Renato Cocchi M.D., Ph.D. (Sociology)
Italian Journal of Intellective Impairment 10 (2): 115-20 (1997)
Paper presented during the 6th World Down Syndrome Congress, Madrid October 1997
A poster on first 360 cases was presented during the 8th Congress of I.A.S.S.M.D., Dublin, August 1988 and a letter on the same cases has been printed in Giorn. Neuropsich. Età Evol. 1988, 8: 256-257.
  Reprinted with the permission of Renato Cocchi
Via A. Rabbeno, 3
42100 Reggio Emilia, Italy
+39 0522 320 716
Mobile +39 348 5145 520
URL: http://www.stress-cocchi.net

Summary

An epidemiological investigation on the possible influence of seasonal factors on conceptions of Trisomy 21 subjects was carried out, as the following of a previous paper (Cocchi R., Occhialini O.: It. J. Intellect. Impair. 1989, 2 1-7) The monthly conception's periods of a representative sample of 487 Italian Down Ss born between 1973 and 1992 were compared with the same monthly periods of all Italian live-born babies of the same years, minus 1/700 as the rate of Down births.
The seasonal distribution of conceptions of Down Ss has been found significantly different (.018), having increased rates particularly in spring and in first and second month of autumn and decreased in 3rd autumn and 2nd winter month. Cutting off the sample according to mother's age on giving birth, the conceptions of Down Ss born from less than 36 year old mothers present the same seasonal distribution (.0009), while in more than 35 year old mothers' sons a seasonal influence seems masked by the well-known maternal age effect.
Strong significant correlation exists between the monthly distribution of the whole sample conceptions and the distribution of the children born from less than 36 years mothers, while this does not occurs for children of mothers with 36 or more years at delivery.

Key words: Down syndrome; epidemiology; conception; seasonal influence, stress.

In a first research [3] we found a higher prevalence of Down's syndrome births in the months of July and November, while there was a reduction in September. Also with regards probable months of conception, a higher number of conceptions were found in the months of August and October, while deficit was evident in the months of July, November and December.

Although having focused attention to this discordance, the subject was merely touched upon rather than attempting a deeper analysis, also because of our awareness that the methods with which we had investigated the problem were too simplistic. We returned to the subject in a following investigation with more rigorous criteria and a larger sample [2]. Now I come back to it with an increased sample and 20 years' period of survey.

Materials and methods

The medical records of 535 Ss affected by Down's syndrome were reexamined. These constitute a consecutive non-selected series of 21 trisomic Ss, coming from all parts of Italy and which I visited personally, between January 1979 and April 1997 inclusive. I extracted the records of those individuals born between 1973 and 1992 and the following information was collected:

sex; date of birth; chromosomal diagnosis; period, expressed in weeks, of any non-pathological or pathological prematurity or postmaturity (reference: 40 weeks); maternal age on giving birth

The Ss were then subdivided according to the month in which they were born.

With the help of an obstetrician's rule, starting from the day of birth and adding or subtracting any weeks of prematurity or postmaturity, I identified the presumable day of conception.

Using the available ISTAT yearbooks [6], I found the total number of living births in Italy, month by month, between 1973 and 1992 inclusive. From each of these monthly totals, 1/700 was subtracted which represents the total number of Down's syndrome births for that same period. It is known that approximately one new-born child in 700 suffers from Down's syndrome. It was assumed that prematurity and postmaturity in non-Down babies are both distributed equally, each one in itself throughout the entire population.

Given that a birth is always preceded by a conception (though the opposite is not sometimes true) and calculating that this conception took place 280 days earlier, from the monthly totals of such new-borns I was able to arrive at the number of respective conceptions.

The monthly conception period was considered to be between the 8th day of a month and the 7th day of the next one; (e.g.: Punctual births or those which, if not premature or postmature, should have happened in the month of January, had their conceptions between 8th of April and 7th of May in the preceding year).

I therefore reconstructed the monthly total of conceptions of non-Down living births using 20-23 parts (= days) of the total number of births in a specified month ( according to its total days number) and 7 parts (= days) of the total number of births in the following month, moving the period back 280 days.

