Founder effect and genetic disease in Sottunga, Finland

AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 77:335-346 (1988)

Founder Effect and Genetic Disease in Sottunga, Finland ELIZABETH O’BRIEN, L.B. JORDE, BJORN RONNLOF, JOHAN 0. FELLMAN, AND ALDUR W. ERIKSSON Department of Human Genetics, Uniuersity of Utah School of Medicine, Salt Lake City, Utah 84132 (E. O’E., L.B. J.); Samfundet Folkhiilsans Genetiska Institut, 00101 Helsinki 10, Finland (E.R., J. O.F., A. W.E.) Institute of Human Genetics, Free Uniuersity, 1007 MC Amsterdam The Netherlands (A. W E . )

KEY WORDS Genetic drift, von Willebrand disease

ABSTRACT Pedigree data are analyzed in order to determine the factors responsible for the high frequencies of certain genetic disorders in an isolated Swedish-speaking population of Finland’s Aland archipelago. The founders of Sottunga are identified, and the genetic contributions of each founder to descending birth cohorts are estimated. Founders born before 1700 have far more descendants in the contemporary gene pool than do more recent founders. However, because of migration and depopulation since 1900, the expected genetic contributions of the early founders to the present-day population are similar to those of later founders.

A descendant in the contemporary population has a 2% chance of having inherited a particular gene from the founder who makes the largest single contribution to the gene pool. This corresponds approximately to a 2% probability of inheriting an autosomal dominant disease gene from this founder. Given an average inbreeding coefficient of 0.0016, the probability of inheriting two recessive disease genes from this founder is 0.000032. The incidence of autosomal dominant von Willebrand disease in Sottunga is greater than 10% while that of autosomal recessive tapetoretinal disease is 1.5%. We conclude, therefore, that the high frequencies of these diseases are not due to the disproportionate genetic contribution of one or a few particular founders. It is more likely that these disease genes occurred in high frequency in the initial population or were introduced repeatedly through time.

Studies of disease distributions in isolated populations have often revealed high frequencies of genetic diseases that are rare in other populations. Examples include Ellis van Creveld syndrome in the Old Order Amish (McKusick, 1978), tyrosinemia among French Canadians in the Lac St. Jean region (Laberge and Dallaire, 1967), and several recessive diseases in Finnish populations (Norjo et al., 1973). When genealogical information is available for these populations, re- searchers are presented with a unique opportunity to investigate the evolutionary forces responsible for this phenomenon (Thompson, 1981). Such studies can ulti- mately lead to a better understanding of the historical, demographic, cultural, and evolutionary factors that have shaped the distribution patterns of genetic diseases in human populations.

The Aland Islands have been of interest to geneticists because certain genetic disorders, normally considered to be rare, are found there in high frequency. Autosomal dominant diseases, notably von Willebrand disease, autosomal recessive disorders including various forms of taRetoretina1 degeneration, and the X-linked Aland eye disease have been documented in high frequency among islanders (Eriksson $t al., 1980). A survey of genetic disorders in Aland completed in 1960 produced 26 cases of autosomal recessive tapetoretinal disease among the inhabitants of the outer islands (Forsius et al., 1980). These cases constituted an incidence of 1/65 affected individuals. In larger European populations von Willebrand disease affects 11

Received December 23, 1987; accepted March 9, 1988

0 1988 ALAN R. LISS, INC.

336 E. O’BRIEN ET AL.

20,000 individuals (Biggs, 1983). The frequency of this disease in Aland varies from < 1% on the main island and its immediate neighbors to > 10% in Sottunga (Lehmann et al., 1980).

In a previous 9nalysis of inbreeding in Sot- tunga, one of Aland’s small outer islands, the random and nonrandom components of inbreeding together did not account for the atypical frequencies of autosomal recessive disorders (O’Brien et al., 1988). The following analysis considers founder effect and genetic drift as alternatives to inbreeding in order to explain the elevated frequencies of both von Willebrand disease and tapetoretinal degeneration.

Controversy surrounds the role of founder effect with respect to the distribution of some widespread genetic diseases which also demonstrate ethnic and local variation (Chase and McKusick, 1972; Wagener et al., 1978; Diamond and Rotter, 1987). Disorders such as Tay-Sachs disease, cystic fibrosis, and phenylketonuria belong to this category of genes maintained in some populations at frequencies higher than can be explained by mutation alone. Without means of recon- structing a population history, and in the absence of good evidence of heterozygote advantage, the prevalence and distributions of these diseases remain enigmatic. Newer applications of restriction fragment length polymorphisms associated with disease genes show promise for resolving this controversy (DiLella et al., 1987; Henderson et al., 1987; Sankila et al., 1987).

