Genetic models have been implicated previously in the etiology of
dementia of Alzheimer type (DAT). Studies of pedigrees with multiple affected
individuals have established the existence of a familial form of the disease
with an apparent autosomal dominant pattern of inher-itance [Cook et al.,
1979]. In addition, studies in families of DAT cases have shown an increased
risk of dementia in first- and second-degree relatives of probands with DAT
[Matsuyama et al., 1985]. It is of interest that such families appear to have a
high risk not only of dementia but also of Down syndrome (DS). Heston et al.
 have studied medical disorders among the relatives of post-mortem
confirmed DAT probands. When compared to the general population, these
relatives show an in-creased frequency not only of DAT itself, but also of DS
and immunoproliferative disorders. Heston's result were confirmed in a case
control study by Heyman et al. , who found an increased frequency of
mental retardation, mainly related with DS, among relatives of clinically
diagnosed patients as compared to controls. An excess of DS in DAT relatives
has not been confirmed by others [Whalley et al., 1982; Chandra et al., 1987;
Heymann et al., 1983; Breitner and Folstein, 1984; Amad-ucci et al., 1986; Huff
et al., 1988]. As DS is not frequent (about 1.4 cases per 1,000 live births in
the general population), large family sample sizes are necessary to evidence
small differences in its frequency. Because the studies in the literature have
sample-size limitations for this purpose, it is not clear whether the two
disorders are associated in some families.
Other studies that support interest in the relationship between DAT
and DS are those that show that nearly all individuals with DS develop
Alzheimer-type changes in their brains by age 30 years [Whalley, 1982; Oliver
and Holland, 1986]. Although the neuropathological studies are in general
agreement, the proportion of individuals with DS who can be said to be demented
and have the manifestations of DAT has not been as well established [Thase et
al., 1984; Wisniewski et al., 1985].
Extensive research by molecular genetic techniques are now being
conducted in DAT [St Georges Hyslop et al., 1987a,b; Tanzi et al., 1987;
Delabar et al., 1986]. The results of studies on genes localized on chromosome
21 are conflicting. The aim of this study was to ascertain whether frequency of
severe dementia, including possible DAT, was increased in the relatives of
individuals with DS. We conducted a case-control study and compared the
familial history of children with DS and controls.
Materials and methods
Selection of Cases and Controls
The study was conducted in the outpatient department of the Hopital
Necker-Enfants Malades (Paris). Referred children were under age 18 years.
Their parents (mother and/or father) were invited to participate in the study.
Parents were interviewed at the hospital. Interviewers were not blind to the
status of the child. Parents were not told of the specific interest in DAT but
rather that the study concerned various health problems, particularly those
related to aging, within families of children with various disorders. It was
made clear that active coopera-tion of the family was needed, not only during
this interview, but after as well in order to collect missing data. A form was
used to record, at home, information that was not available at the time of the
interview. Parents of children were encouraged to obtain data from their own
parents when possible. Phone calls were made for verification and clarification
of missing or incomplete responses.
Nearly all the parents approached agreed to participate. Only
families that had lived in France for more than two generations were included.
This restriction was necessary to reduce the risk of unreliable information due
to language problems or to the removal of family. Children whose parents were
separate or whose father was unknown were excluded.
The cases were children with DS. Each of them have been karyotyped.
Nondisjunction of chromosome 21 dur-ing meiosis, leading to trisomy 21 in
children, was at the origin of nearly 95% of the DS cases. The other cases were
related to mosaicism (n = 6) or translocation of a portion of chromosome 21 on
another chromosome (n = 3). Although the study of these two types of DS would
have been interesting, the number of cases was too small to warrant a separate
study of these cases and they were excluded. Thus, to study an homogeneous
group of children, the cases included were restricted to those cases of DS
related to classical trisomy 21.
The controls were children with normal phenotype referred to the
outpatient departments of general pediatry and pneumology-allergology in the
same hospital. They were affected by ordinary diseases, mostly asthma and
We recorded demographic and medical characteristics of the children
and their parents (mother and father). A structured interview concerning the
ancestors of the child (grandparents and great-grandparents) was arranged. We
constructed a family pedigree recording all ancestors with available vital
statistics (date of birth and death). For each ancestor, living or dead, we
noted medical infor-mation in an open questionnaire. This information was
obtained from members of the families without the aid of medical records.
