Awn. Hum. Genet., Lond. (1978), 42, 153
153
Printed in. Greut Britain
Reproductive compensation in families segregating for Cooley's anaemia in Ferrara BY SILVANA AGUZZI," CALOGERO VULLOt AND ITALO BARRAI" * Department of Zoology, University of Ferrara, Ferrara, Italy, and t Microcythemia Centre, Arcispedab St Anna, Ferrara, Italy Cooley's anaemia, is prevalent in the province of Ferrara and nearby Rovigo, in the Po delta; more than 250 children are being regularly treated at the thalassaemia centre in Ferrara (Cucchi, Vullo & Barrai, 1977). The gene frequency for thalassaemia in the province of Ferrara, obtained from data published by Silvestroni & Bianco (1975), was estimated at about 4.5 yo,appreciably lower than previous estimates based on earlier data (Barrai & Mandel, 1971). The gene frequencies encompass a range which goes from 1.6 to 9.7 %, in an irregular gradient which increases from the western side of the province toward the Adriatic. Over the province, one person in eleven may be heterozygous and, under random mating, about one marriage in 120 is expected to involve two carriers. Since a generation ago the thalassaemia centre has specialized in the treatmen$ of ohildren affected by the anaemia; at present, children reach their teens, and the improvements in transfusion techniques hold promise that most will reach young-adult ages. Treatment of children is free for the families; however, the presence of one child affected by the disease exacts a heavy toll from the parents, who soon learn that there is no final recovery from the disease. When a child is diagnosed as affected by the anaemia the parents are given genetic counselling, are informed of the chances they have of producing another affected child in subsequent pregnancies, and of the available options to prevent it. The genetic counselling is then retrospective. Although screening a t school ages is practised almost every year in the province, only recently has a group of individuals large enough to study the effects of prospective genetic counselling become available; research on this subject is in progress. The families segregating for Cooley's anaemia are instead a captive group, which is in periodic contact with the centre. I n 1975 a local organization of these families was created, with the purpose of continuous information on the possible courses of action foi the control of the disease. From 1951 to 1970,423 children have been diagnosed at the centre as affected by the anaemia (see also Barai & Vullo, 1976); their distribution per year of birth is given in Table 1. Since.the children come also from Rovigo, their distribution is compared with the total births in the provinces of Ferrara and Rovigo. There is no obvious change of the incidence per year (xFI9,= 21.42, P < 0.40); if genetic counselling were to grow more and more effective, one would expect a change in t)hecourse of time, although it might take much longer than a generation to observe a significant change. Also, retrospective genetic counselling would have an effect on the births subsequent to the affected; we wanted then to identify, if possible, the characteristics of the reproductive behaviour of couples who have produced at least one child with Cooley's anaemia. The purpose of the present work is the assessment of such behaviour.
154
SILVANA ACUZZI,C. VULLOAND I. BARRAI Table 1. Diagnosis of C00ley’s anaemia per year at the thalassaemia centre of Fermra, limited to children born in the provinces of Ferrara and Rovigo, 1951-1970
Year
I951 I952 I953 I954 I955 I956 I957 1958 ‘959 I 960 1961 I 962 I963 I964 I965 1966 1967 I 968 I969 I970 Total
Diagnoses 31
26 22
23 13 21
Births I 3 402 12848 I 3 221 12811 12631
Proportion (per thousand) diagnoses/birtlis 2‘ 3 2’0
2’4
1.8 I ‘0
I 2 802
I *6
2.6 1‘7 I .6
20
12405 11631 I I 650 1 1 593 10901 10818 I I 063 10989 10383
21
10215
14
9523 9 I97 9278 8714 226075
32 20
I9 23
I9 17 I4 31
21
20
16 423
dsl = 21-42.
2’0
1’7 I .6 1’3
2.8 1’9 2’ I
1‘5
2.3 2’2
1.8 1‘9
P < 0.40.