Founder effect refers to sampling variation which occurs when a small group of individuals become the founders of a new population (Mayr, 1963). These individuals establish gene frequencies in the new population unlike those in the larger population from which they originated. Unusual frequencies of some genetic diseases often occur in population isolates founded by a small group of individuals (Roberts and Beighton, 1986). With appropriate records, pedigree analyses can demonstrate the process by which disease gene frequencies are elevated in these populations because of the genetic contributions of carrier or affected individuals among the founders (Botha and Beighton, 1983; Klinger, 1983; McKusick et al., 1964; Norio et al., 1973; Thompson, 1981). Alternatively, disease gene frequencies can be augmented in a population through a purely random process of gametic sampling without a high frequency among founders. These two re-

lated aspects of sampling variation, or drift, derive from different processes and can have differing impact on genetic evolution in a small population.

Where vital registers are available, an ef- fective means of investigating the process by which disease genes are distributed within an isolate is to track the genetic contributions in descending gene pools of each founder,(Ftoberts, 1968; Roberts and Bear, 1980; Cazes, 1986). Their contributions over many generations describe the probable introduc- tion, survival, and extinction of genes in the population. In addition, variability in the genetic contributions of founders to descending gene pools can identify individuals who might account for unusual disease gene frequencies.

Drawing upon pedigree information ex- tending back over two centuries, a detailed profile of the genetic contributions descending from each of the known founders of Sot- tunga is presented. In order to describe the composition of Sottunga’s gene pool through time, the legacy of each founder is deter- mined by 1) the number of descendants of the founder and 2) his or her probable genetic contribution. The population of Sottunga was not founded by a single group of contemporary individuals; through time off-island individuals entered the population as new founders. In addition, the island population experienced growth and decline during the period of analysis, 1700-1950.

This analysis first describes the genetic composition of Sottunga through time-a profile of the relative survival and extinction of founder genes as births, deaths, and migration caused the population to grow and shrink. This description is then brought to bear on the composition of the population at the time of the island survey which mea- sured the incidence of genetic disorders. Three alternative mechanisms generally subsumed under the term “drift” are evalu- ated as explanations for elevated disease gene frequencies: 1) atypically higher gene frequencies among the founders themselves (founder effect); 2) sampling variation due to differential genetic contributions among founders, one or a few of whom carried the disease genes; and 3) random departures from the expected transmission probability of 1/2.

The Aland Islands, located in the Baltic Sea between Finland and Sweden, are known to have supported a sizable population of over

BACKGROUND

FOUNDER EFFECT AND GENETIC DISEASE 337

10,000 until the Great Northern War of 1700- 1721. Many of the islands were evacuated during the war and later resettled by some of the prewar inhabitants and others from the Swedish mainland. The 1720s postwar census of Aland was about 6,000 individuals (Mielke e$ al., 1987). The contemporary na- tives of Aland are des$endants of the postwar population. Most Alanders trace their early origins to the Scandinavian population of the west Baltic coast or to more recent migrants from Sweden (Eriksson, 1980). In- deed, genetic distance studies demoFstrate a close affinity between Swedes and Alanders (Jorde et al., 1982).

Today the islands are subdivided into 16 Lutheran parishes; Sottunga is one of five parishes dispersed among the outer reaches of the epstern archipelago. Further details about Aland’s population origins, settle- ment, and demographic history can be found in other studies (Eriksson, 1980; Jorde et al., 1982; Mielke et al., 1976, 1987).

It has been estimated that until abouto1900, only 2.5% of all marriage partners in Aland came from outside the population, and parish endogamy was high, as much as 86% (Work- man and Jorde, 1980). The process of isolate breakdown begap by 1900. At that point migration among Aland, Finland, and Sweden began to increase, and patterns of migration among the islands changed. The more remote islands have become increasingly de- populated since the beginning of the century, while Aland’s single urban center, Marie- hamn, has grown (Eriksson, 1980). Sottun- ga’s population size has always been fairly small, between 300 and 400 (Jorde et al., 1982; Mielke et al., 1976).

The outer-island parishes accoynted for much of the genetic variability in Aland and showed the strongest potential for drift until 1900 (Workman and Jorde 1980; Jorde et al., 1982). Sottunga had the largest average genetic distance from other parishes based on three generations of gene frequency data from Alanders (Eriksson et al., 1973). Sot- tunga’s unusual frequencies of Rh and ABO alleles distinguished it from other parishes (Mielke et al., 1976; Eriksson et al., 1980; Carmelli and Jorde, 1982). Its outlier status, particularly in earlier periods, has been at- tributed to its distance from the main island region and scant immigration from other parishes.

Migration patterns and the progress of isolate breakdown also had a demonstrable effect on inbreeding in Sottunga. Average

inbreeding increased in the island population from 1700 until 1900 and then declined precipitously (O’Brien et al., 1988). Most inbreeding in Sottunga was due to remote consanguinity, i.e., parental relationships more removed than the third-cousin level. A dra- matic buildup of remote consanguinity occurred in the population as pedigrees grew more complex over many generations. However, the buildup of inbreeding and remote consanguinity was mitigated by immigration.

