Abstract
Objective: To estimate the prevalence of frontotemporal dementia (FTD) and other degenerative early-onset dementias in a geographically defined population.
Background: Early-onset dementia (at age <65 years) results in high psychiatric morbidity and caregiver burden. Prevalence figures are available for early-onset AD but not for FTD, a dementia that is almost invariably of early onset.
Methods: Case ascertainment was by review of case records of three specialist clinic databases and inpatient admissions at a university hospital in Cambridge, United Kingdom, for patients with dementia who were <65 years of age, living in Cambridge City or East or South Cambridgeshire (population 326,019) on May 30, 2000. All the relevant health services in the area were also contacted for potential cases. Diagnosis of various dementias was based on published criteria. All patients with potential FTD were examined by the study investigators and underwent structural neuroimaging. The 1998 population estimates for the area were used to calculate age and sex prevalence with confidence intervals for AD, FTD, and other causes of dementia.
Results: A total of 108 patients (66 men and 42 women) with dementia with onset before they were 65 years of age were identified, of whom 60 were <65 years on the census date, giving an overall prevalence of 81 (95% CI, 62.8 to 104.5) per 100,000 in the 45- to 64-year age group. The prevalences of early-onset FTD and AD were the same: 15 per 100,000 (8.4 to 27.0) in the 45- to 64-year-old population. The mean age at onset of FTD was 52.8 years and there was a striking male preponderance (14:3). It is possible case ascertainment methods resulted in a relative underrepresentation of some forms of dementia.
Conclusions: Frontotemporal dementia is a more common cause of early-onset dementia than previously recognized and appears to be more common in men.
Early-onset dementia (at age <65 years), though uncommon, is an important disorder that deserves special attention. Accurate epidemiologic data are required for planning of health services because the psychiatric and medical morbidity is particularly great in this group.1,2⇓ A few epidemiologic studies have specifically addressed the incidence3-5⇓⇓ and prevalence4 of early-onset AD but not other forms of dementia.
Of the non-AD dementias, frontotemporal dementia (FTD) is especially important for several reasons. Until recently, it was considered a rare disorder indistinguishable from AD and was frequently misdiagnosed even in specialist settings.6 With advances in neuroimaging techniques, it has been increasingly possible to make a diagnosis during life, leading to development of diagnostic criteria.7 FTD invariably has a presenile onset and, in neuropathologic series, has been suggested as a common cause of dementia in the younger population.8,9⇓ Behavioral disturbances (rather than cognitive impairment) are a hallmark of FTD, resulting in greater caregiver burden and increased dependency.1 Interest in FTD has increased considerably with the discovery of causative tau gene mutations in familial cases.10-12⇓⇓ To date, the only unpublished study to estimate the prevalence of FTD in a community, based on case note review, found that 12% of patients had FTD in contrast to 34% with AD in a sample of 185 patients with presenile dementia.13
We sought to estimate the prevalence of FTD and other forms of degenerative dementias in patients aged <65 years in Cambridgeshire, United Kingdom. We also examined the demographic features of un-selected groups with FTD and AD, which we predicted were likely to be the commonest degenerative causes of dementia in this age group.
Methods.
Patients and population base.
The study was approved by the local research ethics committee. We chose three areas covered by the Cambridgeshire Health Authority (Cambridge City, East Cambridgeshire, and South Cambridgeshire) as the catchment area for the study. This area includes urban and rural populations with a full range of socioeconomic circumstances, but ethnic minorities are underrepresented. These areas have well-organized health services including community resource teams, geriatric psychiatrists, and the services of a presenile dementia coordinator with minimal cross-boundary referral. Because of the close links between the community services and the specialist memory and early-onset dementia clinics, case ascertainment is likely to have been reasonably comprehensive. The total population of this area is 326,019, of whom 72,815 were aged 45 to 64 years. The target area includes a home specializing in the care of patients with advanced Huntington’s disease.
Index cases met criteria for dementia from the Diagnostic and Statistical Manual of Mental Disorders, 3rd ed.—rev.,14 with age at onset <65 years and resident in the target area, alive on the census day (May 30, 2000). Patients with dementia who moved away from the area before the project census day were excluded from the study. The onset of dementia was defined by age at appearance of the first well-defined change in cognition or behavior in the patient, as given by the caregiver. Age at diagnosis of dementia was also recorded in all patients.
