J. biosoc. Sci. (2003) 35, 287–302 2003 Cambridge University Press Printed in the United Kingdom DOI:10.1017/S0021932003002876
A COMPARATIVE STUDY OF ZYGOTIC TWINNING AND TRIPLET RATES IN EIGHT COUNTRIES, 1972–1999
Y. IMAIZUMI
Faculty of Health Science, Hyogo University, Kakogawa City, Hyogo Prefecture, 675–0101 Japan
Summary. Annual changes in twinning and triplet rates by zygosity were investigated in eight countries during the period 1972–1999 using vital statistics. The monozygotic (MZ) twinning rates in Denmark, Switzerland and the Slovak Republic remained more or less constant throughout this period, whereas those in England and Wales, the Federal Republic of Germany (Germany), the Netherlands, the Czech Republic and Japan increased significantly year by year. With the exception of the Slovak Republic, the dizygotic (DZ) twinning rate increased significantly year by year in each country. It was 2·9 times higher in Denmark and 1·5 times higher in Germany in 1999 than in 1972, and within the same range in the other countries. With two exceptions, the MZ triplet rates remained more or less constant in each country. On the other hand, the DZ and trizygotic (TZ) triplet rates increased significantly year by year in each country. The TZ rate increased 30-fold in Germany, 16·6-fold in Japan, 11·7-fold in Switzerland, 9·7-fold in the Czech Republic, 8·7-fold in the Netherlands, 6·4-fold in Denmark, 5·6-fold in England and Wales and 3·5-fold in the Slovak Republic. The higher DZ twinning rate and higher DZ and TZ triplet rates since 1983 have been attributed to the higher proportion of mothers being treated with ovulation-inducing hormones and in vitro fertilization (IVF) in Denmark, England and Wales, Germany, the Netherlands, Switzerland and Japan. After the introduction of fertility drugs and IVF, variations in the DZ twinning and triplet rates and the TZ triplet rates were not only due to biological factors, but also depended on the popularity of fertility drugs and IVF in each country. In the Slovak Republic, where human fertility might not be affected by some adverse environmental factors, the DZ:MZ ratio remained constant during the period 1972–1999.
Introduction It has long been considered that the monozygotic (MZ) twinning rate is 3–4 per 1000 births throughout the world, and that the dizygotic (DZ) twinning rate varies among 287 288 Y. Imaizumi races (Bulmer, 1970). However, the introduction of ovulation-inducing hormones (Wyshak, 1978) and in vitro fertilization (IVF; Elsner et al., 1997) has increased the incidence of multiple births. Furthermore, according to two recent studies (Werler et al., 1997; Ceizel, 1998), twinning rates increase significantly after periconceptional multivitamin supplementation. According to Bressers et al. (1987), the MZ twinning rate increased in many European countries during the period 1960–1984, a time when the use of oral contraceptives was widespread. In Japan, the MZ twinning rate remained constant during the period 1975–1994 (Imaizumi & Nonaka, 1997a), whereas the DZ twinning rate increased from 1987 to 1994. In the Czech and Slovak Republics, the MZ and DZ twinning rates remained constant during the period 1972–1995 (Imaizumi, 1999). In Japan (Imaizumi & Nonaka,1997b) the MZ triplet rate remained constant during the period 1975–1994, but the TZ triplet rate increased during that period. In England and Wales (Murphy et al., 1997) the TZ triplet rate increased after 1971–1975, and increased rapidly after 1982–1984 up to 1991–1994. The present study focuses on the trends in zygotic twinning and triplet rates in eight countries during the period 1972–1999. According to Tong, Caddy & Short (1997), the DZ:MZ twinning ratio can provide a useful measure of human fertility. This study assesses the validity of this measure of human fertility.
Methods Vital statistics data on twin and triplet births were obtained courtesy of the staff of the statistics sections of eight countries: the Federal Statistics Office of the Federal Republic of Germany (Germany), the Swiss Federal Statistical Office, the Office for National Statistics in England and Wales, the Central Bureau of Statistics in the Netherlands, Statistics Denmark, the Czech Statistical Office, the Statistical Office of the Slovak Republic and the Health and Welfare Statistics and Information Department in Japan. Details of estimated numbers for sex compositions of pairs of twins and sets of triplets in Japan were reported by Imaizumi & Nonaka (1997a,b). Zygotic twinning rates were estimated using the Weinberg method (1901). Triplet rates by zygosity were estimated using the Allen method (1960), as follows: No. of MZ triplets=L(U+D)/3, No. of DZ triplets=D=2n(MZ twinning rate)(DZ twinning rate), No. of TZ triplets=T=4(UD/2)/3, where L is the number of like-sexed triplets, U is the number of unlike-sexed triplets and n is the total number of live and stillbirths.