MATERIALS AND METHODS

According to the Swedish ecclesiastical law of 1686, each minister was required to keep birth and marriage records for the parish. Birth registration began in Sottunga in 1722 (Mielke et al., 1976). This study is based upon genealogies reconstructed by one of us (B.R.) from information contained in Sottunga’s parish registry. The genealogical data con- sist of 3,292 individuals in over 800 nuclear families. Pedigree information for these individuals spans three centuries and up to 15 generations for some individuals. Over half of those in the database (1,860 individuals) were born in Sottunga between 1690 and 1986.

A founder is defined in this analysis as any individual who appears in the database but whose parents do not. By this criterion we identify individuals who constitute the ter- minal points of every line of ascent and the set of unique genomes from which all descendant genomes derive. Most founders born before 1750 were among the postwar settlers who emigrated from Sweden. Some of these individuals lived in Sottunga or had rela- tives there before the war. More recent foun- 4ers are emigrants from Sweden, other Aland islands, or from unknown locations.

This method of identifying founders ob- viously depends upon the quality of the genealogical information. The records for individuals born on the island are believed to be very complete since the early 1700s (Mielke et al., 1976, 1987). The greatest potential source of error in identifying founders is the possibility of unknown remote relationships between founders born on nearby islands and founders born in Sottunga. Our selection of founders included all individuals meeting the founder identification criterion, regardless of whether they had actual descendants in the population. However, some of the analyses presented below pertain only to those who had descendants.


TABLE I . Founder and descendant cohort sizes by date of birth

Birth vear Founders Descendants

< 1700 59 91 1700-1750 78 198 1750-1800 89 428 1800-1850 117 632 1850-1900 96 743 1900-1950 135 480 > 1950 3 142

Founders were assigned to seven cohorts according to year of birth. Analyses are con- fined to the first six cohorts because the post- 1950 family data are incomplete for a 50-year interval. Descendants are all those who have parents in the database; they too were assigned to 50-year cohorts according to year of birth. The number of founders and descendants per cohort is reported in Table 1. Be- cause the cohort intervals are large, founders may have descendants in their own birth cohort.

The genetic contribution of each founder to each descendant in every cohort was estimated by using the method of Roberts (1968) and Roberts and Bear (1980). The probability that an individual received a gene ,from a founder via a particular path is (1/2)n1, where ni is the number of generations separating a founder from a descendant through a pedigree path, i. The total contribution of a founder to a descendant cohort is then given by summing over all k pedigree paths and all m members of the descendant cohort:

m k

The term “generations” here denotes gene transmissionevents. For example, two transmission events separate grandparents and grandchildren. Substituting this path for n, it is apparent that a grandparent contributes one-quarter of his genes on average to a grandchild. It must be emphasized that equation (1) gives an expectation of the number of contributed gene copies. Because of random gametic sampling, there is a variance about this expectation.

Estimates of the genetic contributions of founders can be expressed in two ways. First, the sum of a founder’s contribution given by (1) gives the expected number of copies of a gene (for any randomly chosen locus) that

derive from the founder and are present among the descendant cohort. Second, the contribution of a founder can be expressed as the probability that, for a randomly chosen descendant, a randomly chosen gene descends from the founder. This probability is obtained simply by dividing the quantity in equation (1) by m, the number of individuals in the descendant cohort.

For purposes of comparison, and because of our focus on particular disease genes, the convention of Roberts and Bear (1980) for calculating and interpreting founder contributions is followed in this analysis. However, careful scrutiny of (1) shows that the sum used here, as in Roberts and Bear (1980), corresponds to the expected number of copies of a particular allele at a randomly chosen locus. Since a founder must transmit one allele at each locus to his immediate offspring, the total expected number of genes at a locus contributed by a founder to a given descendant is

k

An assumption implicit in this approach is that a given gene of interest is selectively neutral. For disease genes, this would generally not be true. Although von Willebrand disease and tapetoretinal degeneration are not usually lethal, they would confer some degree of selective disadvantage. Barring higher fitness among heterozygotes, the expected contributions of founders for a disease locus should be somewhat less than the values derived from (1).

RESULTS

Figure 1 shows the number of founders from each cohort who have descendants in each birth cohort. A striking difference can be seen between the earliest founders and all subsequent founders. While the number of contributing founders from the pre-1700 cohort remains stable through time, the number of contributors from subsequent cohorts decreases substantially. Each of the post-1700 founder groups contributes in greater number than the pre-1700 founders at some point in time but subsequently falls below that of the pre-1700 group. Proportionally, 70% of the pre-1700 founder group has descendants in each succeeding cohort. Whereas more recent founder cohorts also reach this proportion of contributors at maximum, none

FOUNDER EFFECT AND GENETIC DISEASE

225

200

175

150

125

100

339

-

--

--

~-

--

--

100

80

60

F o u n d e r s 40

20

0

n t S

1700 1750 1800 1850 1300

Cohort

Fig. 1. Number of founders from six founder groups who make genetic contributions in each descendant cohort (labeled “Cohort”).