MATERIALS AND METHODS
One hundred families were selected by order of presentation for transfusion of a child at the thalassaemia centre, starting from 1 November 1974. At the time of transfusion the children are accompanied by one of the parents, usually the mother. The accompanying parent was interviewed by one of us (S.A.), using a questionnaire where questions were asked about the reproductive performance of the couple. The main questions referred to such variables as the dates of birth of the parents, the date of their marriage, the dates of birth of the children, their sex, and the dates of such events as stillbirths, miscarriages and infant deaths. The informants were also asked for the address of any married sibs they and their spouses might have. I n so doing we obtained the addresses of 127 married sibs either of the father cr the mother, which constituted our control group, the families against which to compare the reproductive performance of the couples having at least one child affected by Cooley’s anaemia and being treated at the centre. We believe that a group formed by the families of the sibs of the parents is as close as possible socially, culturally and genetically to the sample of segregating families, and therefore has a high chance of constituting the most adequate control group for comparison of reproductive behaviour of the segregating families. Throughout this paper we shall use the adjective ‘thalassaemic’ to indicate the families segregating for Cooley’s anaemia and ‘control’ to indicate the families of the sibs of the parents of the affected children. The control families were asked to fill in a questionnaire by mail; the questionnaire was the
Reproductive compensation in families segregating f o r Cooley’s anaemia
155
Table 2. Distvibution of the number of births in thalassaemic and control families Number of families 7 7
Number of births
Thalassaemic
Control
I 2
27 39
3
I7
35 53 30
4
I0
6
5 6
5
3
7
I
Total
I
I00
127
same as the one used for the thalassaemic families. Answers obtained by mail were cross-checked when t,he number of children in the control families did not match with the number declared for them by the thalassaemic sib/sib-in-law.
Variables available from the questionnaires From the questionnaires, we extracted three types of variables. The first type obviously refers to the reproductive performance of the couple; it includes the number of children ever born, the number of living children by sex, diagnosis and order of birth. The second group is composed by those variables which might have an eflect on fertility; they are the ages of spouses at marriage, duration of marriage, and the intervals between the births of adjacent children. The third group refers specifically to those variablw which might be ajfected by the birth of a child with GooZey’s anaemia, such as the number of children born afkr an affected child, the interval between the affected child and the preceding and following sib, the interval between the first and second sib when the first is affected, and the interval between the first and second sib when the second is affected. RESULTS AND DISCUSSION
Reproductive performance In the following we study the reproductive performance, based on the number of births in thalassaemic and control families. The results are summarized in Tables 2 and 3. The average number of births is not significantly different between thalassaemics and controls; thalassaemics have an average of 2-34 births and controls 2.13; the t for comparison is 1.44,not significant. But the variance is 1-60 for thalassaemics and 0.90 for controls. The variance ratio, F = 1.77, is significant,with 99 and 126 degrees of freedom at the 1 % level. Although means are not significantly different, variances are. To test further for a difference in the shape of the distributions, we fitted a truncated Poisson distribution to both distributions. There is a good fit of the Poisson model to the thalassaemics (x$ = 6.86, P < 0.30) whereas there is no fit for the controls = 14.95, P < 0.005). Th0 indication obtained from fitting the same distribution to both groups is that the shapes obsemed are significantly different and may reflect a difference in the reproductive behaviour. It seems that in controls reproduction is arrested after a preferred number of children is obtained; this is also
o&
156
SILVANA AGUZZI,C. VULLOAND I. BARRAI Table 3. Distribution of living children in thalassaemic and control families Number of families Number of children Thalassaemic 29
45 16 7 2
Control 39 51
28 6 3
I
Total
I00
'27
indicated by the narrower variance of the controls. Tn the thalwssaemics, reproduction continues toward a higher number of births. We studied the same distributions considering only living children instead of children ever born. The distributions are given in Table 3. The average number of children is the same in both distributions :thalassaemics have an average of 2-11 living children and controls of 2.08; the variances are 1.07 and 0.93 respectively, not significantly different, Neither distribution fits a truncated Poisson distribution. It seems that the distribution of living children is similar in thalassaemic and control families; the same shape of distribution is achieved by the thalassaemic families through loss of children. Variables which might affect fertility Since the shape of the distribution of children ever born is different in the two groups, and skce the main difference in the two groups is the presence in one group of families of at least one child with Cooley's anaemia in the family offspring, we consider that there is some degree of compensative reproduction in the thalassaemic families. However, such a compensation would require a reproductive behaviour which might be also influenced by other variables which may affect fertility and fecundity. The most important among such variables are the following : (1) Age a t marriage in males and females. (2) Difference in age at marriage between spouses. (3) Age of the mother at the birth of the fist child. (4)Duration of marriage. In Table 4 we give the comparisons between thalassaemic and control families for the four variables considered. It is apparent, from the observed data, that there is no difference between ages at marriage in thalassaemics and controls, nor in the age difference between spouses, nor in the age of mothers at first birth, nor in the duration of marriage. The two samples seem very similar for these demographic variables. We also regressed the number of children born on the duration of marriage; the parameters of the regressions are given in Table 5. Although the regression coefficient of the thalassaemics is larger, the difference is not significant. The indication is toward a faster rate of reproduction per time unit in thalassaemics than in controls.