Results Twins Table 1 shows the number of twins by sex combination and the estimated twinning rate by zygosity in Germany, the Netherlands, Switzerland, England and Wales, Denmark, the Czech Republic, the Slovak Republic and Japan during the period l972–1999. The MZ twinning rates were about 4 per 1000 births in each country Zygotic twinning and triplet rates in eight countries 289 during the entire period. The linear regression coefficients (SE) of the MZ twinning rate on the year were 22·932.46 in Germany, 12·342.44 in the Netherlands, 0·040.05 in Switzerland (not significant, ns), 23·863.37 in England and Wales, 1·094.26 in Denmark (ns), 0·020.01 in the Czech Republic, 0·010.01 in the Slovak Republic (ns) and 0·020.003 in Japan. The regression coefficients for the above five countries were significant at the 5% level. On the other hand, the DZ twinning rate remained more or less constant from 1972 to 1982 but increased thereafter, with the DZ rate in Denmark increasing rapidly after 1990. The linear regression coefficients (SE) of the DZ twinning rates on the year between 1972 and 1999 in corresponding countries were: 4·290.45, 3·340.25, 5·130.30, 5·090.48, 2·470.34, 0·120.03, 0·0030.01 (ns) and 0·110.01, respectively. The regression coefficient was significant at the 1% level, except for that of the Slovak Republic. The DZ twinning rate was 1·6 times higher in England and Wales and 2·9 times higher in Denmark in 1999 than in 1972. It was within the same range in the other countries, with the exception of the Slovak Republic, where the value was 0·94. The DZ twinning rate was at its highest in Denmark between 1993 and 1996.
Triplets Table 2 shows the number of triplets by sex combination and the estimated triplet rate by zygosity in Germany during the period 1972–1999, in England and Wales for the period 1974–1999 and in Japan for the period 1974–1998. In Germany and England and Wales, the MZ triplet rate was lower than the DZ and TZ triplet rates during this period. On the other hand, with one exception, the MZ rate in Japan was higher than the DZ and TZ rates during the period 1974–1985. The TZ rate was at its highest in Germany after 1981, in England and Wales after 1977, and in Japan after 1986. In Germany, the DZ rate was highest during the periods 1972–1976 and 1978–1979, whereas that in Japan was lowest for the period 1977–1988. The TZ triplet rate remained constant from 1972 to 1979 (17–41) and rapidly increased thereafter up to 1998 (509) in Germany. In England and Wales, the TZ rate remained constant from 1974 to 1985 (37–80) and rapidly increased up to 1997 (339) and decreased thereafter. In Japan, the TZ triplet rate remained more or less constant from 1974 to 1988 (12–52) and rapidly increased thereafter up to 1998 (202). The TZ triplet rate was higher in England and Wales than in Germany from 1974 to 1980, but after 1982 the rates were similar up to 1994 for both countries. After 1996, the TZ rate was higher in Germany than in England and Wales. During the period from 1972 to 1999, the TZ rate increased 30-fold in Germany, 5·6-fold in England and Wales and 16·4-fold in Japan. In Germany, the linear regression coefficients (SE) of the triplet rate on the year were 1·950.16 for DZ and 15·931.49 for TZ, and the corresponding values were 1·800.12 and 12·691.03 in England and Wales, and 1·040.10 and 8·900.82 in Japan, respectively. All regression coefficients were significant at the 5% level in each country. The DZ triplet rate was 2–3 times higher in European countries than in Japan. Similarly, with three exceptions, the TZ triplet rate was 1·2–3·9 times higher in European countries than in Japan. The exceptions were 1978–1979 in Germany and in 1982 for England and Wales, where the TZ rate was slightly higher in Japan than in European countries. Table 1. Number of twin pairs and twinning rate according to zygosity in eight countries, 1972–1999 290