50 75 1 .+

+

+

+ 1700

- 1750

0 1800

+. 1850

* 1900

- 1950

1950

.t 1700

- 1750

8 1800

i t . 1850

* 1900

- 1950

25

0 1700 1750 1800 1850 1900 1950

Cohort

Fig. 2. Average number of descendants of each founder group in each descendant cohort.

sustains this level of representation through The average number of descendants of foun- time. ders born <1700 increases steadily until

The distribution of the average number of 1900, reaches a maximum average of 225, descendants of each founder group in SUC- and then declines. The number of descen- ceeding cohorts is shown in Figure 2 . Here dants of post-1700 founders follows a similar again founders born by 1700 appear to be pattern on a reduced scale; no post-1700 foun- outliers compared to more recent founders. der group achieves an average number of

340 E. OBRIEN ET AL.

TABLE 2. Mean number of descendants and expected copies of a particular allele contributed to the 1900-1950 gene pool by each group of founders (founder group sizes and ranges for descendants and gene

copies are shown)

Mean Range Gene Gene

Cohort Size Descendants copies Descendants copies

< 1700 59 185.48 1.06 0-423 0-7.5 1700-1750 78 42.32 .61 0-292 0-5.5 1750-1800 89 24.82 1.09 0-243 0-9.7 1800-1850 117 7.00 .80 0-81 0-9.8 1850-1900 96 3.08 1.06 0-29 0-6.7 1900-1950 135 .91 .46 0-8 0-4.0

3.00

2.50

2.00

Gene Copies 1.50

1 .oo

0.50

0.00

.+. 1700

- 1750

8 1800

Y. 1850

* 1900

- 1950

1700 1750 1800 1850 1900 1950

Cohort

Fig. 3. Average number of gene copies contributed by each founder group to each descendant cohort.

descendants greater than 50 in any time period.

The largest number of descendants (in a cohort) of any single founder is 543 individuals. The founder in question was born before 1700 and made expected genetic contributions to 73% of the 1850-1900 birth cohort.

The 1900-1950 birth cohort is roughly contemporary with the observed cases of tapetoretinal degeneration and von Willebrand disease among the outer islands. During this 50-year interval 480 individuals were born, and 230 founders make genetic contributions to this cohort. The pre-1700 founder cohort has on average 185 descendants in the 1900- 1950 birth cohort. This figure is several times larger than the average number of descen-

dants of any other founder group. The average number of descendants and

average number of gene copies contributed to the 1900-1950 birth cohort are reported for each founder group in Table 2. Ranges for the mean values are included in the table. The values in Table 2 are reported without significance tests because founders are not independent in their genetic contributions or in number of descendants.

Based on number of descendants, the earliest founders appear distinct in their level of genetic contribution to Sottunga's gene pool. This suggests the potential for a sampling effect that might account for the unusual frequencies of certain genetic disorders observed on the island. However, the advan-

Founders

FOUNDER EFFECT AND GENETIC DISEASE

80

70

60

50

40

30

20

10

0

(1700 Founders

H700 Founders 1 I

341

i 2 3 4 5 5 7 a 9 10

Gene Copies 1900-1950

Fig. 4. Distribution of founders’ contributions to the 1900-1950 descendant cohort. For comparison, founders

born before 1700 are separated from all those born after 1700.

0.030

0.025

0.020 Probable C o n t r i b u t i o n

0.015

0.010

0.005

0.000 1

+.

.t. 1700

- 1750

8 1800

+. 1850

* 1900

- 1950

I 1750 1800 1850 1900 1950

Cohor t

Fig. 5 . The probability that a member of each descendant cohort inherited a gene from a member of each founder group.

tage which the early founders appear to have cendants, the latter estimates show greater is placed in slightly different perspective by equality among founders than the absolute estimates of their genetic contributions number of their descendants suggests. De- (equation 1). Because more generations sep- spite their much larger number of descen- arate the pre-1700 founders from their des- dants, the pre-1700 founders do not make

342 E. OBRIE ,N ET AL

demonstrates little separation among founder groups. The maximum probable contribution of any founder to the 1900-1950 birth cohort is .02. This value corresponds to a 2% chance that a randomly chosen individual carries a copy of a particular gene from the founder who makes the largest genetic contribution to the population.

In order to explain some of the observed patterns of founder contributions, genetic contributions and descendant populations were estimated for male founders according to occupation. Farmers have far more descendants and make larger contributions to the gene pools of each descendant cohort than do nonfarmers. In the 1900-1950 birth cohort, farmers from all cohorts have an average of 74 descendants and 1.2 gene copies compared to 2.5 descendants and .36 gene copies for nonfarmers from all founder cohorts. This advantage is in part due to the fact that the farming occupation is the earliest one in Sot- tunga; nonfarmers have no descendants and make no genetic contributions to descendant cohorts prior to 1800.