The interviewers asked specifically for various symp-toms of mental
deterioration. Possible dementia cases were identified by the presence of
severe memory dis-turbance, language disorders, or disorientation leading to
dependence on others. For relatives who met the criteria of possible dementia,
the age at which the symptoms began was estimated. Description of onset (sudden
or progressive) and progression of symptoms were noted. We considered only
symptoms with an evolution longer than 6 months.
The possible dementia cases were subdivided into four groups
according to mental history:
1. The possible DAT group consisted of cases with insidious onset,
gradual progression of dementia, and absence of neurological symptoms.
2. The second group consisted of possible mental de-terioration due
to stroke. It was composed of individuals with very sudden onset dementia with
or without neuro-logical symptoms.
3. The Parkinson dementia (PD) group included Parkinson disease
cases with a history or mental deteriora-tion.
4. The "undetermined" group gathered cases difficult to classify
because of unspecific or unreliable history of dementia. This group probably
included multi-infarct dementia (MID) and combined vascular and degenerative
dementia (mixed dementia). A fraction of DAT cases was probably classified in
this group. In addition, dementia due to specific condition or disorder known
to sometimes cause dementia (essentially alcoholism and depression) were
included in this group.
Statistical analysis of categorical data was performed using the
chi-square test and eventually the Fisher exact probability test. Variance
analysis was used to compare quantitative terms. Rates of dementia were
standardized by time at risk for dementia. Crude rates are presented by age
group (seven classes) along with their confidence interval.
On the hypothesis that rates of DAT were 3% in control families and
6% in DS families, 1,030 individuals in each group were required for a
sufficient study power (a = ß = 5%).
One hundred ninety-four families in each group were included in this
study. For six DS and nine controls, information was not obtained in spite of
efforts to do so by post or phone. This study reports data on the families of
188 DS children and 185 controls.
Some of the characteristics of children and parents are shown in
Table I. There were no significant differences in age or sex between DS
children and controls. As expected, birth order in DS children was higher than
in controls. Moreover, maternal and paternal age at birth were both elevated in
the DS group.
Table I - Sample characteristics
|DS cases(n == 188)||Controls
|Mean age (years)||7.4||7.9||NS
|Sex (% male)||54.8||60.5||NS
|Mean maternal age at
|Mean paternal age at
|NS = nonsignificant P value.
Available Family History
Relatives who reached the age of 60 years constituted the population
at risk of dementia. Medical information recorded on ancestors under age 60
years did not show evidence of possible dementia cases. Only seven parents of
IBS children and none of the parents of controls had reached the age of 60
years at the time of the interview. Hence, our analysis was restricted to
grandparents and great-grandparents.
Table II records the number of ancestors traced and the information
available on them. The response rate for vital statistics on ancestors was
significantly higher in DS families than in control families. In control
families, we obtained significantly more information on females than on males.
Because of the lack of reliable data on mental function, we had to exclude 6.6%
of ancestors over age 60 years in the DS group and 4.6% in the control group (P
< .05). All dementia rates have been calculated for this remaining
population of ancestors over age 60 years for whom data on mental function were
available; 1,336 subjects in the DS group and 1,113 subjects in the control
As parents of DS children were older than were the parents of
controls, the "age effect" led to differences in the vital statistics of
grandparents and great-grandparents.
1. More ancestors were deceased in the DS families. Being born
earlier than the ancestors of controls, they were more "likely to be dead."
Mean age at death of female ancestors was similar between the DS and control
group. Male DS relatives died at an age older than did male control ancestors
(P < .05).
2. For subjects alive at the time of the study, we found, similarly,
an older age in DS ancestors than in the control ancestors. This increase was
observed for both sexes.
Nevertheless, the same percentage of subjects over age 60 was
observed in both groups.