Year Germany The Netherlands Switzerland England & Wales Denmark Czech Republic Slovak Republic Japan MZ DZ MZ DZ MZ DZ MZ DZ MZ DZ MZ DZ MZ DZ MZ DZ
Number of twin pairs 1972 3070 5488 866 1352 371 398 — — 318 366 519 1006 311 592 — — 1973 3093 4606 774 1232 343 446 — — 288 414 552 1172 325 594 — — 1974 3072 4482 750 1052 353 402 2405 3746 296 404 709 1158 262 692 8302 3948 1975 2846 4340 687 1028 344 410 2165 3744 329 362 704 1070 297 656 7494 3720 1976 3028 4460 823 1004 298 408 2242 3296 245 378 602 1124 358 560 7073 3654 1977 2949 4192 786 1010 296 398 2311 3138 265 328 503 1174 333 634 7265 3732 1978 3155 4088 716 1028 298 412 2543 3316 304 318 544 1116 291 660 6943 3644 1979 3211 4160 742 1090 289 386 2691 3408 264 362 484 1028 337 636 7001 3692 1980 3425 4542 845 1082 369 378 2554 3754 262 320 445 996 246 642 6644 3654 Imaizumi Y. 1981 3173 4608 854 1078 326 400 — — 225 294 439 850 239 580 6570 3536 1982 3465 4402 834 1020 311 398 2678 3524 240 292 442 866 262 600 6543 3546 1983 3246 4546 690 1140 312 402 2547 3746 198 320 437 884 272 556 6471 3502 1984 3415 4494 771 1220 274 442 2677 3644 233 334 456 764 257 590 6457 3440 1985 3221 5028 752 1306 281 484 2850 3850 265 322 428 852 261 546 6181 3334 1986 3498 5294 812 1382 320 480 2779 4190 259 318 436 712 251 538 5773 3294 1987 3542 5644 874 1376 242 564 2838 4348 239 378 453 802 210 480 5752 3226 1988 3529 5954 892 1454 283 576 3002 4450 276 386 405 782 225 490 5636 3300 1989 3580 5964 840 1640 344 618 2927 4652 254 436 394 722 246 464 5386 3368 1990 3850 6174 926 1764 305 644 3190 4744 256 456 422 730 245 492 5037 3552 1991 3340 6280 908 1886 347 626 2978 5182 270 570 371 784 233 420 5017 3804 1992 3648 5916 871 1918 337 666 2926 5388 317 606 385 766 259 508 5108 4006 1993 3490 6336 919 1906 256 700 2972 5330 318 704 372 708 177 512 5023 4316 1994 3376 6368 872 1946 282 712 3087 5364 257 880 338 660 192 434 5421 4986 1995 3507 6588 1090 1980 293 682 2959 5790 335 788 306 704 190 388 5319 5138 1996 3689 7374 1059 2138 333 730 2817 5798 259 948 334 700 181 378 5497 5526 1997 3689 7776 977 2268 333 664 2977 5922 — — 346 820 137 396 5415 5618 1998 3404 8298 1090 2466 323 714 2870 5906 — — 296 1140 175 390 5283 5762 1999 3231 8250 938 2696 348 782 2722 5914 — — 398 1022 180 356 — — Table 1. Continued
Year Germany The Netherlands Switzerland England & Wales Denmark Czech Republic Slovak Republic Japan MZ DZ MZ DZ MZ DZ MZ DZ MZ DZ MZ DZ MZ DZ MZ DZ
Zygotic twinning rate per 1000 births 1972 3·4 6·0 4·0 6·3 4·0 4·3 — — 4·2 4·8 3·2 6·1 3·5 6·7 — — yoi wnigadtiltrtsi ih countries eight in rates triplet and twinning Zygotic 1973 3·8 5·6 3·9 6·3 3·9 5·1 — — 4·0 5·7 3·0 6·4 3·5 6·3 — — 1974 3·8 5·5 4·0 5·6 4·2 4·7 3·7 5·8 4·1 5·6 3·6 5·9 2·7 7·0 3·9 1·9 1975 3·6 5·5 3·8 5·7 4·4 5·2 3·6 6·1 4·5 5·0 3·7 5·6 3·0 6·7 3·7 1·9 1976 3·8 5·6 4·6 5·6 4·0 5·5 3·8 5·6 3·7 5·8 3·2 6·0 3·6 5·6 3·7 1·9 1977 3·6 5·2 4·5 5·8 4·0 5·4 4·0 5·5 4·3 5·3 2·8 6·4 3·3 6·3 3·9 2·0 1978 3·9 5·0 4·1 5·8 4·2 5·7 4·2 5·5 4·9 5·1 3·0 6·2 2·9 6·5 3·9 2·0 1979 3·9 5·1 4·2 6·2 4·0 5·3 4·2 5·3 4·4 6·1 2·8 5·9 3·3 6·3 4·1 2·1 1980 3·9 5·2 4·6 5·9 5·0 5·1 3·9 5·7 4·6 5·6 