The statistical significance of occupational influence on genetic contribution was tested by applying the Kruskal-Wallis analysis of variance to differences among occupation groups in terms of family size. In Table 3 the Kruskal-Wallis results pertain to male founders and occupation categories with ten or more cases. In Table 4 the results pertain to all married males and occupation categories with ten or more cases. In both samples farmers have the highest mean family sizes, and both tables report significant differences in family size among occupation categories.

The apparent tradition in Sottunga for farmers to have more offspring and more descendants and to make greater genetic contributions suggests an underlying continuity in landholding practices in Sottunga. Landown- ers have been described as the upper-class stratum of Aland society (Eriksson, 1980). Due to class stratification and/or land avail- ability, we might expect landholdings and, therefore, the farming occupation, to remain within families. The contingency tables shown in Table 5 demonstrate the expected pattern of farmers to be the sons (and grandsons) of farmers. A strong occupational association between fathers and sons is demonstrated, and a somewhat weaker, though still significant, association is shown between grandfathers and grandsons.

TABLE 3. Results of Kruskal-Wallis test for differences in family size among occupation groups (mean family sizes and 95% confidence interuals for means are also

rmorted)*

95% Group Group confidence

Group size mean limits

Farmer 56 4.55 3.81-5.30

Priest 10 3.20 1.44-4.96 Sailor 14 1.43 -0.06-2.91

Odd jobs 11 1.45 -0.22-3.13

*P < .001.

TABLE 4. Results of Kruskal- Wallis test for differences in family size among occupation groups, all married

males included (mean family sizes and 95% confidence intervals for means are also reported)*

95% Group Group confidence

Group size mean limits

Farmer 248 5.00 4.62-5.39 Fisherman 25 3.76 2.54-4.97 Crofter 53 4.47 3.64-5.31 Farmhand 11 2.64 0.80-4.47 Pilot 49 4.02 3.15-4.89 Carpenter 14 1.79 0.16-3.41 Sailor 50 2.04 1.18-2.90 Office 11 2.27 0.44-4.10 Technician 7 1.53 0.05-3.00

+P < 10-10.

larger expected gene contributions to the contemporary population than do other founder cohorts.

Figure 3 gives the expected number of copies of a particular gene contributed by each founder cohort to each descendant cohort. Until 1850, founders born before 1700 make a larger average contribution than other founders. By 1950 (the contemporary population) the expected values for all founders are greatly reduced and nearly converge.

The overall distribution of the expected number of gene copies contributed by all founders to the 1900-1950 birth cohort is shown in Figure 4. The great majority of founders contribute only one or two gene copies to this recent cohort. No founder contributes a disproportionately large number of gene copies to this cohort.

Figure 5 shows the distribution of expected founder contributions, or the probability that a particular gene of an individual in a descendant cohort descends from a given individual in a founder cohort. This distribution

FOUNDER EFFECT AND GENETIC DISEASE 343

TABLE 5. Test for independence o f occupations (farmers us. nonfarmersi ouer one and two generations (Yates’ correction was used in estimating the x2 ualues)

Father Grandfather Son Nonfarmer Farmer Grandson Nonfarmer Farmer

Nonfarmer 143 113 Nonfarmer 61 164 Farmer 7 56 Farmer 4 49 x2 = 38.9, P < 10 ’ iz = 8.1, P < ,005

DISCUSSION

Previous studies of the Aland Islands demonstrated that the small outer islands, Sot- tunga among them, experienced greater isolation until more recent times than the larger central islands. Drift effects on gene frequencies were evident in a study showing Sottunga’s outlier status. Total inbreeding, including the random and nonrandom components, was shown to be low for a small island population and did not account for the observed incidence of autosomal recessive disorders. However, inbreeding analysis did disclose a small set of pedigrees that are very extensive and include highly complex, albeit remote, consanguineous relationships. These pieces of evidence suggested a possible opportunity to identify one or more founders whose disproportionate contributions to the 1900- 1950 gene pool might account for the amplified preyalence of the genetic disorders af- fecting A land’s outer island population.