For example: Of those born in the January period, about 3/4th were conceived in last 22 days of the previous April period, while about 1/4th were conceived in first 7 days of the previous May. So, remembering that conception usually takes place 9 months and 10 days previously, the total number of conceptions in the September 8-October 7 period, through these calculations, constitutes about 3/4th of all the living births in the month of December between 1973 and 1992, plus 1/4th of all the living births in the month of January between 1973 and 1992.

The conceptions dates of all the Down subjects which make up our sample were subdivided according to the same monthly time periods (e.g.: April 8 - May 7).

We then further subdivided the Down subject conceptions on the basis of the mother's age when the child was born. In order to do this we chose the age of 36 as a limit, which corresponds approximately with a conception when the mother was 35 year old.

This age is considered by obstetricians to be the starting point of increased risk and therefore a prenatal diagnosis is advised. To evaluate the significance of the various factors we used Chi Square test and "r" coefficient of correlation.

Results

The results of this research are as follows:
Suitable Ss for the investigation: 487 out of 535 = (91.03%)
This is due to the fact that 40 Ss were born between 1944 and 1972 and 8 Ss after 1992, therefore all they do not fall within the period in question.
Gender distribution: M: 284 Ss = 57.29%
F: 203 Ss = 42.71%
sex ratio M/F: 137.44
Distribution of the chromosomal anomalies:
standard trisomy 21: 446 Ss = 91.59%
mosaicisms: 16 Ss = 3.28%
translocations: 16 Ss = 3.28%
unknown: 9 Ss = 1.85%
Table 1 shows the stratification of the sample as regards the year of birth.

Table 1: stratification and rates of the sample subjects according to the year of birth
  Year   no.
 of Ss 
%   Year   No.
 of Ss 
%
1973 23 4.72 1983 35 7.19
1974 21 4.31 1984 41 8.42
1975 21 4.31 1985 29 5.95
1976 26 5.34 1986 22 4.52
1977 23 4.72 1987 19 3.90
1978 26 5.34 1988 17 3.49
1979 25 5.13 1989 18 3.70
1980 26 5.34 1990 15 3.08
1981 37 7.80 1991 10 2.05
1982 40 8.21 1992 13 2.67
The total number of births per months, whether Down subjects considered in this sample or non Down, is presented in Table 2.

Table 2: Monthly distribution and rates of the whole population of Italian livebirths in 1973-1992; the same minus 1/700th, as the rate of Down births; monthly distribution and rates of births of sample subjects, all born in the same time period.
Month Live births (tot) Tot. -1/700th % Sample births %
January 1087867 1086312 8.27 42 8.62
February 986228 984819 7.50 39 8.01
March 1097166 1095599 8.35 40 8.21
April 1051147 1049645 8.00 44 9.03
May 1166822 1165155 8.88 40 8.21
June 1118938 1117340 8.51 39 8.01
July 1179041 1177357 8.97 50 10.26
August 1126481 1124872 8.57 37 7.60
Septem. 1115636 1114042 8.49 32 6.57
October 1187283 1185587 9.02 37 7.60
Novemb. 1009827 1008384 7.68 49 10.06
Decemb. 1020145 1018688 7.76 38 7.80
Totals   13127800 100.00 487 100.00
Chi Square = 9.615 with 11 df and p = .572
The minimax difference in monthly percentages is 1.52 in the whole population vs 3.69 in sample subjects. There is not any significant difference between monthly distributions of the births.

Table 3 lists the total number of monthly conceptions of live births between 8th April 1972 and 7th April 1992 minus 1/700th, together with the total monthly conceptions of our 487 Down subjects, in same periods.