Reproductive compensation in !families segregating for Cooley’s anaemia
157
Table 4.Averages of demographic variables in thalassaemic and control families (in years) Age a t marriage
Thalassaemic Control t
Males
Females
25.8
22.5 22.3
25.6
3’3 3‘3
0.36
0.05
Duration of marriage a t I . i. 75
Age difference Age of mother husband - wife at first birth 23’3 23-2
-0.14
15’4
16.2
0.30
0.73
The tot.al number of degrees of freedom for the comparisons of Table 4 is 223.
Table 5. Parameters of the regression of the number of children born on the duration of marriage in m o n t h Average
, Group
Children
Duration
a
b
Thalassaemic
2’34
5’6
2.13
0.93 1-36
0.0076
(:(JllkOlS
184.8 193‘7
0.005 I
5’1
t
for regression
Table 6. Number of families as a function of the order of birth of the first affected child ever born and sibship size attained at time of investigation Order of birth of child with Cooley’s anaemia
Sibship size A
7
1
I
2
3
4
5
6
7
Total
27
39
I7
I0
5
I
I
I00
Ist
2nd 3rd 4th 5th Total
This is not confirmed, however, by the study of the interval between two successive births in thalassaemics and controls. The average interval, independent of family size and birth order, was 55-2 months in thalassaemics and 51.7 months in control families. Aa important point comes, however, from the comparison of the variances of birth intervals: the variance is 1509.9 in thalassaemics and 1042.6 in controls, indicating a tendency to clustering of birth intervals in thalassaemics ; the variances are significantly different.
Variables aflected by the birth of a child with Cooley’s anaemia The number of families segregating for Cooley’s anaemia is tabulated by sibship size and by the order of birth of the first affected in Table 6. The families on the main diagonal of the table are those in which the affected appears at the last place; however, since the families in the sample may not be completed, the child with Cooley’s anaemia in these families is at the last place at the time of the investigation. The families above the main diagonal are those which have continued reproduction after the birth of the first affected child; there are 35 such families, and 65 which arrested reproduction with the appearance of the affected, limited to the time of the investigation. If we exclude the 27 families of size 1, 38 families have more than one child, and the affected is the last born.
158
SILVANA AGUZZI,C. VULLOAND I. BARRAI Table 7. Comparisons between the birth intervals (inmonths) between the affected and the adjacent sibs Interval between
Average
Affected and second sib Normal and second sib
48.13 58.00
6.47
Sib and first affected First affected and sib
63-23 47'09
6'or} 4'73
S.B.
Difference
Student's t
9.87
1'11
16.14
2'11
P X o t signif.
< 0.05
It would appear that there is a tendency towards two types of behaviour: the one toward continuation of reproduction if the first child with the anaemia appears at the first or at an early pregnancy, and the other towards arresting reproduction if the child appears at a later pregnancy. The expected frequency of the birth order of the first affected child is rectangular provided that the occurrence of an affected child does not influence reproductive behaviour, na,mely we expect that for families of constant size the number of families with the affected at the first place is the same as the number of those with the affected at the second place and so on. I n our data we observe that there is an excess of families with the first affected at the last place ; and the excess is significant for families of size 2 = 4.33, P < 0.05). The excess of families of size 3 with the affected a t the la& place, although indicative, is not statistically significant. This would suggest that in these families further reproduction has been discouraged by the birth of an affected child. For larger family sizes it seems that the &st affected appears at random in any order ; these families seem those which would continue reproduction to replace the child. We examined the profession of the parents in such families but there was no indication that they were different from other families in this respect. We obtained further indications from the study of the intervals between the birth of one child affected by Cooley's anaemia and the birth of subsequent children.