The primary source of case identification was the database of the specialist services—memory, early dementia, and Huntington’s disease clinics at Addenbrooke’s Hospital. The memory clinic, established in 1990, investigates 300 patients per annum referred from local general practitioners (45%) and other consultants (30% neurologists, 20% psychiatrists, and 5% geriatricians). All new referrals are evaluated by a behavioral neurologist (J.R.H.), a clinical neuropsychologist (Dr. D. Caine), and psychiatrist (Dr. G. Berrios). All patients with potential dementia were systematically evaluated using standard protocols, which include structured interview of caregivers for symptoms suggestive of FTD,15 comprehensive neuropsychological evaluation, and neuroimaging by CT or MRI. Patients with suspected AD or FTD routinely undergo functional imaging using hexamethylphosphoramide SPECT. The procedures followed in the memory clinic are described in detail elsewhere.16 The memory clinic acts as a feeder for the early-onset dementia clinic that has been operational since 1997 and has evaluated >200 patients. The latter clinic is also directed by the senior author (J.R.H.) and does not receive direct referrals but rather coordinates the further investigation, counseling, and care of patients with dementia aged <65 years and has the services of a specialist nurse. In all patients seen in the early-onset dementia clinic, diagnosis was made by the memory clinic team. The primary cognitive evaluation instrument used in both the memory and early-onset dementia clinics is the Addenbrooke’s Cognitive Examination, which we have shown is sensitive to both AD and FTD.17 A multidisciplinary Huntington’s disease clinic was established in 1991 and accepts referrals from general practitioners and other consultants. The services of this specialist clinic include management, counseling, genetic testing, and coordination of community and social service resources for patients with Huntington’s disease.
There were multiple secondary sources of case ascertainment. Computerized inpatient hospital records for the last 3 years (1997 through May 30, 2000) at Addenbrooke’s Hospital were searched, considering all potential dementia-related International Classification of Diseases–1018 codes, including organic psychiatric disorders as well as diseases that may result in dementia for patients from Cambridgeshire in the 40- to 70-year age range. The wide coverage ensured the inclusion of the uncommon causes of dementia such as brain tumors, vasculitis, inherited metabolic disorders, and conditions that may mimic FTD, especially organic psychiatric disorders.
In addition we contacted all the 250 general practitioners, seven geriatric psychiatrists, the clinical psychology service, community resource teams, and nursing homes in the target area for patients with early-onset dementia. Information was also gathered from caregiver support groups, e.g., Pick’s disease caregiver group and the Alzheimer’s Society, for potential cases.
Diagnosis.
All the case records were reviewed by one rater (E.R.) and the diagnostic criteria (given here) were applied. Patients in whom dementia had been diagnosed before they were 65 years of age, but who were currently older and alive on the census day, were included in the study. Dementias with specific causes (e.g., B12 deficiency, hypothyroidism) as evident from clinical profile and investigations were categorized as secondary dementias. Standard diagnostic criteria were applied for the others: AD,19 FTD,7 progressive supranuclear palsy,20 diffuse Lewy body disease,21 vascular dementia,22 alcoholic dementia,14 PD,23 and multisystem atrophy.24 Corticobasal degeneration (CBD) was diagnosed in the presence of apraxia or alien hand and asymmetric parkinsonism. Patients with FTD were further subclassified as having frontal variant FTD, semantic dementia, and progressive nonfluent aphasia in keeping with local and international criteria.7,15⇓ In the small number of instances in which the diagnosis was unclear, the cases were discussed with another investigator (J.R.H.) and a consensus was reached. Patients still under evaluation or with unclear diagnosis were classified as having disease of undetermined cause. Follow-up of cases in the clinics varied from 6 months to 8 years. Information on indeterminate cases, such as patients with mild cognitive impairment, available after census day was not taken into account in the classification of cases.
Although pathologic confirmation of the diagnosis in the current cohort is still lacking, clinicopathologic study of 80 consecutively studied cases reaching postmortem in Cambridge indicates that AD and FTD cases can be differentiated in life with a high degree of accuracy. Of the 25 cases with a clinical diagnosis of AD, all have had pathologic confirmation, some with coexistent vascular or Lewy body pathology. Of the 24 cases with a clinical diagnosis of FTD coming to autopsy, 22 have shown non-AD pathology: features of classic Pick disease’s (tau-positive intraneuronal inclusions) in eight, ubiquitin-positive motor neuron disease–associated inclusions in seven, changes of CBD in three, and neuronal loss and spongiosis without distinctive histologic change in four patients. Two patients with an in vivo diagnosis of FTD had AD. The spectrum of non-AD pathology is in keeping with the experience reported by other groups.25-27⇓⇓
Statistical analysis.