2·9 6·4 2·6 6·7 4·0 2·2 1981 3·7 5·3 4·8 6·0 4·4 5·4 — — 4·2 5·5 3·0 5·9 2·5 6·2 4·1 2·2 1982 4·1 5·1 4·8 5·9 4·1 5·3 4·3 5·6 4·5 5·5 3·1 6·1 2·8 6·4 4·1 2·2 1983 3·9 5·5 4·0 6·7 4·2 5·4 4·0 5·9 3·9 6·3 3·2 6·4 2·9 6·0 4·1 2·2 1984 4·2 5·5 4·4 7·0 3·7 5·9 4·2 5·7 4·5 6·4 3·3 5·6 2·8 6·5 4·1 2·2 1985 3·9 6·2 4·2 7·3 3·8 6·5 4·3 5·8 4·9 6·0 3·1 6·2 2·9 6·0 4·1 2·2 1986 4·1 6·2 4·4 7·5 4·2 6·3 4·2 6·3 4·7 5·7 3·3 5·3 2·9 6·1 4·0 2·3 1987 4·1 6·5 4·7 7·3 3·2 7·3 4·1 6·3 4·2 6·7 3·5 6·1 2·5 5·7 4·1 2·3 1988 3·9 6·6 4·8 7·8 3·5 7·1 4·3 6·4 4·7 6·5 3·0 5·9 2·7 5·9 4·1 2·4 1989 4·1 6·8 4·4 8·6 4·2 7·6 4·2 6·7 4·1 7·1 3·1 5·6 3·1 5·8 4·1 2·6 1990 4·2 6·8 4·7 8·9 3·6 7·6 4·5 6·7 4·0 7·2 3·2 5·6 3·0 6·1 4·0 2·8 1991 4·0 7·5 4·6 9·4 4·0 7·2 4·2 7·4 4·2 8·8 2·9 6·0 3·0 5·3 3·9 3·0 1992 4·5 7·3 4·4 9·7 3·9 7·6 4·2 7·8 4·7 8·9 3·2 6·3 3·5 6·8 4·1 3·2 1993 4·4 7·9 4·7 9·7 3·0 8·3 4·4 7·9 4·7 10·4 3·1 5·8 2·4 7·0 4·1 3·5 1994 4·4 8·2 4·4 9·9 3·4 8·6 4·6 8·0 3·7 12·6 3·2 6·2 2·9 6·5 4·2 3·9 1995 4·6 8·6 5·7 10·4 3·6 8·3 4·5 8·9 4·8 11·2 3·2 7·3 3·1 6·3 4·3 4·2 1996 4·6 9·2 5·6 11·2 4·0 8·8 4·3 8·9 3·8 14·0 3·7 7·7 3·0 6·3 4·4 4·4 1997 4·5 9·5 5·1 11·7 4·1 8·2 4·6 9·2 — — 3·8 9·0 2·3 6·7 4·4 4·6
1998 4·3 10·5 5·4 12·3 4·1 9·0 4·5 9·2 — — 3·3 12·6 3·0 6·7 4·3 4·6 291 1999 4·2 10·7 4·7 13·4 4·4 10·0 4·4 9·5 — — 4·4 11·4 3·2 6·3 — — Table 2. Secular change of triplet deliveries and triplet rates by zygosity in Germany, 1972–1999, in England and Wales, 292 1974–1980 and 1982–1999 and in Japan, 1974–1998
Germany England & Wales Japan
Year Triplet deliveries Triplet rate* Triplet deliveries Triplet rate* Triplet deliveries Triplet rate*
Total MMM MMF MFF FFF MZ DZ TZ Total MMM MMF MFF FFF MZ DZ TZ Total MMM MMF MFF FFF MZ DZ TZ
1972 64 12 14 16 22 12·8 40·7 16·8 — — — — — — — — — — — — — — — — 1973 81 32 15 18 16 30·9 42·0 25·4 — — — — — — — — — — — — — — — — 1974 78 23 15 20 20 24·6 41·5 29·6 60 13 20 17 10 2·1 43·0 47·6 124 36 20 15 48 29·2 14·6 12·2 1975 72 21 11 20 20 25·7 39·7 25·9 72 17 18 24 13 11·7 43·6 62·8 123 37 16 25 43 28·5 13·9 18·0 1976 74 21 16 16 21 25·0 41·8 25·2 76 24 13 16 23 49·1 42·5 37·2 119 33 17 17 46 30·4 13·8 14·2 1977 75 18 22 18 17 14·2 37·6 40·7 68 16 17 22 13 13·2 43·9 61·2 127 42 17 24 39 31·1 15·8 19·0 1978 60 16 9 19 16 14·9 38·9 19·9 62 14 14 21 13 10·1 46·6 46·5 122 36 16 26 42 30·4 15·7 20·7 1979 60 14 14 16 16 11·2 39·5 22·3 76 15 13 33 15 8·0 44·3 65·8 123 32 24 25 39 25·9 17·4 26·3 Imaizumi Y. 1980 94 21 26 20 27 23·8 41·0 43·1 91 32 15 25 19 42·4 43·9 51·4 123 40 19 25 35 30·6 17·7 23·6 1981 103 24 25 34 20 15·1 38·9 64·8 — — — — — — — — 149 39 34 18 53 40·4 18·0 31·1 1982 95 17 27 27 24 13·0 41·7 56·4 70 15 19 15 21 23·3 47·6 40·3 161 44 40 33 38 30·1 18·3 48·9 1983 76 17 22 26 11 0·2 42·6 48·5 89 18 22 31 18 13·1 47·7 79·9 137 39 24 22 42 35·5 18·1 26·7 1984 124 29 32 28 35 38·6 46·1 67·3 80 22 17 25 16 21·6 47·6 55·7 135 41 28 22 40 35·1 18·2 30·5 1985 137 19 40 48 30 7·9 48·5 111·2 93 13 22 23 35 33·2 50·4 57·3 129 34 19 27 41 33·7 18·3 28·7 1986 138 26 36 41 35 24·5 51·1 86·5 123 33 31 37 22 31·1 52·7 101·3 126 36 30 24 33 29·2 18·1 37·6 1987 158 38 42 56 22 13·8 52·6 114·8 125 21 39 44 21 3·4 52·6 126·5 151 40 32 31 43 37·7 18·7 47·1 1988 217 40 63 68 46 29·8 52·3 160·0 157 29 52 46 30 19·5 55·0 150·8 149 35 29 38 39 31·0 19·7 51·9 1989 215 48 65 62 40 33·5 54·7 155·2 183 37 59 62 25 12·3 57·0 195·5 157 36 45 41 27 19·2 21·4 73·8 1990 204 32 68 61 43 16·0 57·6 150·9 201 44 62 55 40 43·4 60·1 179·8 211 42 53 59 45 31·6 22·0 102·4 1991 234 53 69 76 