(xe,
Intervals between adjacent births We had available 131 intervals between two successive births in the thalassaemic families and 137 intervals in the control families. As noted above, the average intervals were 55.2 & 3.4 months for thalassaemic families and 51.7 & 2.8 for control families. The difference was not significant; the variance was found to be significantly larger in the thalassaemic families, indicating a tendency to clustering in successive births. We studied the interval between births in thalassaemic families as a function of the birth of a child affected by Cooley's anaemia, in those families having more than one birth. We compared the birth interval between the first- and the second-born when the fist-born was affected and the second-born a normal sib, with the birth interval when the fist-born was norms1 and the second affected; and the interval between the affected and the following sib with that between the affected and the preceding sib, irrespective of the order of birth. The results of the comparisons are given in Table 7. The difference between the birth intervals when the Cooley is the first-born and the interval when the affected is the second born is not significant. There is indication that the interval is shorter when the first child is affected than when it is normal.
Reproductive compensation in families segreggting for Cooley’s anaemia
159
The difference between the birth interval normal-&ected and the interval affected-normal, irrespective of the order and family size, is significant at the 5 % level. It would appear that, if a decision for another birth is taken, the child is replaced sooner than on average ; and we believe that this behaviour also indicates that there is some degree of reproductive compensation in these data. CONCLUSION
We believe that the main indication which emerges from our analysis is that the families segregating for Cooley’s anaemia show two main patterns of reproductive behaviour after the birth of the affected child. If the affected child is at the last, or at one of the late places in a sibship of a desired size, reproduction is discontinued; if the child is at an early place, or at the first place, and replacement is decided upcn, the birth is planned sooner thacnit would have been had the child not been affected. We are aware that, when a family is large, it is likely to be terminated in any case, be the child affected or not; however, the excms of families terminating with an affected child in these data seems to indicate a systematic behaviour. Our conclusion is based on the following facts: the distribution of births in thalassaemics fits a truncated Poisson distribution whereas the controls do not; the variance is higher in thalassaemics, and significantlyso. The thalassaemic and control parents do not differ in the demographic parameters which could affect fertility and fecundity. The thalassaemic families who replace a child do so sooner than the average birth interval. We tentatively advance that some degree of reproductive compensation is operating in thalassaemic families as a consequence of the presence of an affected child; it might also be advanced that the role, if any, of the genetic counselling to which the families have been exposed was not predominant in determining their reproductive behaviour. This is in no way a negative conclusion for activities of genetic counselling, and we are well aware t h a t the observed behaviour may be limited in time to the particular set of families, and in space to the area of the Po delta. SUMMARY
The reproductive performance of 100 families segregating for Cooley’s anaemia was studied and compared to the performance of 127 control families. The control families were those of the sibs of the parents in the segregating families. No difference was found in the variables which might affect fertility and fecundity in control and thalassaemic families, although a different shape of the distribution of the number of births per family was found ;the distribution had a higher variance in thalassaemicsthan in controls. The interval between the birth of a normal and an affected child is significantly longer than the interval between the birth of an affected child and a subsequent one. It was thought that these findings are indicative of reproductive compensation in families segregating for Cooley’s anaemia, and of the moderate effect of the genetic counselling given to them. This work was possible because of the co-operation of the parents of the affected children. To them, and to the medical and paramedical personnel that over a generation has helped the children at the Microcythemia Centre, the work is respectfully dedicated. The work was supported by grants from the World Health Organization and the Italian NRC to the Dopartnient of Zoology at the University of Ferrara; and by a grant of the Province of Ferrara to the Microcytheniia Centre at the Arcispedale St Anna in Ferrara. I1
HGE
42
160
SILVANAAGUZZI,C. VULLOAND I. BARRAI REFERENCES
BARRAI, I. & MANUEL,S. P. H. (1971). Equilibrium under inbreeding in balanced polymorphism. Bulletin of the World Health Organization 45, 201-8. BARRAI, I. & VULLO, C. (1976). La funzione del consultorio genetic0 nella prevenzione del morbo di Cooley. Atti V I Giornate Mediehe Daune, Foggia, 27-29 Maggio. CUCCHI,P., VULLO, G. & BARRAI, I. (1977). Population genetics in the province of Ferrara. 11. Survival of children with Cooley’s anemia. American Journal of Human. Genetics 29, 178-83. SILVESTRONI, E. & BIANCO, I. (1976). Screening for microcythemia in Italy: analysis of data collected in the past 30 years. American Journal of Human Cenetica 27, 198-212.