Table 3: Monthly distribution of conceptions of live births non-Down subjects and conceptions of 487 sample subjects from 8 April 1972 to 7 April 1992
 Monthly period  No. of nonDown
conceptions
% No. of sample
conceptions
%
8 Apr. - 7 May 1045179 7.96 47 9.65
8 May - 7 June 1081647 8.24 57 11.70
8.June - 7 July 1055074 8.04 29 5.95
8 July - 7 Aug. 1109608 8.45 42 8.62
8 Aug. - 7 Sept. 1158445 8.82 47 9.65
8 Sept.- 7 Oct. 1112958 8.48 50 10.27
8 Oct. - 7 Nov. 1158967 8.83 44 9.03
8 Nov. - 7 Dec. 1091323 8.31 25 5.13
8 Dec. - 7 Jan. 1162007 8.85 32 6.57
8 Jan. - 7 Feb. 1138245 8.67 43 8.83
8 Feb. - 7 Mar. 960913 7.32 36 7.39
8 Mar. - 7 Apr. 1053434 8.02 35 7.19
Totals 13127800 100.00 487 100.00
Chi Square = 22.952; df = 11; p < 0.018
The distribution of the conceptions significantly differs. The minimax difference in monthly percentages is 1.53 in the whole population vs 6.57 in sample subjects.

Table 4 shows data referring to the conceptions of 472 of our 487 Down subjects, grouped as far as the mother's age on giving birth is concerned, having placed the cut-off at 36 years of age.

Table 4: Monthly distribution of conceptions of live births non-Down Ss and conceptions of 472 (*) sample Ss according to mother's age at delivery
 Monthly period  a b c d
No. of non No. of sample Sample Ss: maternal age
Down Ss Ss < 36 years ≥ 36 years
8 Apr. - 7 May 1045179 47 34 13
8 May - 7 June 1081647 55 35 20
8 June - 7 July 1055074 28 12 16
8 July - 7 Aug. 1109608 40 30 10
8 Aug. - 7 Sep. 1158445 46 22 24
8 Sep. - 7 Oct. 1112958 50 32 18
8 Oct. - 7 Nov. 1158967 41 23 18
8 Nov. - 7 Dec. 1091323 23 9 14
8 Dec. - 7 Jan. 1162007 31 17 14
8 Jan. - 7 Feb. 1138245 41 26 15
8 Feb. - 7 Mar. 960913 35 12 23
8 Mar. - 7 Apr. 1053434 35 24 11
Totals 13127800 472 276 196
(*) maternal age at delivery was lacking in the records of 15 Ss.

Chi Square: a vs b = 22.081; df = 11 and p = .011
a vs c = 37.636; df = 11; p < 0.0009
vs d = 14.545; df=11; p = .205 N.S.a

"r" coefficient of correlation:
[b vs c: r = .885; t = 6.013; df = 10; p < .0009]*
[b vs d: r = . 333; t = 1.122; df = 10; p = .288]*
c vs d: r = - 0.143; t = - .456; df 10; p = .658

*being c and d subsamples of b, tests are debatable

Down individuals born to mothers of less than 36 years old parallels the same seasonal trend with regards conceptions and this distribution leads to a highly significant Chi Square figure. The same cannot be said of Down subjects born to mothers of 36 years or older. A non significant correlation was also found between the monthly conception distribution of all 487 Down subjects in this sample and the monthly distribution of those born to mothers with 36 years and more, although the latter is a subsample of the former. There is non significant correlation between the distributions of the two subsamples.

Discussion

The M/F ratio, which slightly exceed the Italian trends at birth [1]; the distribution of chromosomal diagnoses which is well within both Italian and International averages [1;5;9-10], the nationwide provenance indicate that these 487 cases can be considered a representative sample of Italian Down subjects.

The results punctually confirm the previous research [3]. A seasonal influence on a biological process such as the one leading to chromosomal non disjunction during the gamete formation phase (meiotic phase) and during the first mitotic divisions (for mosaic forms), should not seem either impossible or extraordinary.

If we look at schizophrenia, it has been reported since more than 20 years that there exists a significant but unexplainable excess of cases born in the winter months.

This is a definite fact, confirmed and reconfirmed by a whole series of researchers [8], and which is difficult to reconcile with a purely psychological etiology.