36 28·7 60·5 191·8 208 45 63 64 36 33·9 63·5 198·7 224 42 61 73 38 19·9 23·5 124·6 1992 240 41 90 67 42 15·9 65·5 214·2 202 30 72 68 32 0·1 66·9 225·0 285 57 99 58 51 35·6 25·9 149·2 1993 276 49 89 96 42 13·6 68·9 262·0 234 40 69 82 43 24·9 70·1 250·5 281 57 97 80 36 18·1 28·5 172·3 1994 262 38 99 71 54 21·7 72·0 245·3 260 53 72 86 49 48·7 75·3 264·9 346 62 98 117 53 22·8 32·9 201·8 1995 280 56 87 86 51 38·1 78·2 248·0 282 52 84 97 49 35·8 79·8 317·1 337 64 110 95 45 21·0 36·4 198·6 1996 386 67 118 130 71 40·8 85·1 356·8 259 55 86 79 39 34·2 76·6 285·8 323 64 94 98 58 33·5 39·1 179·4 1997 401 70 136 142 53 8·5 86·2 396·9 295 56 101 86 52 42·5 84·4 329·4 322 59 96 106 49 19·7 40·1 192·0 1998 491 80 159 178 74 22·6 90·9 509·4 297 63 109 80 45 42·7 82·9 338·9 338 62 109 104 54 23·1 39·4 202·3 1999 486 91 172 154 69 36·7 89·0 502·3 267 55 79 72 61 77·6 82·4 267·1 — — — — — — — —
*Triplet rate per million births. Zygotic twinning and triplet rates in eight countries 293
Fig. 1. Yearly changes in the monozygotic triplet rate (3-year moving average) per million births in Germany, England and Wales and Japan.
Because of a small number of total triplets as well as MZ triplets in each year for Germany and England and Wales, the 3-year moving average for MZ triplet rates was computed for each year and for each country (Fig. 1). The MZ triplet rate was 1·1–3·8 times higher in Japan than in Germany and 1·1–4·2 times higher in Japan than in England and Wales during the period 1974–1988. After 1994, the MZ rate in England and Wales was the highest among the three countries. The linear regression coefficients (SE) of the MZ triplet rate on the year were 0·260.12 in Germany, 1·200.22 in England and Wales and 0·300.14 in Japan (ns). The coefficients were significant at the 5% level in the former two countries. Table 3 shows the number of triplets by sex combination and the estimated triplet rate by zygosity in the Netherlands, Switzerland, the Czech Republic, the Slovak Republic and Denmark during the period 1972–1999. During this period, the TZ rate increased 11·7-fold in Switzerland, 9·7-fold in the Czech Republic, 8·7-fold in the Netherlands, 6·4-fold in Denmark and 3·5-fold in the Slovak Republic. In the Netherlands, the TZ rate increased from 1972–1973 to 1990–1991 and decreased thereafter. The linear regression coefficients (SE) of the triplet rate on the year were 0·020.15 for MZ, 0·280.03 for DZ and 0·040.01 for TZ triplets. The coefficients for the DZ and the TZ triplets were significant at the 1% level. In Switzerland, the linear regression coefficients (SE) were 0·370.49 for MZ, 0·660.14 for DZ and 0·100.01 for TZ triplets. The corresponding values were 1·410.99, 0·350.22 and 0·090.04 in the Czech Republic, 0·440.71, 294 Y. Imaizumi Table 3. Trends of triplet sets and triplet rates by zygosity in the Netherlands, Switzerland, Czech Republic, Slovak Republic and Denmark, 1972–1999
Triplet sets Triplet rate per million births Year Total MMM MMF MFF FFF Monozygotic Dizygotic Trizygotic
The Netherlands 1972–73 41 15 7 12 7 21·4 49·7 28·2 1974–75 50 9 11 12 18 37·9 44·4 54·0 1976–77 53 8 15 12 18 30·8 52·0 67·3 1978–79 49 11 16 8 14 31·6 49·6 57·6 1980–81 53 9 16 12 16 24·6 56·0 65·8 1982–83 52 13 15 8 16 43·5 55·5 52·0 1984–85 63 10 18 21 14 10·6 61·2 105·8 1986–87 79 7 32 23 17 -7·1 66·8 152·0 1988–89 120 19 40 35 26 27·9 75·1 214·7 1990–91 233 36 82 73 42 38·0 84·2 462·1 1992–93 200 35 82 58 25 4·5 87·9 414·4 1994–95 177 41 58 50 28 50·9 102·3 302·9 1996–97 180 31 61 56 32 21·9 121·8 325·2 1998–99 144 27 48 35 34 39·7 129·7 189·0