Table II. Available Family History
|Male ancestor||Female ancestor||Total ancestor
|No. ancestors with vital statistics (a)||918
(81.4%)***||lc 734(66.1%)||932 (82.6%)***||791
(71.3%)||1,850 (82.%)***||1,525 (68.7%)
|Mean age at
|Mean age at time of
|No. ancestors with age over 60 years (b)||692
(75.4%)||532 (72.5%)||739 (79.3%)||635
(80.3%)||1,431 (77.4%)||1,167 (76.5%)
|Ancestors over age 60 years with available data on mental
function (c)||641 (92.7%)||501 (94.2%)||695
(94.0%)*||612 (96.4%)||1,336 (93.4%)*||1,113 (95.4%)
|a Response rate. b Percent of ancestors with
vital statistics; c Percent of ancestors over age 60 years. P values for
comparison between DS ancestors and control ancestors: * P a0.05, ** P a
0.01, *** P a 001.
Dementia in Ancestors Over Age 60 Years
Delay between age 60 years and age at death or age at time of
interview was considered as the time at risk for dementia. This time was
significantly greater in DS than in control ancestors for both sexes (Table
After a standardization by this risk time, similar rates of possible
dementia, 5.55% in DS ancestors and 6.19% in control ancestors, were found. A
higher frequency of dementia was noted for females in both groups.
There was no aggregation of cases in families. The 78 cases in DS
were found in 63 families, 11 families with two cases and two families with
three cases. The 66 cases in controls were found in 52 families, eight families
with two cases, three families with three cases.
Study of dementia rates by classes according to clinical history did
not show any difference between the two groups. The insidious onset form,
suggestive of DAT, was present in 1.96% of DS ancestors and 2.63% of control
ancestors. An excess of cases in females was again found in the two groups.
Only one family with two possible DAT cases related at first degree was
reported in the DS group compared to three in the control group.
Distribution by age class for possible dementia of all types is
presented in Figure 1. Ascending values from 0.5% before age 65 years to 10.6%
after age 90 years were observed in the DS group. Similarly, rates varied from
0.5% before age 65 years to 16.1% after age 90 years in controls. In two age
classes, we noted significant differences: between ages 70 and 74 years, more
possible dementia cases were reported in DS families, but conversely, between
ages 75 and 79 years, the increase was in control families. Possible DAT rates
by age group are presented in Figure 2. They increased similarly with age in
the two groups.
Age of onset was sometimes difficult to determine and was not
estimated for all the subjects ( 74 of the 78 DS relative cases; 54 of the 66
control relative cases). Of the dementia cases in DS families, 13 (17.6%) began
before age 70 years compared to three (5.6%) in control families (P < .05).
But of the 13 cases with onset before age 70 years, eight were possible
dementia due to stroke, two PD, two "undetermined," and only one a possible
DAT. Of the five cases in DS relatives who began before age 60 years, two were
possible dementia suggestive of stroke, one a PD, and two "undetermined."
Table III. Dementia in Ancestors Over Age 60 Years: No.
Cases and Rates Standardized by Time at Risk of Dementia
|Male ancestors||Female ancestors||All ancestors
|DS (n = 641)||Control (n = 501)||DS (n =
695)||Control (n = 612)||DS (n = 1,336)||Control (n =
|Mean time at risk for dementia (after age 60
|All dementia||29 (4.27%)||16
(3.41%)||49 (6.70%)||50 (8.52%)||78 (5.55%)||66
|Progressive onset||6 (0.94%)||5
(1.08%)||22 (3.02%)||23 (3.91%)||28 (1.96%)||28
|Sudden onset||14 (2.15%)||2
(0.44%)||14 (1.88%)||10 (1.74%)||28 (2.0%)||12
|Parkinson dementia (PD)||4 (0.58%)||1
(0.22%)||3 (0.43%)||1 (0.14%)||7 (0.50%)||2
|Undetermined||5 (0.71%)||8 (1.67%)||10
(1.37%)||16 (2.74%)||15 (1.08%)||24 (2.24%)
|P value for comparison between DS ancestors and
control ancestors: * P a 0.05.