153
Printed in. Greut Britain
Reproductive compensation in families segregating for Cooley's anaemia in Ferrara BY SILVANA AGUZZI," CALOGERO VULLOt AND ITALO BARRAI" * Department of Zoology, University of Ferrara, Ferrara, Italy, and t Microcythemia Centre, Arcispedab St Anna, Ferrara, Italy Cooley's anaemia, is prevalent in the province of Ferrara and nearby Rovigo, in the Po delta; more than 250 children are being regularly treated at the thalassaemia centre in Ferrara (Cucchi, Vullo & Barrai, 1977). The gene frequency for thalassaemia in the province of Ferrara, obtained from data published by Silvestroni & Bianco (1975), was estimated at about 4.5 yo,appreciably lower than previous estimates based on earlier data (Barrai & Mandel, 1971). The gene frequencies encompass a range which goes from 1.6 to 9.7 %, in an irregular gradient which increases from the western side of the province toward the Adriatic. Over the province, one person in eleven may be heterozygous and, under random mating, about one marriage in 120 is expected to involve two carriers. Since a generation ago the thalassaemia centre has specialized in the treatmen$ of ohildren affected by the anaemia; at present, children reach their teens, and the improvements in transfusion techniques hold promise that most will reach young-adult ages. Treatment of children is free for the families; however, the presence of one child affected by the disease exacts a heavy toll from the parents, who soon learn that there is no final recovery from the disease. When a child is diagnosed as affected by the anaemia the parents are given genetic counselling, are informed of the chances they have of producing another affected child in subsequent pregnancies, and of the available options to prevent it. The genetic counselling is then retrospective. Although screening a t school ages is practised almost every year in the province, only recently has a group of individuals large enough to study the effects of prospective genetic counselling become available; research on this subject is in progress. The families segregating for Cooley's anaemia are instead a captive group, which is in periodic contact with the centre. I n 1975 a local organization of these families was created, with the purpose of continuous information on the possible courses of action foi the control of the disease. From 1951 to 1970,423 children have been diagnosed at the centre as affected by the anaemia (see also Barai & Vullo, 1976); their distribution per year of birth is given in Table 1. Since.the children come also from Rovigo, their distribution is compared with the total births in the provinces of Ferrara and Rovigo. There is no obvious change of the incidence per year (xFI9,= 21.42, P < 0.40); if genetic counselling were to grow more and more effective, one would expect a change in t)hecourse of time, although it might take much longer than a generation to observe a significant change. Also, retrospective genetic counselling would have an effect on the births subsequent to the affected; we wanted then to identify, if possible, the characteristics of the reproductive behaviour of couples who have produced at least one child with Cooley's anaemia. The purpose of the present work is the assessment of such behaviour.
154
SILVANA ACUZZI,C. VULLOAND I. BARRAI Table 1. Diagnosis of C00ley’s anaemia per year at the thalassaemia centre of Fermra, limited to children born in the provinces of Ferrara and Rovigo, 1951-1970
Year
I951 I952 I953 I954 I955 I956 I957 1958 ‘959 I 960 1961 I 962 I963 I964 I965 1966 1967 I 968 I969 I970 Total
Diagnoses 31
26 22
23 13 21
Births I 3 402 12848 I 3 221 12811 12631
Proportion (per thousand) diagnoses/birtlis 2‘ 3 2’0
2’4
1.8 I ‘0
I 2 802
I *6
2.6 1‘7 I .6
20
12405 11631 I I 650 1 1 593 10901 10818 I I 063 10989 10383
21
10215
14
9523 9 I97 9278 8714 226075
32 20
I9 23
I9 17 I4 31
21
20
16 423
dsl = 21-42.
2’0
1’7 I .6 1’3
2.8 1’9 2’ I
1‘5
2.3 2’2
1.8 1‘9
P < 0.40.