The maximum expected number of copies of a particular gene in the 1900-1950 gene pool (480 individuals) contributed by any one founder is 9.8. This is equal to a 2% chance that a randomly selected individual from the contemporary population inherited a particular gene (at any locus) from the founder who makes the largest genetic contribution to the gene pool a t this time. The probability that a descendant in the 1900-1950 population received two copies of a gene identical by descent from an ancestor is .0016 (O’Brien et al., 1988). These figures suggest, first of all, that a descendant from the contemporary population has a 2% chance of having an autosomal dominant disorder by descent from the ancestor who made the largest genetic contribution to the population. Second, a descendant born in 1900-1950 has a .0032% chance of having an autosomal recessive disorder if the gene was introduced (in single copy) by that founder. Since the actual fre-

quencies of von Willebrand and tapetoretinal diseases in Sottunga are greater than 10 and 1.5%, respectively (Lehmann et al., 1980; Forsius et al., 19801, it is likely that multiple founders are responsible for these genes in the contemporary population. Either several gene carriers were present among the initial founders, or the disease genes were introduced to the island repeatedly by immigrants. It is important to note that, because of its selective neutrality assumption, equation (1) produces an overestimate of the expected contribution of a disease gene carrier. This overestimate leads to a conservative un- derestimate of the number of disease genes among founders.

It is also possible, however, that the prevalence of genetic disorders in Sottunga is the result of stochastic variation due to random sampling of gametes (i.e., random deviation from the expected transmission probability of 1/2). Livingstone (1987) has simulated random drift under population conditions not unlike those in Sottunga. The results demonstrated that a single copy of a recessive disease gene introduced by one of a small group of founders will reach polymorphic frequency in an appreciable proportion of cases.

As in our previous examination of inbreeding in Sottunga, gaps in the pedigree data obscure associations between some individuals and depress contributions of some founders to their descendants. It is inevitable that immigrants are especially affected by a lack of pedigree information. Of particular con- cern are immigrants from nearby islands, a few of which share Sottunga’s prevalence of von Willebrand and tapetoretinal disease. Neighboring islands would be likely sources of immigrants who might have reintroduced disease genes into Sottunga’s population over time. In fa$, if all affected individuals throughout Aland are descended from the same founder(s), sufficient pedigree information might link informative individuals from neighboring islands to Sottunga’s genealogy.


Roberts and Bear’s study (1980) of the founders of Tristan da Cunha demonstrated changes in founder contributions resulting from two events that caused reductions in population size-once as a result of emigra- tion and once as a result of shipwreck. In general, founder contributions in Sottunga are an order of magnitude lower than those of Tristan. This appears to be due to a sub- stantial difference in the number of founders in the two populations (46 in Tristan vs. 340 in Sottunga).

The frequent integration of immigrants into Sottunga’s gene pool resulted in a large number of founders compared to its size and compared to other populations in which similar data have been examined. Among the Jicaque Indians of Honduras, six generations of pedigree information showed large differences in genetic contributions among the original seven founders of this small population (Jacquard, 1974). In addition, later immigrants made especially large contributions because of higher fertility and diminished the proportional contributions of early founders. A different scenario developed over three centuries among the Kel Kummer Tuareg of Mali (Jacquard, 1974). With 156 founders, a population size of 376 in 1970, and very deep pedigree information, this population bears greater similarity to Sottunga. Among the Kel Kummer, however, large differences in genetic contributions among founders developed, and early founders maintained larger genetic contributions through time compared to later immigrants. In particular, 15 founders accounted for 80% of the genes in the contemporary population. The Kel Kummer were considered nobility within their tribal heirarchy. Although fe- males were absorbed into the population as spouses, the offspring of immigrants were excluded from the local mate pool by strict rules governing mate choice among castes. The genetic contributions of immigrants were, therefore, temporary.

Previous studies of founder effect and random drift in small isolates suggest that demographic characteristics play an important role in establishing polymorphic frequencies of usually rare genes (Roberts, 1968; Living- stone, 1987; Neel and Thompson, 1978; Thompson and Neel, 1978). These studies have demonstrated that bottlenecks followed by rapid expansion can greatly amplify the frequencies of rare genes in a population. Even with migration rates like those found

in Sottunga, this outcome is not improbable (Livingstone, 1987).

Sottunga’s demographic history has not been examined in detail together with founder contributions in this study. It is known, however, that the pedigree information dates to a major bottleneck caused by the Great Northern War followed by 150 years of population growth and tben decline. The demographic history of Aland since 1700 (and before) is known to have been punctuated by population growth and decline due to war, famine, and epidemics interspersed with periods of recovery (Mielke, 1980; Mielke et al., 1984, 1987). These same processes surely left their mark on the small island populations prior to 1700 and may have created the appropriate conditions under which drift and founder effects established the frequencies of genetic disease found in the contemporary population.

More recent applications of restriction fragment length polymorphisms offer a new approach to examining founder effect (Henderson et al., 1987; Kere et al., 1987; Sankila et al., 1987). The identification of polymorphic alleles closely linked to a disease gene provides disease-associated haplotypes. In relatively young populations, associations between disease genes and their markers should not be disturbed by cross- overs. Therefore, the number of different haplotypes associated with a disease gene indicates the number of founders who introduc$d it to the population. Future research in Aland will include an analysis of restriction fragment length polymorphisms which have been identified within and near the von Willebrand disease locus (Bernardi et al., 1987; Ianuzzi et al., 1987; Quadt et al., 1986; Verweij et al., 1985).

