Norway. Tel: ⫹475558-8518; Fax: ⫹47-5597-4979. E-mail: [email protected] Received April 30, 2003; accepted October 13, 2003. 쑕 2004 Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010
of nails. Other studies have indicated an effect of father’s age for other birth defects (4–6). A recent study found an elevated risk among older fathers for the combined category of neural tube defects, for hydrocephaly and spina bifida, congenital cataracts, and reduction of upper limb defects (7). The same study reported increased risks among offspring of young fathers (⬍20 years) for neural tube defects, anencephaly, and hydrocephaly. Studies from New York and Beijing indicated an association of young paternal age with congenital heart defects (6, 8). The age of the father and the mother are highly correlated. Careful adjustment is therefore needed to separate the effects of the two if there are effects of maternal age (9). There are several examples of such adjustments (10, 11). Mother’s age is a well-known risk factor for chromosomal aberrations. We previously studied the effect of paternal age on Down’s syndrome and gastroschisis (including omphalocele) separately (12, 13). We did not find conclusive evidence of an association between paternal age and Down’s syndrome. However, there were some indications that young paternal age increased the risk of gastroschisis. Using a large population-based data set we analyze here a broad spectrum of other defects categories to look for effects of paternal age and possible consistencies with previous studies. We adjusted the analyses for maternal age, year of birth of the child, parity, and place of birth. 1047-2797/04/$–see front matter doi:10.1016/j.annepidem.2003.10.003
AEP Vol. 14, No. 8 September 2004: 566–570
Selected Abbreviations and Acronyms CI ⫽ confidence interval CNS ⫽ central nervous systems OR ⫽ odds ratio
MATERIALS AND METHODS The Medical Birth Registry of Norway is a populationbased registry covering all births since 1967. Norwegian law requires notification to the registry of all pregnancies exceeding 16 weeks within the first week after the birth. By 1998, there were a total of 1,869,388 births recorded. In this study, the following categories of birth defects, classified according to The International Classification of Diseases (ICD-8) were included: Anencephaly, spina bifida, hydrocephaly, other CNS, ear/face/neck, heart, circulatory, respiratory, total cleft lip, anal, renal, limb defects, and cleft palate defects. We also included a category for “any birth defect” that accounts for a child with any of the above categories of birth defects. We used major categories of malformations as described in previous studies (14–16). There is considerable variation in the ascertainment of birth defects in the Medical Birth Registry of Norway. By comparing with independently collected data, it has been estimated that as much as 90% of cases with neural tube defects and 80% with cleft lip are ascertained (17). Ascertainment by the Medical Birth Registry of Norway for internal defects like heart defects is probably more incomplete. For 7% (130,536) of the children, father’s identity was unknown and paternal age therefore unknown. These births were excluded from the analysis. Logistic regression models were used to estimate the effect of paternal age measured by odds ratios (ORs). Separate analyses were performed for the different categories of birth defects. Both maternal and paternal ages were computed as age in days but scaled to completed years. Paternal age was first treated as a continuous variable in the analyses to assess a linear trend in risk of birth defects with advancing paternal age and then also categorized into 5-year intervals as shown in the tables. Adjustment was made in the logistic models for several potential confounders, most importantly for a possible effect of maternal age (18). To minimize the possibility of residual confounding within broader categories of maternal age, maternal age was used as a continuous regression variable in the models. Since an effect of maternal age could have a non-linear relationship with risk, we also entered secondand third-degree terms of maternal age in the models for each category of birth defects. We also fitted models with independent first- and second-order terms of maternal age
Kazaura et al. PATERNAL AGE AND BIRTH DEFECTS
567
above and below 30 years as an alternative mode of adjustment for all categories of birth defects (12). Since maternity institutions across Norway have slightly different procedures of diagnosis and reporting of birth defects, place of birth was also adjusted for. Five categories of institutions were distinguished in the adjustment: 1) Ulleval and Aker hospitals; 2) Haukeland Hospital; 3) Trondheim regional hospital; 4) Stavanger central hospital; and 5) all of the remaining institutions. These specific hospitals cover the largest regional hospitals in Norway. It was also expected that the prevalence of various birth defects has also changed over time partly because of diagnostic improvements (e.g., ultrasound). We also adjusted for child’s year of birth. These birth periods were categorized into 5-year intervals, with the exception of the first (1967– 1970) and the last category (1996–1998). Previous studies also controlled for paternal race (3, 6, 7). Racial disparity was, however, considered negligible in Norway. Many children born to older fathers are higher parity births. Adjustment for parity was therefore also performed. All sibships can be identified in our registry, and we adjusted for the correlation between siblings using a method of robust calculation of confidence intervals and p-values in logistic regression models. We applied each of the categories of birth defects as an outcome variable in the binary logistic regression model with all analyses performed using the statistical software STATA (19).