Prevalence estimates were calculated by dividing the number of patients with dementia by the appropriate age and sex denominators determined from census data for the target area (1998 mid-year estimates). The 95% CI were calculated for age and sex prevalence ratios using a recommended method for small numbers.28 Student’s t-test was used to compare differences across groups. Difference in prevalence ratio between the present study and published studies was inferred from the 95% CI.28
Results.
A total of 125 patients with early-onset dementia were identified from the primary sources; of these 43 were excluded, leaving 82 eligible cases (figure). In addition, 172 potential cases were ascertained from the secondary sources, but of these only new 20 patients (all with secondary dementias) eventually fulfilled the inclusion criteria: as illustrated in the figure, a high proportion (41/172) were already known via the primary source. Six new patients were seen at the memory clinic during the study period before the census date.
Figure. Flow chart to indicate sources of identification of prevalent cases of early-onset dementia in Cambridgeshire. *Same patients (3) ascertained from different sources. CRT = community resource team.
A total of 239 general practitioners from 57 practices were contacted individually. Only 46 practitioners responded. A second letter was sent to the practice managers and an additional two practices responded. The computerized search of medical records of patients with possible dementia identified 135 potential cases. Of these, we were able to trace and examine notes of 101 and 16 were included in the study, all with nondegenerative causes of dementia. Of the 34 untraceable records, seven were coded as brain tumors, 11 PD, five metabolic disorders, two anoxic brain damage, and nine miscellaneous medical diseases with accompanying dementia. It is unlikely that any cases of FTD were missed by this route. Of note is the fact that all cases with AD and FTD were already known to the specialist clinics. All had, therefore, undergone evaluation by one of the authors (J.R.H.) together with detailed neuropsychological and radiologic investigation.
Overall, 108 patients (66 men and 42 women) had the first symptom of dementia before 65 years. If onset was considered as age at diagnosis, 88 patients (55 men and 33 women) could be included in the study (table 1). Thus 20 patients became symptomatic before 65 years but their disease was diagnosed when they were between 65 and 68 years. Of the 108 patients with early-onset dementia, primary degenerative dementias accounted for 71%, of which 35% were AD and 22% FTD. The FTD group included 13 patients with a frontal variant of FTD, two each with semantic dementia and nonfluent aphasia.7 The ages at symptom onset—52.8 (8.7) years in FTD and 57.7 (5.3) years in AD—and at diagnosis—56.1 (8.6) years in FTD and 60.7 (5.6) years in AD—were significantly lower in the FTD group as compared with AD. The duration of illness in the two groups was, however, similar: 3.3 (1.9) in FTD vs 3.0 (2.2) in AD. Men (14:3) were predominantly affected in the FTD group, whereas women (16:11) were predominant in the AD group.
Differential diagnosis of early-onset dementias in Cambridgeshire
Twenty-eight patients (25.9%) were in residential care. Fourteen of the 21 patients with Huntington’s disease were inpatients at the nursing homes specializing in the care of patients with Huntington’s disease, many of whose cases had been diagnosed elsewhere in United Kingdom and transferred to the home. Family history was seen in 26 of the patients (24.1%), 13 with Huntington’s disease and five with FTD.
Forty-eight of the 108 patients were >65 years on the census date. There was only one patient with dementia (Huntington’s disease) aged 40 to 45 years. The overall prevalence was thus estimated in the 45- to 64-year age group for both men and women in the different disease subgroups of AD, FTD, vascular dementia, Huntington’s disease, and parkinsonian syndromes using the 1998 population estimates for the target area (table 2). The prevalence of FTD was higher for men than for women and the converse was seen in AD. A comparison of age-specific prevalence for AD, FTD, and vascular dementia in the current study with published figures from other studies is shown in table 3.
Prevalence of different types of dementia in the 45- to 64-year age group for women and men per 100,000 in Cambridgeshire
Comparison of prevalence of AD and frontotemporal dementia per 100,000 in the 45- to 64-year age group between studies
Discussion.