Fig. 1 - Dementia rate by age group (seven
Fig. 2 - Rate of
progressive dementia suggestive of dementia of Alzheimer type (DAT) by age
group (seven classes)
Dementia in Grandparents Over Age 60 Years
Second-degree relatives were analyzed separately. Demographic data
were available for mast of the grandparents: 98% in the DS group and 95% in the
control group. We obtained sufficient data on mental function of 505
grandparents (95.5% of the 529 who reached age 60 years) in the DS group and
453 (94.2% of the 481 who reached age 60 years) in the control group. Mean tine
at risk for dementia was greater, for both sexes, in grandparents of DS
children than in those of the control (Table IV). After standardization by time
at risk, rate of dementia in DS grandparents was significantly higher than in
controls: 3.63 vs. 0.90% (P < .01). This increase was significant for
grandfathers (4.93 vs. 0.45%, P < .01) but not for grandmothers (2.78 vs.
No possible DAT cases were described in DS grandfathers or in the
control grandfathers. Nine of the 13 cases (69%) were related to mental
deterioration due to stroke. They were observed in the maternal branch of the
families for ten of the 13 cases.
Table IV. Dementia in Grandparents Over Age 60 Years: No.
Cases and Rates Standardize by Time at Risk for Dementia
|DS (n = 258)||Control (n = 220)||DS (n =
247)||Control (n = 233)||DS (n = 505)||Control(n =
|Mean time at risk for dementia (after age 60
|All dementia||13 (4.93%)***||1
(0.45%)||9 (2.78%)||3 (1.38%)||22 (3.63%)**||4
|Progressive onset dementia suggestive of
|P values for comparison between DS ancestors
and control ancestors: P a 0.05, ** P a 0.01, *** P a .001.
Dementia in Ancestors of Children With Maternal Age Under
Elevated maternal age at birth is considered to be the most
important risk factor for DS, and it is hypothesized that the mechanism at the
origin of trisomy 21 (nondisjunction of chromosome 21 originated from the
mother in 80% of the cases) might differ according to maternal age at birth
[Penrose, 1983; Erickson, 1978].
We compared families of DS children with maternal age under 35 years
to families of controls with same maternal age. There were 118 DS children and
167 controls with mother's age under 35 years at birth. Available data on
mental function were provided for 93.5% (n = 857) of the DS ancestors with age
over 60 years and 95% (n = 982) of the control group's ancestors (Table V).
Dementia rates standardized by time at risk were similar in the two
groups: 4.98% in the DS and 6.36% in the control group. No excess of possible
DAT cases was observed.
Of the cases with known age of onset (42 in DS, 50 in controls), an
earlier onset was noted for Down syndrome, seven cases (four possible dementia
due to stroke, two PD, one "undetermined") before age 70 years compared to two
(possible dementia due to stroke) in control ancestors (P = .05).
Table V. Dementia in Relatives of Proband With Maternal Age
at Birth < 35 Years: No. Cases and Rates Standardized by Time at Risk for
|Male ancestors||Female ancestors||All ancestors
|DS (n = 415)||Control (n = 445)||DS (n =
442)||Control (n = 537)||DS (n = 857)||Control (n =
|Mean time at risk for dementia (after
|All dementia||15 (3.58%)||14
(3.28%)||29 (6.25%)||46 (9.02%)||44 (4.98%)||60
|Progressive onset dementia suggestive of DAT||6
(1.40%)||4 (0.92%)||12 (2.54%)||23 (4.49%)||18
|P value for comparison between DS ancestors and
control ancestors: P a.05.
The present study found no excess of severe dementia in the families
of DS cases compared with the families of controls. Among the various types of
dementia, we did not observe an increased frequency of occurrence of possible
DAT in ancestors of probands with DS.
In discussing these negative results, some aspects of the study should
be considered. First, ascertainment of dementia cases is affected by 1) the
motivation of the respondents to collect vital statistics and medical
information on their families; and 2) the personal bias of the respondent (or
the interviewer) regarding the reporting of dementia cases. The respondents of
FMS cases provided more vital statistics on their families than did the
controls. However, comparable information about medical history and mental
functioning in ancestors over age 60 years was obtained for both DS cases and
controls. Strict guidelines were defined to minimize the effect of the
interviewer regarding the diagnosis of dementia, and mild and moderate
dementing conditions were excluded. Although some of these might be true
dementia cases, there is no reason to suspect that comparability is affected by
this underestimating. However, it seems easier to obtain information of an
equal quality for families of both DS cases and the controls if ascertainment
is restricted to severe demented illness. Thus, the proportion of
false-positive and false-negative errors in the diagnosis of dementia cases in
families is likely to be identical in both groups. Because interviewers were
aware of the hypothesis of the study, we cannot totally exclude a recall bias.