MATERIALS AND METHODS
One hundred families were selected by order of presentation for transfusion of a child at the thalassaemia centre, starting from 1 November 1974. At the time of transfusion the children are accompanied by one of the parents, usually the mother. The accompanying parent was interviewed by one of us (S.A.), using a questionnaire where questions were asked about the reproductive performance of the couple. The main questions referred to such variables as the dates of birth of the parents, the date of their marriage, the dates of birth of the children, their sex, and the dates of such events as stillbirths, miscarriages and infant deaths. The informants were also asked for the address of any married sibs they and their spouses might have. I n so doing we obtained the addresses of 127 married sibs either of the father cr the mother, which constituted our control group, the families against which to compare the reproductive performance of the couples having at least one child affected by Cooley’s anaemia and being treated at the centre. We believe that a group formed by the families of the sibs of the parents is as close as possible socially, culturally and genetically to the sample of segregating families, and therefore has a high chance of constituting the most adequate control group for comparison of reproductive behaviour of the segregating families. Throughout this paper we shall use the adjective ‘thalassaemic’ to indicate the families segregating for Cooley’s anaemia and ‘control’ to indicate the families of the sibs of the parents of the affected children. The control families were asked to fill in a questionnaire by mail; the questionnaire was the
Reproductive compensation in families segregating f o r Cooley’s anaemia
155
Table 2. Distvibution of the number of births in thalassaemic and control families Number of families 7 7
Number of births
Thalassaemic
Control
I 2
27 39
3
I7
35 53 30
4
I0
6
5 6
5
3
7
I
Total
I
I00
127
same as the one used for the thalassaemic families. Answers obtained by mail were cross-checked when t,he number of children in the control families did not match with the number declared for them by the thalassaemic sib/sib-in-law.
Variables available from the questionnaires From the questionnaires, we extracted three types of variables. The first type obviously refers to the reproductive performance of the couple; it includes the number of children ever born, the number of living children by sex, diagnosis and order of birth. The second group is composed by those variables which might have an eflect on fertility; they are the ages of spouses at marriage, duration of marriage, and the intervals between the births of adjacent children. The third group refers specifically to those variablw which might be ajfected by the birth of a child with GooZey’s anaemia, such as the number of children born afkr an affected child, the interval between the affected child and the preceding and following sib, the interval between the first and second sib when the first is affected, and the interval between the first and second sib when the second is affected. RESULTS AND DISCUSSION
Reproductive performance In the following we study the reproductive performance, based on the number of births in thalassaemic and control families. The results are summarized in Tables 2 and 3. The average number of births is not significantly different between thalassaemics and controls; thalassaemics have an average of 2-34 births and controls 2.13; the t for comparison is 1.44,not significant. But the variance is 1-60 for thalassaemics and 0.90 for controls. The variance ratio, F = 1.77, is significant,with 99 and 126 degrees of freedom at the 1 % level. Although means are not significantly different, variances are. To test further for a difference in the shape of the distributions, we fitted a truncated Poisson distribution to both distributions. There is a good fit of the Poisson model to the thalassaemics (x$ = 6.86, P < 0.30) whereas there is no fit for the controls = 14.95, P < 0.005). Th0 indication obtained from fitting the same distribution to both groups is that the shapes obsemed are significantly different and may reflect a difference in the reproductive behaviour. It seems that in controls reproduction is arrested after a preferred number of children is obtained; this is also
o&
156
SILVANA AGUZZI,C. VULLOAND I. BARRAI Table 3. Distribution of living children in thalassaemic and control families Number of families Number of children Thalassaemic 29
45 16 7 2
Control 39 51
28 6 3
I
Total
I00
'27
indicated by the narrower variance of the controls. Tn the thalwssaemics, reproduction continues toward a higher number of births. We studied the same distributions considering only living children instead of children ever born. The distributions are given in Table 3. The average number of children is the same in both distributions :thalassaemics have an average of 2-11 living children and controls of 2.08; the variances are 1.07 and 0.93 respectively, not significantly different, Neither distribution fits a truncated Poisson distribution. It seems that the distribution of living children is similar in thalassaemic and control families; the same shape of distribution is achieved by the thalassaemic families through loss of children. Variables which might affect fertility Since the shape of the distribution of children ever born is different in the two groups, and skce the main difference in the two groups is the presence in one group of families of at least one child with Cooley's anaemia in the family offspring, we consider that there is some degree of compensative reproduction in the thalassaemic families. However, such a compensation would require a reproductive behaviour which might be also influenced by other variables which may affect fertility and fecundity. The most important among such variables are the following : (1) Age a t marriage in males and females. (2) Difference in age at marriage between spouses. (3) Age of the mother at the birth of the fist child. (4)Duration of marriage. In Table 4 we give the comparisons between thalassaemic and control families for the four variables considered. It is apparent, from the observed data, that there is no difference between ages at marriage in thalassaemics and controls, nor in the age difference between spouses, nor in the age of mothers at first birth, nor in the duration of marriage. The two samples seem very similar for these demographic variables. We also regressed the number of children born on the duration of marriage; the parameters of the regressions are given in Table 5. Although the regression coefficient of the thalassaemics is larger, the difference is not significant. The indication is toward a faster rate of reproduction per time unit in thalassaemics than in controls.