This study, the first to address the issue of prevalence of FTD in a defined population, identified 17 patients with FTD from a population of approximately 300,000 of whom 11 were still aged <65 years. This produced an age-specific (45 to 64 years) prevalence of 15 per 100, 000 (95% CI, 8.4 to 27.0). This figure is remarkably similar to an earlier unpublished study.13 In our study, however, all the patients were examined by the study investigators and also had structural and functional neuroimaging. Our prevalence estimate for FTD must, however, be regarded as a minimum prevalence, as patients with the frontal variant of FTD typically present with changes in personality and behavior (e.g., disinhibition, apathy, reduced empathy, poor self-care) leading to referral to psychiatric services.15 The close relationship between the memory clinic, which has been in existence for a decade, and the local psychiatric services, however, lessens the likelihood that considerable numbers of cases remained undiscovered. As discussed more fully later, the methods of case ascertainment may have resulted in a relative underrepresentation of some other causes of dementia.
Almost one-third of patients with FTD (29%) had a positive family history, a figure similar to that reported (38%) in a nationwide survey of familial FTD conducted in the Netherlands.29 The Dutch study estimated that first-degree relatives of patients with FTD are at a 3.5 times higher risk of developing dementia, which may develop 11 years earlier than in the general population.29
The proportion of AD to FTD in the present study was 1.6:1, compared with the widely quoted ratio of 4:1.30 One factor affecting this ratio was the rather low prevalence of AD (15 per 100,000 for AD in the 45- to 64-year age group), which fell below that generally reported in the literature (see table 3).4,13,31-34⇓⇓⇓⇓⇓ This difference may be attributable, in part, to methods of case ascertainment. Most studies used medical case note review only, although the Finnish study used a two-stage stratified cluster sampling method,31 and that from Copiah County in the United States used a door-to-door survey for household population but also ascertained cases from institutions of AD.34 Moreover, inclusion criteria have varied, only two being based on National Institute of Neurological and Communicative Disorders guidelines.13,32⇓ All patients with FTD and AD in the current study underwent neuroimaging and neuropsychiatric and neuropsychological assessment by the same team. It is possible that this comprehensive assessment excluded some of the patients with memory problems due to depression and mild cognitive impairment. For instance, we excluded five patients with mild cognitive impairment, referred as possible AD, because serial neuropsychological assessment and neuroimaging showed, as yet, no progression. It is also highly probable that prior studies have classified some patients with FTD as having AD. Even the specificity of National Institute of Neurological and Communicative Disorders criteria in differentiating AD from FTD has been found to be very low (0.23), as most patients with FTD also fulfill these criteria for AD.35 Before the establishment of clear diagnostic criteria, the clinical diagnosis of FTD was inaccurate. For instance, only 14% of cases of pathologically confirmed FTD were diagnosed as FTD during the patient’s life in one study,6 and in another the sensitivity for diagnosis of FTD among neurologists was 43%.36 False-negative misdiagnoses mainly involved AD. A recent study found that use of specific diagnostic criteria for FTD7 improved clinical accuracy, with both specificity and sensitivity as high as 97%.37 Our own clinicopathologic experience confirms that AD and FTD can be reliably differentiated in the majority of cases in clinical settings.
Another important factor in comparing studies is the definition used of disease onset. Although we ascertained 27 cases of AD, only 15 patients were aged <65 years on the census date, resulting in a much lower prevalence estimate. Most of the AD cases were diagnosed when the patients were 60 to 65 years of age, in contrast to the majority of FTD cases, which are diagnosed when the patient is aged <65 years. The age at symptom onset was also significantly higher in the AD compared with the FTD group. This suggests that FTD, in the vast majority of cases, is a true early-onset dementia. The prevalence of AD increases with age and the proportion of patients aged <60 years of age would be low. Previously reported prevalence studies (except one)4 have defined the age at onset of dementia as the age at appearance of first symptom.13,31-34⇓⇓⇓⇓ It is, however, difficult to accurately time the onset as dementias are of insidious onset. This is even truer of patients with FTD who present with only personality changes, which may be subtle in the beginning and are given importance only in retrospect when the changes become more prominent. The prevalence as estimated by age at diagnosis is likely to be an underestimate and may affect comparisons of age-specific rates across various studies.