RESULTS The overall birth prevalence of ascertained serious birth defects (excluding Down’s syndrome and all known chromosomal anomalies) in Norway was 2.5%. There was no clear overall pattern of increased risk by increasing paternal age. No effects were observed for maternal age (Table 1). TABLE 1. Prevalence of birth defectsa at birth by parental age groups in Norway, 1967–1998 Prevalence by fathers age
Prevalence by mothers age
Age group (years)
Cases
Percent
Cases
Percent
⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹ Total
253 5862 14392 12675 6287 2336 753 255 42813
2.44 2.29 2.45 2.54 2.51 2.47 2.45 2.33 2.46
3026 13522 16502 9705 3414 628 27 46824
2.34 2.43 2.55 2.60 2.55 2.47 2.02 2.51
a
Excluding Down’s syndrome and other known chromosomal aberrations.
568 Kazaura et al.
AEP Vol. 14, No. 8 September 2004: 566–570
PATERNAL AGE AND BIRTH DEFECTS
Using logistic regression models, we estimated the effect of father’s age adjusting for mother’s age, parity, maternity institution, and birth year of the child. Only minor differences were seen across all defect categories in estimates of the effect of father’s age depending on mode of adjustment for mother’s age. We have therefore tabulated effects estimated from models adjusting for first- and second-degree terms of mother’s age in Tables 2, 3, and 4. We found the highest risk (OR ⫽ 2.5; 95% CI: 1.2–5.5) among fathers aged between 45 and 49 years compared with those aged 25 to 29 years, for a category of “other central nervous systems” (CNS) defects (Table 2). Although there was no clear monotonic pattern of risks for “other CNS” defects with advancing paternal age, a test for linear trend across the whole range of paternal age was significant (p ⬍ 0.05). We found similar results of elevated risks for respiratory, renal, and ear/face/neck defects with advancing father’s age (Tables 2, 3, and 4). A test for linear trend showed that only the ear/face/neck and renal defects categories were positively associated with paternal age (p ⬍ 0.05). However, the average risk of these defects within paternal age groups was weak. There were indications of an association between low paternal age (less than 25 years) and risks of neural tube defects and heart birth defects. For example, we estimated that fathers aged between 20 and 24 years had a 1.4-fold (95% CI: 1.1–1.8) risk of fathering a child with anencephaly as compared with fathers aged between 25 to 29 years. Although not significant, a similar pattern of risks was seen for spina bifida.