Reproductive compensation in !families segregating for Cooley’s anaemia
157
Table 4.Averages of demographic variables in thalassaemic and control families (in years) Age a t marriage
Thalassaemic Control t
Males
Females
25.8
22.5 22.3
25.6
3’3 3‘3
0.36
0.05
Duration of marriage a t I . i. 75
Age difference Age of mother husband - wife at first birth 23’3 23-2
-0.14
15’4
16.2
0.30
0.73
The tot.al number of degrees of freedom for the comparisons of Table 4 is 223.
Table 5. Parameters of the regression of the number of children born on the duration of marriage in m o n t h Average
, Group
Children
Duration
a
b
Thalassaemic
2’34
5’6
2.13
0.93 1-36
0.0076
(:(JllkOlS
184.8 193‘7
0.005 I
5’1
t
for regression
Table 6. Number of families as a function of the order of birth of the first affected child ever born and sibship size attained at time of investigation Order of birth of child with Cooley’s anaemia
Sibship size A
7
1
I
2
3
4
5
6
7
Total
27
39
I7
I0
5
I
I
I00
Ist
2nd 3rd 4th 5th Total
This is not confirmed, however, by the study of the interval between two successive births in thalassaemics and controls. The average interval, independent of family size and birth order, was 55-2 months in thalassaemics and 51.7 months in control families. Aa important point comes, however, from the comparison of the variances of birth intervals: the variance is 1509.9 in thalassaemics and 1042.6 in controls, indicating a tendency to clustering of birth intervals in thalassaemics ; the variances are significantly different.
Variables aflected by the birth of a child with Cooley’s anaemia The number of families segregating for Cooley’s anaemia is tabulated by sibship size and by the order of birth of the first affected in Table 6. The families on the main diagonal of the table are those in which the affected appears at the last place; however, since the families in the sample may not be completed, the child with Cooley’s anaemia in these families is at the last place at the time of the investigation. The families above the main diagonal are those which have continued reproduction after the birth of the first affected child; there are 35 such families, and 65 which arrested reproduction with the appearance of the affected, limited to the time of the investigation. If we exclude the 27 families of size 1, 38 families have more than one child, and the affected is the last born.
158
SILVANA AGUZZI,C. VULLOAND I. BARRAI Table 7. Comparisons between the birth intervals (inmonths) between the affected and the adjacent sibs Interval between
Average
Affected and second sib Normal and second sib
48.13 58.00
6.47
Sib and first affected First affected and sib
63-23 47'09
6'or} 4'73
S.B.
Difference
Student's t
9.87
1'11
16.14
2'11
P X o t signif.
< 0.05
It would appear that there is a tendency towards two types of behaviour: the one toward continuation of reproduction if the first child with the anaemia appears at the first or at an early pregnancy, and the other towards arresting reproduction if the child appears at a later pregnancy. The expected frequency of the birth order of the first affected child is rectangular provided that the occurrence of an affected child does not influence reproductive behaviour, na,mely we expect that for families of constant size the number of families with the affected at the first place is the same as the number of those with the affected at the second place and so on. I n our data we observe that there is an excess of families with the first affected at the last place ; and the excess is significant for families of size 2 = 4.33, P < 0.05). The excess of families of size 3 with the affected a t the la& place, although indicative, is not statistically significant. This would suggest that in these families further reproduction has been discouraged by the birth of an affected child. For larger family sizes it seems that the &st affected appears at random in any order ; these families seem those which would continue reproduction to replace the child. We examined the profession of the parents in such families but there was no indication that they were different from other families in this respect. We obtained further indications from the study of the intervals between the birth of one child affected by Cooley's anaemia and the birth of subsequent children.