CONCLUSIONS

Several factors could be responsible for the high frequencies of genetic diseases seen in this population. First, the disease genes may have existed in relatively high frequency among the initial founders of the population, due to nonrandom sampling from a larger host population. Second, one or very few founders who carried disease genes could have contributed disproportionately to the contemporary population. Third, genetic drift, in the sense of stochastic variation in gametic sampling, could have amplified the frequencies of these particular disease genes.

FOUNDER EFFECT A

The second of these possibilities is unlikely based on the analysis presented here. The third factor j s also unlikely because other parishes in Aland, which were largely separated from Sottunga, also have high frequencies of some of the same genetic diseases. It is improbable that drift would, by chance, increase the frequencies of the same diseases in different populations. It appears more likely, therefore, that multiple copies of the gene for these disorders were present in the initial population or introduced to it through time. Further evidence regarding founder effect and random drift in this population will be gained by follow-up examination of the genealogies of affected individuals and by an analysis of restriction fragment length polymorphism haplotype distributions at the von Willebrand disease locus.

LND GENETIC DISEASE 345

ACKNOWLEDGMENTS

We are grateful for comments and assis- tance from James Mielke, Kari Pitkanen, Alan Rogers, and Elizabeth Thompson. This research was supported by NSF grant BNS- 8319448 and by grants from the Sigrid Juse- lius Foundation, Helsinki, Finland.

LITERATURE CITED

Bernardi F, Marchetti G, Bertagnolo V, Faggioli L, and del Senno L (1987) Two Taql RFLPs in the human von Willebrand factor gene. Nucleic Acids Res. 151347.

Biggs, R (1983) Defects in coagulation. In AE Emery and DL Rimoin (eds.): Principles and Practice of Medical Genetics, Vol. 2. New York: Churchill Livingstone, pp. 1065-1075.

Botha MC, and Beighton P (1983) Inherited disorders in the Afrikaner population of southern Africa. S. Afr. Med. J. 64t609-612.

Carmelli D, and Jorde LB (1982) A nonparametric distance analysis of biochemical genetic data from the A land Islands, Finland. Am. J. Phys. Anthropol. 57~331-340.

Cazes MH (1986) Genetic origins of the Dogon population in the Arrondissement of Boni (Mali). Am. J. Hum. Genet. 39t96-111.

Chase GA, and McKusick VA (1972) Controversy in human genetics: Founder effect in Tay-Sachs disease. Am. J. Hum. Genet. 24r339-340.

Diamond JM, and Rotter JI (1987) Observing the founder effect in human evolution. Nature 329:105-106.

DiLella AG, Marvit J, Brayton K, and Woo SL (1987) An amino-acid substitution involved in phenylketonuria is in linkage disequilibrium with DNA haplotype 2. Nature .727:333-338.

Eriksson AW (1980) Genetic studies on Aland: Geo- graphical, historical and archival data and some other potentialities. In AW Eriksson, HR Forsius, HR Nev- anlinna, PL Workman, and RK Norio (eds.): Popula- tion Structure and Genetic Disorders. New York: Academic Press, pp. 459-470.

Eriksson AW, Eskola M-R, Workman -PL, and Morton NE (1973) Population studies on the Aland Islands. 11. Historical population structure: Inference from bioas- say of kinship and migration. Hum. Hered. 23t511- 534.

Eriksson AW, Fellman JO, and Forsius, HR (1980) Some genetic and clinical aspects of the Aland Islanders. In AW Eriksson, HR Forsius, HR Nevanlinna, PL Work- man, and RK Norio (eds.): Population Structure and Genetic Disorders. New York Academic Press, pp. 509- 536.

Forsius HR, Eriksson AW, and Damsten M (1980) Reces- sive tapetoretinal degeneration with varying diag- noses in Aland. In AW Eriksson, HR Forsius, HR Nevanlinna, PL Workman, and RK Norio (eds.): Popu- lation Structure and Genetic Disorders. New York: Academic Press, pp. 553-558.

Henderson HE, Landon SV, and Berger GM (1987) Low- density lipoprotein-receptor gene haplotypes in Afri- kaans-speaking patients with homozygous familial hy- percholesterolaemia. S. Afr. Med. J. 71:218-220.

Ianuzzi MC, Konkle BA, Ginsburg D, and Collins FS (1987) RsaI RFLP in the human von Willebrand factor gene. Nucleic. Acids Res. 155909.

Jacquard A (1974) The Genetic Structure of Populations. New York: Springer-Verlag.

Jorde LB, Workman PL, and Eriksson AW (1982) Genetic microevolution in the Aland Islands, Finland. In MH Crawford and JH Mielke (eds.): Current Developments in Anthropological Genetics, Vol. 2. New York: Plenum Press, pp. 333-365.