Although FTD has been considered to have an equal sex incidence,9,29⇓ we found it was four times more common in men than women (14:3). As the numbers were small, we cannot draw firm conclusions for this difference, which is worthy of further study. A small genetic study on 18 families of Pick disease also reported a higher risk for men.38 The difference is not easily explained, as the age at diagnosis and duration of illness between men and women were comparable. This is in contrast to the higher prevalence and incidence in women for AD that was seen in some earlier studies on presenile AD.3,31,32⇓⇓
We identified 12 cases of parkinsonian syndromes including three cases of CBD, but only five were in patients aged <65 years on the census date, giving a prevalence of 6.9 per 100,000 (95% CI, 2.9 to 16.1). The prevalence of CBD, PD, and multisystem atrophy was similar: 1.4 (0.2 to 7.8) per 100,000. Two patients had an unspecified parkinsonism-plus syndrome. There is considerable clinical and pathologic overlap between CBD and FTD,39,40⇓ prompting their classification as one entity by some workers.41
The prevalence of Huntington’s disease in the United Kingdom across various studies has varied between 3 and 10 per 100,000.42 The high prevalence of Huntington’s disease in our study was probably due to several reasons. A great number of patients (14/21) were in residential care in a special home in the target area catering exclusively to patients with Huntington’s disease, which takes referrals from throughout southern England. The presence of a specialist clinic at the hospital with availability of genetic testing may have contributed to an increase and early case detection. Some studies in the past have suggested that the original gene carriers for Huntington’s disease came from the East Anglian region of Britain, although this has been disputed.42,43⇓ It should be noted that our figures relate to prevalence and not to incidence. The long survival of patients with Huntington’s disease will elevate the prevalence.42
The limitations of the study include the relatively small size of the population surveyed and a number of potential biases in case ascertainment. Multiple sources of case ascertainment were used in order to minimize the chances of missing potential cases. Capture–recapture analysis was not performed because of lack of independence of the source of cases used. Some hospital case records were missing even after an exhaustive search, but all of the missing records appeared to relate to possible secondary dementias, which may, therefore, be underrepresented, but it is unlikely that cases with degenerative dementia seen in the hospital service were missed. The response from general practitioners (21.6%) was admittedly poor, but no additional cases of degenerative dementia were identified. Other studies on early-onset dementia have also reported on a very low ascertainment of degenerative dementia from general practices.4,13⇓ We suspect that many of the practitioners had not responded because they were unaware of cases. It is possible that AD cases were underrepresented in the sample: families of patients with AD may be less concerned than those facing the behavioral changes typical of FTD and hence less likely to seek a medical opinion. It is less likely that general practitioners would fail to refer on potential cases of AD because the multidisciplinary memory clinic in Cambridge has been in operation for more than a decade and assesses a large number of local referrals from general practitioners (see “Methods”). A community-based study with neuropsychological screening of a substantial sample of the target population would be required to estimate the true prevalence of early-onset dementia and proportion of cases due to AD and FTD. Such a study would clearly represent a huge undertaking given the low overall prevalence in this age group and the varied presentations of FTD.
The average survival of patients with early-onset AD and FTD is about 5 to 10 years.3-5,9⇓⇓⇓ Up to a quarter of patients with early-onset dementia are receiving residential care 1 year after diagnosis.2,4,13⇓⇓ At present, most of these homes are for geriatric patients and are not specialized for young patients with dementia.2 Yet patients with FTD present particular problems because of their high rate of neuropsychiatric and behavioral symptoms. These issues have important implications in service planning for early-onset dementias. We identified 17 patients with FTD in a relatively small area comprising a total population of 326,000. If these figures are extrapolated, then there are at least 3,500 patients with FTD in the United Kingdom and >10,000 in the United States.
Acknowledgments
E.R. was supported by a fellowship from the Commonwealth Commission.
Acknowledgment
The authors thank the general practitioners, practice managers, nursing homes, community resource teams, geriatric psychiatrists (Drs. C. Gregory, D. Girling, C. Lawton, T. Dening), Ms. Sue Elliott of the Alzheimer’s disease group, the Pick disease support group, Dr. Roger Barker, Dr. Anne Rossor, and Mr. Peter Driscoll (hospital information systems), Medical records, and Ms. Lynne Macdonald for help in case ascertainment; Cambridgeshire Health Authority, Cambridgeshire City Council, and Richard Wilcox from the Institute of Public Health, Cambridge, for information on geographic boundaries and census; and Judith Pride for help in contacting all the health services and secretarial assistance.
Footnotes
See also pages 1585, 1608, and 1622
- Received July 16, 2001.
- Accepted in final form February 8, 2002.
References
Letters: Rapid online correspondence
REQUIREMENTS
You must ensure that your Disclosures have been updated within the previous six months. Please go to our Submission Site to add or update your Disclosure information.
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