On the other hand, being the chromosomal anomaly in Down Ss a process that has its developing point in the conception, although the anomaly in itself nearly always has a gamete origin, it seems possible to approximate the problem only by taking into consideration the conception's periods. Some investigations based on birth's periods conducted to debatable results [7;11-13], in my opinion, due to the excess presence of prematurity among deliveries of Down Ss as well as to the additional or prevalent effect of the maternal age. As far as this survey is concerned, I found an excess of trisomy 21 conceptions in spring months and in the first two autumn months.

Together with this we found two other peaks six months nearly apart from each other, namely August-September, and January-February.

The seasonal trend for Down subjects born to mothers under 36 years old is almost super impossible to conceptions' distribution of the whole sample and significantly different from the distribution of non Downs' conceptions. Regarding the subjects born to mothers of 36 years or older, no significant Chi Square figure and coefficient of correlation have been found in comparison with the seasonal trend for the population as a whole. This leads to infer that in this subgroups the seasonal influence on conceptions, if any, is completely masked by the well known maternal age effect.

The causes of incomplete chromosomal disjunctions, in Down subjects too, are still elusive [4], but if a seasonal influence should be finally confirmed, the field of stress reactions, mainly the parasympathetic ones, seems to me the next step to be investigated, even in 36 years and older mothers.

Conclusion

This investigation into the influence on the conception of Down subjects, as far as our 487 Italian individuals born in 1973-1992 are concerned, leads to confirm a different seasonal distribution in conceptions of 21 trisomics as compared to the seasonal distribution of all non-Down living births in Italy in the same time-period. This divergence was seen significant both for the whole sample and the sub-group of subjects born to mothers of less than 36 years old, but not for subjects born from mothers of 36 years and older. These results suggest that there exists a complex influence which we still are unable to well understand in its various aspects.

References

[1] Camera G., Mastroiacovo P.: Epidemiologia della Sindrome di Down. In:Ce.Pi.M.: Aspetti epidemiologici, genetici, clinici, riabilitativi e sociali della Sindrome di Down. Ce.Pi.M., Genova 1984: 225-230.

[2] Cocchi R., Occhialini O.: Is there any seasonal influence in the conceptions of Down's syndrome subjects? It. J. Intellect. Impair. 1989, 2: 1-7.

[3] Cocchi R., Occhialini O., Cocchi Cercolani P.: Alcuni dati epidemiologici su una serie consecutiva, non selezionata, di 241 soggetti Down. Rass. Studi Psichiat. 1985, 74: 785-790.

[4] Editor's writing: The elusive cause of Down's Syndrome. Lancet 1983, May 21: 1143-1144.

[5] Hook E.B.: Down Syndrome: Frequency in human population and factors pertinent to variation in rates. In: De La Cruz F.F., Gerald P.S. (eds): Trisomy 21 (Down Syndrome) research perspectives. University Park Press. Baltimore 1981.

[6] ISTAT: Annuari di Statistiche Demografiche 1973-1992. ISTAT, Roma 1974-1993.

[7] Iselius L., Lindsten J.: Changes in the incidence of Down syndrome in Sweden during 1968-1982. Hum. Genet. 1986, 72: 133-139.

[8] Kendall R.E., Kemp I.W.: Winter-born v summer-born schizophrenics. Br. J.Psychiat. 1987, 151: 499-505.

[9] Lambert J.L., Rondal J.A.: Le mongolisme. Mardaga, Bruxelles 1979.

[10] Lindsten G., Marsk L., Berklund K., Iselius L., Ryman N., Anneren G., Kjessler B., Mitelman F., Walstroem J., Vejlens L.: Incidence of Down's Syndrome in Sweden during the years 1968-1977. In: Burgio G.R., Fraccaro M., Tiepolo L., Wolf U. (eds): Trisomy 21. Human Genet. 1981 (suppl. 2): 195-210.

[11] Rothman K.J., Fabia J.J.: Place and time aspects of the occurence of Down's syndrome. Am. J. Epidem. 1976, 103: 560-564.

[12] Seifert C., Sommer A.: A summertime peak of Down's syndrome in Franklin County, Ohio. Am. J. Dis. Child. 1986, 140: 822-824.

[13] Videbech P., Neilsen J.: Chromosomal anomalies and season of birth. Human Gen. 1984, 65: 221-231.