This could have led to an overdetection of dementia cases in DS families, but
the negativity of our results is against such a bias.
The frequency of occurrence of dementia observed in this study is
about 6%. This frequency increased more than tenfold between ages 60 and 90
years in relatives of both DS and controls. These endings are comparable to
prevalence rates for severe dementia that have been reported in other studies
[Mölsä et al., 1982; Kay et al., 1964]. Despite inherent differences in study
design, which make a direct comparison of our results with those of other
studies difficult, this similarity provides confirmatory evidence for the
reliability of our data. The present study, similar to previous studies
[Sulkava et al., 1985; Broe et al., 1976], found higher rates of severe
dementia in females.
The assignment of type of dementia, based on the reporting of history
of onset and progression of the dementing disorder, was made with great
caution. On the whole, less than 40% of demented cases were diagnosed as
probable DAT, a proportion that is very close to those found in other
epidemiological studies [Mölsä et al., 1982; Sulkava et al., 1985; Rorsman et
An important concern in interpreting these negative results is whether
the study power was sufficient to detect a difference in the frequencies of
occurrence. In this study, the number of subjects in each group was large and
sufficient to detect weak association between DS and dementia (study power =
In this study, we observed very few early-onset dementia cases, that
is, before age 60 years. The slight excess of cases with onset before age 70
years observed in DS relatives was related mostly to the cases with onset
suggestive of possible dementia due to stroke and not to possible early-onset
DAT cases. Yatham et al. [ 1988] suggested recently that the rate of presenile
DAT cases was higher in relatives of DS probands than in the general population
and discussed this result with respect to the very small number of presenile
dementia cases (n = 4) that they observed. Our study did not find a similar
One possible explanation of differences between our results and those
observed in studies conducted by Heston et al.  and Heyman et al. 
is that DAT could be associated with some special conditions of DS. For
example, in families of DAT, translocation or duplication of part of chromosome
21 could be looked for. We think that this hypothesis is not very likely, but
it cannot be totally excluded. When the investigation is focused on classical
free trisomy 21, there is no evidence of excess of DAT in ancestors of Down
Amaducci LA, Fratiglioni L, Rocca WA, Fieschi C, Livrea P, Pedone D,
Bracco L, Lippi L, Gandolfo C, Bino G, Prencipe M, Bonnati ML, Girotti F,
Carella F, Tavolato B, Ferla S, Lenzi GL, Carolei A, Gambi A, Grigoletto F,
Shoenberg BS (1986) : Risk factors of clinically diagnosed Alzheimer disease: A
case-control study of an Italian population. Neurology 36:922-931.
Breitner JCS, Folstein MF (1984): Familial Alzheimer dementia; A
prevalent disorder with specific clinical features. Psychol Med 14:64-80.
Broe GA, Akhtar AJ, Andrews GR, Caird FI, Gilmore AJJ, Mc Lennan WJ
(1976): Neurological disorders in the elderly at home. J Neurol Neurosurg
Chandra V, Philipose V, Bell PA, Lazaroff A, Shcenberg BS (1987):
Case-control study of late onset probable Alzheimer's disease. Neurology
Cook R, Ward B, Austin J (1979): Studies in aging of the brain: IV.
Familial Alzheimer disease; relation to transmissible dementia, aneuploidy and
microtubular defects. Neurology 29:1402-1412.
Delabar JM, Lamour Y, Gegonne A, Davous P, Roudier M, Nicole A,
Ceballon I, Amouyel P, Stehelin D, Sinet PM (1980): Rearrangement of chromosome
21 in Alzheimer's disease? Ann Genet (Paris) 29:226-228.
Erickson JD (1978): Down's syndrome, paternal age, maternal age and
birth order. Ann Hum Genet 41:289-298.
Heston LL, Mastri AR, Anderson E, White J (1981): Dementia of the
Alzheimer type. Clinical genetics, natural history, and associated conditions.