DISCUSSION The current study was aimed at assessing the effect of paternal age on the risk of selected categories of birth defects. We found little or no evidences for such effect. Unlike with one previous study (7) that found increasing risks for a combined category neural tube (and weak effects for anencephaly and spina bifida categories) with advancing paternal age, such patterns were not consistent in our study. We found an indication of an effect of advanced paternal age on broad category of birth defects categorized as “other CNS” whose major subcategories include encephalocele, microcephaly, and other defects of the brain, spine, or CNS. This association has never been reported in other similar studies. The association would therefore need confirmation from subsequent studies. In addition, the observed positive linear trend of paternal age with categories of ear/ face/neck and renal defects would require further attention from other similar studies. We did not find paternal age
TABLE 2. Relative risks for selected birth defects by paternal age in Norway, 1967–1998 Paternal age (years) and categories of birth defects Neural tube defects ⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹ Anencephaly ⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹ Spina bifida ⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹ Hydrocephaly ⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹ Other CNS ⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹
Number of cases
Risk (per 10,000 births)
Adjusted odds ratioa
95% confidence intervalb
12 222 447 347 166 78 19 12
11.6 8.7 7.6 7.0 6.6 8.3 6.2 11.0
1.4 1.3 1.0 0.9 0.8 0.9 0.8 0.9
0.9–2.5 1.1–1.5 Reference 0.8–1.0 0.6–1.0 0.7–1.2 0.5–1.3 0.4–1.9
4 93 183 133 63 27 5 4
3.9 3.6 3.1 2.7 2.5 2.9 1.6 3.7
1.4 1.4 1.0 0.8 0.8 0.8 0.3 1.3
0.6–3.4 1.1–1.8 Reference 0.6–1.0 0.6–1.1 0.5–1.3 0.1–1.0 0.4–3.7
8 129 264 214 103 51 14 8
7.7 5.0 4.5 4.3 4.1 5.4 4.5 7.3
1.4 1.2 1.0 0.9 0.8 0.9 1.0 0.7
0.8–2.8 1.0–1.5 Reference 0.8–1.1 0.6–1.0 0.6–1.4 0.6–1.8 0.2–2.0
5 65 185 183 96 33 12 4
4.8 2.5 3.2 3.7 3.8 3.5 3.9 3.6
1.2 0.8 1.0 1.0 1.1 0.8 1.2 0.5
0.6–2.6 0.6–1.0 Reference 0.8–1.2 0.8–1.5 0.5–1.3 0.6–2.2 0.1–2.0
3 52 98 75 63 16 11 4
2.9 2.0 1.7 1.5 2.5 1.7 3.6 3.7
1.8 1.4 1.0 0.9 1.4 1.1 2.5 2.6
0.6–4.9 1.0–2.0 Reference 0.7–1.3 0.9–2.0 0.6–1.9 1.2–5.5 1.0–7.1
a
Adjusted for mother’s age, parity, maternity institution, and birth year of the child. Using robust estimation of variances accounting for correlation between siblings.
b
effects for some previously reported birth defects like hydrocephaly, spina bifida, and limb defects (3, 7). A study conducted in British Columbia also reported association of young paternal age with the risk of anencephaly, spina bifida, and the combined category of neural tube defects (7). This is consistent with indications in our study that these birth defects may indeed have a weak association
AEP Vol. 14, No. 8 September 2004: 566–570
Kazaura et al. PATERNAL AGE AND BIRTH DEFECTS
569
TABLE 3. Relative risks for selected birth defects by paternal age in Norway, 1967–1998
TABLE 4. Relative risks for selected birth defects by paternal age in Norway, 1967–1998
Paternal age (years) and categories of birth defects
Paternal age (years) and categories of birth defects
Ear/Face/Neck ⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹ Heart ⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹ Circulatory ⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹ Respiratory ⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹ Isolated cleft palate ⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹
Number of cases
Risk (per 10,000 births)
Adjusted odds ratioa
95% confidence intervalb
4 127 292 271 168 63 19 9
3.9 5.0 5.0 5.4 6.7 6.7 6.2 8.2
0.7 0.9 1.0 1.2 1.3 1.2 1.4 1.1
0.3–1.5 0.7–1.2 Reference 1.0–1.4 1.0–1.6 0.9–1.7 0.9–2.2 0.5–2.5
23 434 1093 1143 601 258 76 28
22.2 16.9 18.6 22.9 24.0 27.3 24.7 25.6
1.5 1.0 1.0 1.1 1.0 1.1 1.0 1.1
1.0–2.0 0.9–1.2 Reference 1.0–1.2 0.9–1.2 0.9–1.3 0.8–1.3 0.8–1.7
5 71 218 273 132 44 12 9
4.8 2.8 3.7 5.5 5.3 4.7 3.9 8.2
1.7 1.0 1.0 1.2 0.9 1.0 0.9 1.8
0.8–3.6 0.7–1.3 Reference 1.0–1.5 0.7–1.2 0.7–1.4 0.5–1.7 0.9–3.8
3 109 320 345 164 58 21 5
2.9 4.3 5.5 6.9 6.6 6.1 6.8 4.6
0.9 1.1 1.0 1.1 1.0 1.2 1.2 0.7
0.4–2.1 0.9–1.3 Reference 0.9–1.3 0.8–1.3 0.9–1.6 0.7–2.0 0.3–2.0
9 107 298 251 129 59 22 8
8.7 4.2 5.1 5.0 5.6 6.2 7.2 7.3
1.4 0.8 1.0 1.0 0.8 1.1 1.1 1.1
0.6–2.5 0.6–1.0 Reference 0.8–1.2 0.7–1.1 0.8–1.6 0.6–1.8 0.5–2.5
Total cleft lip ⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹ Anal ⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹ Renal ⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹ Limb ⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹ Any birth defectc ⬍20 20–24 25–29 30–34 35–39 40–44 45–49 50⫹
a
a
b
b
Adjusted for mother’s age, parity, maternity institution, and birth year of the child. Using robust estimation of variances accounting for correlation between siblings.