Switzerland 1972–77 39 5 7 11 16 16·9 40·7 21·6 1978–83 87 17 26 23 21 33·6 46·4 117·0 1984–89 108 22 40 26 20 26·0 50·9 155·0 1990–95 140 20 44 51 25 7·3 56·8 211·4 1996–99 111 20 39 34 18 17·6 74·7 252·9 Czech Republic 1972–77 91 18 17 24 32 19·7 39·2 23·2 1978–83 119 31 24 38 26 26·7 36·8 64·0 1984–89 88 16 24 23 25 19·2 37·1 53·4 1990–95 91 23 27 25 16 17·9 38·1 72·5 1996–99 115 14 34 41 26 15·7 77·1 224·6
Slovak Republic 1972–77 46 11 6 10 19 18·9 27·7 5·9 1978–83 49 6 13 19 11 -0·8 23·9 32·7 1984–89 47 8 14 12 13 7·5 19·8 26·3 1990–95 43 12 9 15 7 6·3 18·7 24·1 1996–99 34 4 11 7 12 8·1 10·0 20·7 Denmark 1972–77 53 13 17 12 11 19·3 44·3 62·4 1978–83 39 8 12 5 14 31·8 50·1 33·9 1984–89 60 9 16 19 16 20·1 57·8 99·2 1990–95 170 29 61 54 26 12·6 85·9 322·1 1996 38 8 16 8 6 52·9 106·3 400·0 Zygotic twinning and triplet rates in eight countries 295 1·330.23 and 0·320.50 in the Slovak Republic, and 0·220.29, 0·310.06 and 0·050.01 in Denmark, respectively. The coefficients for DZ and TZ triplets were statistically significant at the 5% level in Switzerland and Denmark, whereas the coefficient for DZ triplets in the Slovak Republic was negatively significant at the 1% level.
Twinning rate by maternal age Figure 2 shows yearly changes in MZ and DZ twinning rates according to maternal age during the period 1982–1999 in England and Wales. With one exception, the MZ twinning rate remained constant during the entire period in each maternal age group. The exception was the 25–29-years age group, where the rate increased significantly year by year. On the other hand, with one exception, the DZ twinning rate increased significantly year by year in each maternal age. The exception was the under-20-years age group. Figure 3 shows yearly changes in MZ and DZ twinning rates according to maternal age during the period 1974–1998 in Japan. With three exceptions, the MZ twinning rate remained constant during the entire period in each maternal age group. The exceptions were the age groups 20–24 and 30–34 years, where the rate increased significantly year by year. On the other hand, with one exception, the DZ twinning rate increased significantly year by year in each maternal age. The exception was under-20-years age group.
Dizygotic to monozygotic twinning ratio Figure 4 shows the age-adjusted (standardized for maternal age) twinning rate in Japan during the period 1975–1998, and in England and Wales for the period 1982–1999. In Japan, the age-standardized DZ:MZ ratio increased significantly from 0·496 in 1975 to 0·956 in 1998 (p<0·01). Similarly, in England and Wales, the age-standardized DZ:MZ ratio increased significantly from 1·316 in 1982 to 1·886 in 1999 (p<0·01). The linear regression coefficients (SE) of the DZ:MZ ratio on the year were 0·0200.002 in Japan and 0·0300.003 in England and Wales, being significant at the 1% level. Figure 5 shows yearly changes in the DZ:MZ crude twinning ratio in eight countries during the period 1972–1999. The ratio remained constant between 1972 and 1986, and increased thereafter. The linear regression coefficients (SE) of the DZ:MZ ratio on the year were 0·0270.004 in England and Wales, 0·0300.005 in Germany, 0·0410.005 in the Netherlands, 0·0510.006 in Switzerland, 0·0330.005 in Denmark, 0·0230.009 in the Czech Republic, 0·0070.007 in the Slovak Republic and 0·0240.002 in Japan. With one exception (the Slovak Republic), these coefficients were statistically significant at the 5% level in each country.