Arch Gen Psychiatry 38:1085-1090.
Heyman A, Wilkinson WE, Hurwitz BJ, Schmechep D, Sigmon AH, Weinberg
T, Helms MJ, Swift ML (1983): Alzheimer's disease: Genetic aspects and
associated clinical disorders. Ann Neurol 14:507-515.
Heyman A, Wilkinson WE, Stafford JA, Helms M, Sigmon AH, Weinberg T
(1984): Alzheimer's disease, a study of epidemiological aspects. Ann Neurol
Huff FJ, Auerbach J, Chakravarti A, Boller F (1988): Risk of dementia
in relatives of patients with Alzheimer's disease. Neurology 38:786-790.
Kay DWK, Beamish P, Roth M (:1964): Old age mental disorders in
Newcastle-upon-Tyne. I. A study of prevalence: Br J Psychiatry 110:146-158.
Matsuyama SS, Jarvik LF, Kumai V (1985): 5 Dementia genetics. In Avie
T (ed): "Recent Advances in Psychogeriatrics." London: Churchill Livingstone,
Mölsä PK, Marttila RJ, Rinne UK (1982): Epidemiology of dementia in
Finnish population. Acta Neurol Scand 65:541-552.
Oliver C, Holland AJ (1986): Down's syndrome and Alzheimer's disease:
A review. Psychol Med 16:307-322.
Penrose LS (1983): The relative effects of paternal and maternal age
in mongolism. J Genet 27:219-224.
Rorsman B, Hagnell 0, Lanke J (1986): Prevalence and incidence of
senile and multi infarct dementia in the Lundby study: A comparison between the
time periods 1947-1957 and 1957-19 7 2. Neuropsycho-biology 15:122-129.
St Georges Hyslop PH, Tanzi RE, Polinski RJ, Haines JL, Nee L, Watkins
PC, Myers RH, Feldman RG, Pollen I), Drachman D, Growdon J, Bruni A, Foncin JF,
Salmon D, Frommelt P, Amaducci L, Sorbi S, Piacentini S, Stewart GD, Hobbs WJ,
Conneally M, Gusella JF (1987a): The genetic defect causing Alzheimer's disease
maps on chromosome 21. Science 235:885-890.
St Georges Hyslop PH, Tanzi RE, Polinski RJ, Neva RL, Pollen D,
Drachman D, Growdon J, Cupples LA, Nee L, Myers; RH, O'Sullivan D, Watkins PC,
Amos JA, Deutsch CK, Bodfish JW, Kinsbourne M, Feldman RG, Bruni A, Amaducci L,
Foncin JF, Gusella JF. (1987b): Absence of duplication of chromosome 21 genes
in familial and sporadic Alzheimer's disease. Science 238:664-666.
Sulkava R, Wikström J, Aromaa A, Raitasalo R, Lehtinen V, Lahtela K,
Palo J (1985): Prevalence of severe dementia in Finland. Neurology
Tanzi RE, Gusella JC, Watkins PC, Bruns GAP, St Georges Hyslop PH, Van
Keuren ML, Patterson D, Pagan S, Kurnitt DM, Neve RL. (1987): Amyloid ß
protein gene: cDNA, mRNA distribution and genetic linkage near the Alzheimer
locus. Science 235:880-884.
Thase ME, Tigner R, Smeltzer DJ, Liss L (1984): Age related
neuropsychological deficits in Down syndrome. Biol Psych 18:571-585.
Whalley LJ (1982): The dementia of Down's syndrome and its relevance
to aetiological studies of Alzheimer's disease. Ann NY Acad Sci 396:39-53.
Whalley LJ, Carothers AD, Collyer S, De Mey R, Frackiewicz A (1982): A
study of familial factors in Alzheimer's disease. Br J Psychiatry
Wisniewski KE, Wisniewski HM, Wen GY (1985): Occurrence of
neuropathological changes and dementia of Alzheimer's disease in Down syndrome.
Ann Neurol 17:278-282.
Yatham LN, McHale, Kinsella A (1988): Down syndrome and its
association with Alzheimer's disease. Acta Psychiatr Scand 77:38-41.