with young paternal age. We also found a similar effect for heart defects. For the combined category of neural tube defects, because older fathers (age 30 and above) in age groups of 5-year intervals had almost equal average risks (Table 2), we pooled them to make a reference group. The increased risk in this
Risk (per 10000 births)
Adjusted odds ratioa
95% confidence intervalb
14 348 815 635 327 143 57 15
13.5 13.6 13.9 12.7 13.1 15.1 18.5 13.7
1.2 1.0 1.0 0.9 0.9 1.0 1.2 0.9
0.8–1.8 0.9–1.1 Reference 0.8–1.0 0.8–1.1 0.8–1.3 0.9–1.6 0.5–1.5
7 108 283 225 119 59 16 5
6.7 4.2 4.8 4.5 4.8 6.2 5.2 4.6
1.4 0.9 1.0 0.8 0.8 1.0 0.6 0.6
0.7–2.6 0.7–1.1 Reference 0.7–1.0 0.6–1.1 0.7–1.4 0.3–1.2 0.2–1.6
4 111 296 340 201 79 20 10
3.9 4.3 5.0 6.8 8.0 8.4 6.5 9.1
1.1 1.1 1.0 1.1 1.2 1.4 1.3 1.6
0.5–2.4 0.9–1.4 Reference 0.9–1.3 0.9–1.4 1.0–1.9 0.8–2.0 0.8–3.2
61 1463 3726 3400 1714 620 210 61
58.7 57.1 63.5 68.2 68.4 65.6 68.2 55.7
0.9 1.0 1.0 1.0 1.0 1.0 1.0 0.9
0.7–1.1 0.9–1.0 Reference 1.0–1.1 1.0–1.1 0.9–1.2 0.9–1.2 0.6–1.1
253 5862 14392 12675 6287 2336 753 255
2.4 2.3 2.5 2-5 2.5 2.5 2.5 2.3
1.0 1.0 1.0 1.0 1.0 1.1 1.1 0.9
0.9–1.2 0.9–1.0 Reference 1.0–1.0 1.0–1.1 1.0–1.1 1.0–1.2 0.8–1.0
Number of cases
Adjusted for mother’s age, parity, maternity institution, and birth year of the child. Using robust estimation of variances accounting for correlation between siblings. Excluding Down’s syndrome and other known chromosomal aberrations.
c
category of birth defects was more evident. For example, very young fathers (less than 20 years) had a 1.7-fold (95% CI: 1.0–2.9) increased risk of a child with neural tube defect compared with at least 30 years old. One of the main advantages of this study is that we used a large data set that produced reasonably stable estimates
570 Kazaura et al. PATERNAL AGE AND BIRTH DEFECTS
of effects even for rare categories of defects. The Medical Birth Registry of Norway, like other population-based registries, does not have complete ascertainments for key categories of defects (for example, neural tube defects). Some cases of birth defects are diagnosed too late and therefore not recorded and some diagnoses are simply missed (17). A general misclassification of defects should therefore produce too low estimates of effects. It is however not likely that under-reporting of birth defects would depend on father’s age. We had little opportunity to assess whether effects of father’s age were missed because of such misclassification of cases. We excluded 7% of births because the age the father was unknown. For a small proportion of births the information on father’s identity and subsequently his age is likely to be wrong. The effect of such errors is likely to be small. A series of previously considered biological, socio-economical, and environmental confounding factors like smoking and the use of medications were not linked in the present data. Smoking has been repeatedly reported as a risk factor for several malformations, for example limb defects, clubfoot, cleft lip and palate (20–24). This could potentially explain some differences between our results and of the previous studies. Our observed effects of young paternal age could, for example, be due to confounding from specific lifestyle factors. In conclusion, our data do not show consistent evidence that the risk of birth defects in general increases with increased paternal age. Except for a mixed category of “other CNS” defects, we found no indications of increased risks by high paternal age for any specific category of defects. Slightly increased risks for neural tube defects and heart defects among offspring of young fathers may, if real, be caused by lifestyle factors that are more prevalent among young couples.