Discussion The MZ twinning rate remained more or less constant in the Czech and Slovak Republics during the period 1972–1995 (Imaizumi, 1999) and in Japan during the 296 Y. Imaizumi
Fig. 2. Yearly changes in the monozygotic and dizygotic twinning rates by maternal age in England and Wales, 1982–1999. Zygotic twinning and triplet rates in eight countries 297
Fig. 3. Yearly changes in the monozygotic and dizygotic twinning rates by maternal age in Japan, 1974–1998. period 1975–1994 (Imaizumi & Nonaka, 1997a). In the present study, the MZ rate remained constant in Denmark for the period 1972–1996, and in Switzerland and the Slovak Republic for the period 1972–1999, but increased significantly in the other five countries. According to Bressers et al. (1987), the MZ twinning rate in fifteen European countries (including seven countries in the present study) increased significantly during the period 1960–1984. They found a temporal association between 298 Y. Imaizumi
Fig. 4. Age-adjusted and crude dizygotic to monozygotic twinning ratios in Japan, 1975–1998, and in England and Wales, 1982–1999. increasing MZ rate and increasing use of oral contraceptives. Wood (1997) analysed the twinning rate by zygosity in England and Wales for the period 1938–1995. She pointed out that ovulation-inducing drugs were first used during the 1960s, when the MZ twinning rate was rising. In the present study, the increased MZ twinning rates observed in five countries might be related to ovulation induction (Derom et al., 1987) and IVF (Edwards et al., 1986). On the other hand, the DZ twinning rate increased significantly year by year in all counties in the present study (with the exception of Zygotic twinning and triplet rates in eight countries 299
Fig. 5. The crude dizygotic to monozygotic twinning ratios in eight countries during the period 1972–1999. the Slovak Republic) during the same period (Table 1). According to Westergaard et al. (1993), 800–1000 children had been born in Denmark by 1993 as a result of IVF. Many children have been born in Germany, the UK and Switzerland as a result of IVF (Cohen, de Mouzon & Lancaster, 1993). The DZ twinning rate has therefore increased in these countries as a result of IVF and the hormonal induction of ovulation. Before the introduction of fertility drugs, the DZ twinning rate was highest in African populations, followed by Caucasian and Oriental populations (Bulmer, 1970). Pollard (1995) studied the twinning rates of fourteen ethnic groups in California in the USA during the period 1982–1988. He found that overall twinning rates were 13·2 per 1000 births for blacks, 10·1 for whites and 7·2 for Asians. DZ twinning rates were higher in Caucasian populations than Japanese populations for each year (Table 1). The DZ rate was 2·3–3·0 times higher in the former than the latter populations in 1980 and the corresponding value was 1·5–2·7 in 1998. After the introduction of fertility drugs and IVF, the variations in DZ twinning rate in each country were not only due to biological factors, but also depended on the popularity of the use of fertility drugs and IVF in each country. In the present study, with two exceptions (the Czech and Slovak Republics), the DZ and TZ triplet rates increased significantly year by year in each country during the period 1972–1999. In recent years, the TZ triplet rate was highest in Germany (509 per million births) in 1998, followed by Denmark (400) in 1996, England and 300 Y. Imaizumi Wales (339) in 1998, Switzerland (253) in 1996–1999 and Japan (202) in 1998. In Denmark, the TZ triplet rate has increased rapidly since 1987. According to Westergaard et al. (1997), the triplet rate increased in Denmark after 1989 as a result of IVF and hormonal induction of ovulation. According to Imaizumi & Inouye (1980), before the introduction of fertility drugs, the MZ triplet rate in Japan was around two times higher than that in other countries (England and Wales, the USA, Italy and Australia), and the DZ and TZ triplet rates of the former were one-third to one-quarter and one-sixth to one-thirteenth of the latter, respectively. In the present study, the corresponding values were 1·3–3·8, half to one-third and five-sixths to one-quarter, respectively (Table 2). Variations in DZ and TZ triplet rates were reduced between the Caucasian and Japanese populations. Imaizumi & Inouye (1982) observed that higher rates of triplets and quadruplets of polyembryonal origin and lower rates of those of polyovulational origin are characteristic features of Japanese compared with Caucasian populations. Therefore, after the introduction of fertility drugs and IVF, these variations in the TZ triplet rates among countries were not only due to biological factors, but also depended on the popularity of the use of fertility drugs and IVF in each country. In Japan, ovulation-inducing hormones were first used in 1966 and the first IVF baby was born in 1983. Japan’s triplet rate was 52·8 per million births in 1967 and 274·5 in 1998, a 5·2-fold increase. A nationwide survey of infertility treatments was conducted in Japan between January and February 2000 through questionnaires (Aono et al., 2000). The total number of infertility treatment facilities was 494, and questionnaires were distributed to all of them. Of the 494, 297 (60·1%) responded. The reported causes of 682 triplet births were: IVF (73·2%), ovulation-inducing hormones (22·1%) and natural pregnancies (4·3%) during the period 1997–1999. Therefore, in Japan, the rising twinning and triplet rates observed since 1974 can almost certainly be attributed to the higher proportion of mothers being treated with ovulation- inducing hormones and IVF since 1988. According to Tong et al. (1997), the DZ to MZ twinning ratio can provide a useful measure of human fertility. Tong (2000) stated that trends in published national twinning rates in developed countries from the 1970s onward no longer reflect natural human fertility. In the present study, the use of fertility drugs and IVF treatments is reflected in the DZ:MZ ratios in seven countries. In the eighth (the Slovak Republic), the DZ:MZ ratio remained constant during the period 1972–1999. It seems that human fertility in the Slovak Republic is not affected by some adverse environmental factors.