Kere J, Savilahti E, Norio R, and de la Chapelle A (1987) Distribution of closely linked restriction fragment length polymorphism alleles in Finnish cystic fibrosis patients and parents. 15th Annual Meeting, European Workshop for Cystic Fibrosis. Oslo, Norway.

Klinger KW (1983) Cystic fibrosis in the Ohio Amish: Gene frequency and founder effect. Hum. Genet. 65194- 98.

Laberge C, and Dallaire L (1967) Genetic aspects of tyrosinemia in the Chicoutimi region. Can. Med. Assoc. J. 97:1099-1100.

Lehmann W, Forsius HR, and Eriksson AW (1980) Von Willebrand-Jurgens syndrome on Aland. In AW Eriks- son, HR Forsius, HR Nevanlinna, PL Workman, and RK Norio (eds.): Population Structure and Genetic Dis- orders. New York: Academic Press, pp. 537-545.

Livingstone FB (1987) Simulation of the founder effect and its role in the determination of the polymorphic frequencies of deleterious genes in human populations. Hum. Biol. 59:59-75.

Mayr E (1963) Animal Species and Evolution. Cam- bridge: Harvard University Press.

McKusick VA (1978) Medical Genetic Studies of the Amish. Baltimore: Johns Hopkins University Press.

McKusick VA, Hostetler JA, Egeland JA, and Eldridge R (1964) The distribution of certain genes in the Old Order Amish. Cold Spring Harb. Symp. Quant. Biol. 29:99-113.

Mielke JH (1980) Demographic aspects of population structure in Aland. In AW Eriksson, HR Forsius, HR Nevanlinna, PL Workman, and RK Norio (eds.): Popu- lation Structure and Genetic Disorders. New York: Academic Press, pp. 471-486.

Mielke JH, Workman PL, Fellman J, and Eriksson AW (1976) Population structure of the Aland Islands, Fin- land. In H Harris and K Hirschhorn (eds.): Advances in Human Genetics, Vol. 6. New York: Plenum Press, pp 241-321.


Mielke JH, Jorde LB, Trapp PG, Anderton DL, Pitkanen K, and Eriksson AW (1984) Historical epidemiology of smallpox in Aland, Finland: 1751-1890. Demography 21:271-295.

Mielke JH, Pitkanen K, Jorde LB, Fellman JO, and Eriksson AW (1987) Demographic patterns in the Aland Islands, Finland, 1750-1900. Yrbk. Popul. Res. Finland 25.57-74.

Neel JV, and Thompson EA (1978) Founder effect and number of private polymorphisms observed in Amer- indian tribes. Proc. Natl. Acad. Sci. USA 75:1904- 1908.

Norio R, Nevanlinna HR, and Perheentupa J (1973) He- reditary diseases in Finland. Ann. Clin. Res. 5:109- 141.

O’Brien E, Jorde LB, Ronnlof B, Fellman JO, and Eriks- son AW (1988) Inbreeding and genetic disease in Sot- tunga, Finland. Am. J. Phys. Anthropol. 75:477-486.

Quadt R, Venveij CL, de Vries CJ, Briet E, and Panne- koek H (1986) A polymorphic XbaI site within the human von Willebrand factor (vWF) gene identified by a vWF cDNA clone. Nucleic Acids Res. 14:7139.

Roberts DF (1968) Genetic effects of population size re- duction. Nature 220: 1084-1088.

Roberts DF, and Bear, JC (1980) Measures of genetic change in an evolving population. Hum. Biol. 52373- 786.

Roberts DF, and Beighton P (1986) Oceanic Islands - a unique opportunity for medical research. S. Afr. Med. J. 69:293-296.

Sankila EM, de la Chapelle A, Karna J, Forsius H, Frants R, and Eriksson AW (1987) Choroideremia: Close linkage to DXYSl and DXYSl2 demonstrated by segregation analysis and historical-genealogical evidence. Clin. Genet. 31:315-322.

Thompson EA (1981) Pedigree analysis of Hodgkin’s disease in a Newfoundland genealogy. Ann. Hum. Genet. 45:279-292.

Thompson EA, and Neel JV (1978) Probability of founder effect in a tribal population. Proc. Natl. Acad. Ski. USA 75r1442-1445.

Verweij CL, Hofker M, Quadt R, Briet E, and Pannekoek H (1985) RFLP for a human von Willebrand factor (vWF) cDNA clone, pvWF1100. Nucleic Acids Res. 13:8289.

Wagener D, Cavalli-Sforza LL, and Barakat R (1978) Ethnic variation of genetic disease: Roles of drift for recessive lethal genes. Am. J. Hum. Genet. 30:262- 270.

Workman PL, and Jorde LB (1980) The genetic structure of the Aland Islands. In AW Eriksson, HR Forsius, HR Nevanlinna, PL Workman, and RK Norio (eds.): Popu- lation Structure and Genetic Disorders. New York: Academic Press, pp. 471-486.

Documents

Founder effect and genetic disease in Sottunga, Finland