(xe,
Intervals between adjacent births We had available 131 intervals between two successive births in the thalassaemic families and 137 intervals in the control families. As noted above, the average intervals were 55.2 & 3.4 months for thalassaemic families and 51.7 & 2.8 for control families. The difference was not significant; the variance was found to be significantly larger in the thalassaemic families, indicating a tendency to clustering in successive births. We studied the interval between births in thalassaemic families as a function of the birth of a child affected by Cooley's anaemia, in those families having more than one birth. We compared the birth interval between the first- and the second-born when the fist-born was affected and the second-born a normal sib, with the birth interval when the fist-born was norms1 and the second affected; and the interval between the affected and the following sib with that between the affected and the preceding sib, irrespective of the order of birth. The results of the comparisons are given in Table 7. The difference between the birth intervals when the Cooley is the first-born and the interval when the affected is the second born is not significant. There is indication that the interval is shorter when the first child is affected than when it is normal.
Reproductive compensation in families segreggting for Cooley’s anaemia
159
The difference between the birth interval normal-&ected and the interval affected-normal, irrespective of the order and family size, is significant at the 5 % level. It would appear that, if a decision for another birth is taken, the child is replaced sooner than on average ; and we believe that this behaviour also indicates that there is some degree of reproductive compensation in these data. CONCLUSION
We believe that the main indication which emerges from our analysis is that the families segregating for Cooley’s anaemia show two main patterns of reproductive behaviour after the birth of the affected child. If the affected child is at the last, or at one of the late places in a sibship of a desired size, reproduction is discontinued; if the child is at an early place, or at the first place, and replacement is decided upcn, the birth is planned sooner thacnit would have been had the child not been affected. We are aware that, when a family is large, it is likely to be terminated in any case, be the child affected or not; however, the excms of families terminating with an affected child in these data seems to indicate a systematic behaviour. Our conclusion is based on the following facts: the distribution of births in thalassaemics fits a truncated Poisson distribution whereas the controls do not; the variance is higher in thalassaemics, and significantlyso. The thalassaemic and control parents do not differ in the demographic parameters which could affect fertility and fecundity. The thalassaemic families who replace a child do so sooner than the average birth interval. We tentatively advance that some degree of reproductive compensation is operating in thalassaemic families as a consequence of the presence of an affected child; it might also be advanced that the role, if any, of the genetic counselling to which the families have been exposed was not predominant in determining their reproductive behaviour. This is in no way a negative conclusion for activities of genetic counselling, and we are well aware t h a t the observed behaviour may be limited in time to the particular set of families, and in space to the area of the Po delta. SUMMARY
The reproductive performance of 100 families segregating for Cooley’s anaemia was studied and compared to the performance of 127 control families. The control families were those of the sibs of the parents in the segregating families. No difference was found in the variables which might affect fertility and fecundity in control and thalassaemic families, although a different shape of the distribution of the number of births per family was found ;the distribution had a higher variance in thalassaemicsthan in controls. The interval between the birth of a normal and an affected child is significantly longer than the interval between the birth of an affected child and a subsequent one. It was thought that these findings are indicative of reproductive compensation in families segregating for Cooley’s anaemia, and of the moderate effect of the genetic counselling given to them. This work was possible because of the co-operation of the parents of the affected children. To them, and to the medical and paramedical personnel that over a generation has helped the children at the Microcythemia Centre, the work is respectfully dedicated. The work was supported by grants from the World Health Organization and the Italian NRC to the Dopartnient of Zoology at the University of Ferrara; and by a grant of the Province of Ferrara to the Microcytheniia Centre at the Arcispedale St Anna in Ferrara. I1
HGE
42
160
SILVANAAGUZZI,C. VULLOAND I. BARRAI REFERENCES
BARRAI, I. & MANUEL,S. P. H. (1971). Equilibrium under inbreeding in balanced polymorphism. Bulletin of the World Health Organization 45, 201-8. BARRAI, I. & VULLO, C. (1976). La funzione del consultorio genetic0 nella prevenzione del morbo di Cooley. Atti V I Giornate Mediehe Daune, Foggia, 27-29 Maggio. CUCCHI,P., VULLO, G. & BARRAI, I. (1977). Population genetics in the province of Ferrara. 11. Survival of children with Cooley’s anemia. American Journal of Human. Genetics 29, 178-83. SILVESTRONI, E. & BIANCO, I. (1976). Screening for microcythemia in Italy: analysis of data collected in the past 30 years. American Journal of Human Cenetica 27, 198-212.