REFERENCES 1. Trasler JM, Doerksen T. Paternal exposures–reproductive risks. Teratology. 1999;60:161–172. 2. Penrose LS. Parental age and mutation. Lancet. 1955;2:312–313. 3. Lian ZH, Zack MM, Erickson JD. Paternal age and the occurrence of birth defects. Am J Hum Genet. 1986;9:648–660.
AEP Vol. 14, No. 8 September 2004: 566–570
4. Harlap S, Paltiel O, Deutsch L, Knaanie A, Masalha S, Tiram E, et al. Paternal age and preeclampsia. Epidemiology. 2002;13:660–667. 5. Olshan AF, Schnitzer PG, Baird PA. Paternal age and the risk of congenital heart defects. Teratology. 1994;50:80–84. 6. Polednak AP. Paternal age in relation to selected birth defects. Hum Biol. 1976;48:727–739. 7. McIntosh GC, Olshan AF, Baird PA. Paternal age and the risk of birth defects in offspring. Epidemiology. 1995;6:282–288. 8. Zhan SY, Lian ZH, Zheng DZ, Gao L. Effect of fathers’ age and birth order on occurrence of congenital heart disease. J Epidemiol Community Health. 1991;45:299–301. 9. Stene J, Stene E. Statistical methods for detecting a moderate paternal age effect on incidence of disorder when a maternal one is present. Ann Hum Genet. 1977;40:343–353. 10. Erickson JD, Bjerkedal TO. Down syndrome associated with father’s age in Norway. J Med Genet. 1981;18:22–28. 11. Martinez FML, Rodriguez PE, Prieto L. Prenatal exposure to salicylates and gastroschisis: A case–control study. Teratology. 1997;56:241–243. 12. Kazaura MR, Lie RT. Down’s syndrome and paternal age in Norway. Paediatr Perinat Epidemiol. 2002;16:314–319. 13. Kazaura MR, Lie RT, Irgens LM, Didriksen A, Kapstad M, Egenaes J, et al. Increasing risk of gastroschisis in Norway: An age-period-cohort analysis. Am J Epidemiol. Submitted. 14. Melve KK, Skjærven R. Families with birth defects: Is birth weight of nonmalformed siblings affected? Am J Epidemiol. 2002;155:932–940. 15. Lie RT, Wilcox AJ, Skjærven R. Survival and reproduction among males with birth defects and risk of recurrence in their children. JAMA. 2001;285:755–760. 16. Lie RT, Wilcox AJ, Skjaerven R. A population-based study of the risk of recurrence of birth defects. N Engl J Med. 1994;331:1–4. 17. Lie RT, Heuch I, Irgens LM. Maximum likelihood estimation of the proportion of congenital malformations using double registration system. Biometrics. 1994;50:433–444. 18. Erickson DJ. Paternal age and Down syndrome. Am J Hum Genet. 1979; 31:489–497. 19. STATA Corporation. STATA User’s Guide. Texas: Stata Press; 2001. 20. Wasserman CR, Shaw GM, O’Malley CD, Tolarova MM, Lammer EJ. Parental cigarette smoking and risk for congenital anomalies of the heart, neural tube, or limb. Teratology. 1996;53:261–267. 21. Chung KC, Kowalski CP, Kim HM, Buchman SR. Maternal cigarette smoking during pregnancy and the risk of having a child with cleft lip/ palate. Plast Reconstr Surg. 2000;105:485–491. 22. Kallen K. Maternal smoking and orofacial clefts. Cleft Palate Craniofac J. 1997;34:11–16. 23. Van den Eeden SK, Karagas MR, Daling JR, Vaughan TL. A case–control study of maternal smoking and congenital malformations. Paediatr Perinat Epidemiol. 1990;4:147–155. 24. Zhang J, Savitz DA, Schwingl PJ, Cai WW. A case–control study of paternal smoking and birth defects. Int J Epidemiol. 1992;21:273–278.