Acknowledgments The author is grateful to the staff of the Health and Welfare Statistics and Information Department in Japan, the Federal Statistics Office in the Federal Republic of Germany, the Swiss Federal Statistical Office, Statistics Denmark, the Central Bureau of Statistics in the Netherlands, the Czech Statistical Office, the Statistical Office of the Slovak Republic and the Office for National Statistics in England and Wales, for providing data on multiple births. Zygotic twinning and triplet rates in eight countries 301 References A, G. (1960) A differential method for estimation of type frequencies in triplets and quadruplets. Am. J. hum. Genet. 12, 210–224. A,T.et al. (2000) The middle report on the nationwide survey: trends of super- twin pregnancies in Japan. In: Study on Reproductive Assisted Medical Service in Japan, pp. 799–806. Edited by T. Yanaihara. Ministry of Health and Welfare of Japan, Tokyo (in Japanese). B,W.M.A.,E,A.W.,K,P.J.&P, P. (1987) Increasing trend in the monozygotic twinning rate. Acta genet. med. gemell. 36, 397–408. B, M. G. (1970) The Biology of Twinning in Man. Clarendon Press, Oxford. C, A. E. (1998) Periconceptional folic acid containing multivitamin supplementation. Eur. J. Obstet. Gynec. Reproduct. Biol. 78, 151–161. C, J., M,J.&L P. (1993) World Collaborative Report 1991. VIIIth World Congress on In Vitro Fertilization and Alternate Assisted Reproduction, Kyoto, September 1993. D, C., D, R., V, R., V B,H.&T, M. (1987) Increased monozygotic twinning rate after ovulation induction. Lancet I, 1236–1238. E, R. C., M,L.&W, D. (1986) Identical twins and in vitro fertilization. J. in vitro Fertil. Embryo Trans. 3, 114–117. E, C. W., T,M.J.,S,C.L.,et al. (1997) Multiple pregnancy rate and embryo number transferred during in vitro fertilization. Am. J. Obstet. Gynec. 177, 350–357. I, Y. (1999) Constant multiple birth rates in the Czech Republic and the Slovak Republic until recently, 1972–1995. Twin Res. 2, 10–15. I,Y.&I, E. (1980) Analysis of multiple birth rates in Japan. III. Secular trend, maternal age effect and geographical variation in triplet rates. Jap. J. hum. Genet. 25, 73–81. I,Y.&I, E. (1982) Analysis of multiple birth rates in Japan. VI. Quadruplets: birth and stillbirth rates. Jap. J. hum Genet. 27, 227–234. I,Y.&N, K. (1997a) The twinning rates by zygosity in Japan, 1975–1994. Acta genet. med. gemell. 46, 9–22. I,Y.&N, K. (1997b) Rising trizygotic triplet rates in Japan, 1975–1994. Acta genet. med. gemell. 46, 87–98. M, M., H, K., B, J., W, B., E,J.D.&B, D. (1997) Infertility treatment and multiple birth rates in Britain, 1938–94. J. biosoc. Sci. 29, 235–243. P, R. (1995) Ethnic comparison of twinning rates in California. Hum. Biol. 67, 921–931. T, S. (2000) Dizygotic to monozygotic twinning ratio at The Royal Women’s Hospital, Melbourne 1947–1997, compared with Australian national twinning incidence. Twin Res. 3, 12–16. T, S., C,D.&S, R. V. (1997) Use of dizygotic to monozygotic twinnning ratio as a measure of fertility. Lancet I 349, 843–845. W, W. (1901) Beitrage zur Physiologie und Pathololgie der Mehrlingsgeburten beim Menschen. Arch. ges. Physiol. 88, 346–430. W, M. M., C, J. D., W, C. R., S, G. M., E,J.D.& M, A. A. (1997) Multivitamin supplementation and multiple births. Am. J. med. Genet. 71, 93–96. W,L.G.,R,P.E.,M, S., I,H.J.,A,A.N., L,J.F.et al. (1993) in vitro fertiliserng. En oversigt over medicinske indikationer og forslag til falles retningslinier ved de offentlige danske fertilitetsklinikker. Ugeskr. Laeg. 155, 2511–2514 (in Danish). 302 Y. Imaizumi W, T., W, J., A,P.&M, M. (1997) Population based study of rates of multiple pregnancies in Denmark, 1980–91. Br. med. J. 314, 775–779. W, R. (1997) Trends in multiple births, 1938–1995. Population Trends 87(Spring), 29–35. W, G. (1978) Statistical findings on the effects of fertility drugs on plural births. In: Twin Research, Part B, Biology and Epidemiology, pp. 17–33. Edited by W. E. Nance, G. Allen & P. Parisi. Alan R. Liss, New York.