The incidence of frontotemporal lobar degeneration in Rochester, Minnesota, 1990 through 1994

Methods.

We ascertained new cases of dementia from 1990 through 1994 using the records-linkage system of the Rochester Epidemiology Project, as described previously.6 We searched computerized patient indexes for 132 specific diagnostic codes that might indicate dementia. Any subject who received at least one of the study codes during the 5 years of the study (or during the subsequent 6 years) was considered as a potential case. All medical records of each potential case were screened by a specifically trained nurse abstracter, and a diagnosis was obtained, as previously described.6 The study neurologists confirmed the presence of dementia, classified the dementia by type, and determined the year of onset.

The principal sources of diagnostic information were the medical history, neurologic examinations, and neuroimaging studies as recorded historically in the patient dossier of the records-linkage system. Diagnostic criteria for dementia and for AD were those of the Diagnostic and Statistical Manual for Mental Disorders (4th ed.).7 We considered a historical diagnosis of FTLD (diagnosis of an FTLD syndrome made by the treating physician or specialist at the time of evaluation) and a study diagnosis of FTLD (retrospective application of the study criteria to the clinical information in the medical record). The study diagnosis was established by the first author using the recent consensus criteria.8,9⇓ The consensus criteria were not available to the treating physicians or specialists who made the historical diagnosis between 1990 and 1994 (or shortly after). We recognized three syndromes of FTLD8,9⇓: the behavioral-dysexecutive syndrome of frontotemporal dementia, the primary progressive aphasia syndrome, and the semantic dementia syndrome. We also used available neuroimaging data to support the diagnosis of FTLD. Finally, we identified all cases from the cohort who had died and had neuropathologic examinations.

We calculated age-specific incidence rates. The number of person-years at risk was estimated from census data with adjustment for prevalent cases of dementia, as described elsewhere.6 Average annual incidence rates are reported by 10-year age classes.

Results.

There were 60 patients who had onset of dementia under age 70 from 1990 through 1994. AD was diagnosed in 21 patients (35%). There were four cases of FTLD, all women, all of the behavioral-dysexecutive type (table). FTLD represented 7% of all the cases of dementia with onset under age 70. The four cases had been recognized and correctly diagnosed historically by their treating physicians. All four had been seen by a Mayo Clinic behavioral neurologist. Three of the four patients had MR scans that demonstrated volume loss in the temporal or frontal lobes or in both. The fourth patient had a structural imaging study that was nondiagnostic; however, she also had an SPECT study that showed reduced perfusion in frontal regions. Two of the four patients had died, and the diagnosis of FTLD was confirmed neuropathologically.

One additional man had a clinical syndrome of frontotemporal dementia with onset at age 43; however, the imaging finding of bilateral putaminal encephalomalacia was inconsistent with our definition of FTLD. No other cases with FTLD were identified historically or on medical record review among the 504 cases who were age 70 or older at onset of dementia from 1990 through 1994. Finally, among the 114 autopsied cases from the entire incidence series (13 with onset below age 70 and 101 who were age 70 or older), we found only the 2 cases of FTLD described above.

The ratio of FTLD relative to AD was approximately 1:5.25 under age 70. The incidence rate of FTLD (new cases per 100,000 person-years) was 2.2 in the 40- to 49-year age class (one case over 45,351 person-years), 3.3 in the 50- to 59-year age class (one case over 30,589 person-years), and 8.9 in the 60- to 69-year age class (two cases over 22,503 person-years) (figure). For comparison, the corresponding figures for AD were 0.0, 3.3, and 88.9 (see the figure).

• Download figure
• Open in new tab
• Download powerpoint

Figure. Age-specific incidence rates of Alzheimer disease (AD) (unpublished findings) and frontotemporal lobar degeneration (FTLD) (current study) in Rochester, MN, 1990 through 1994 (new cases per 100,000 person-years).

Discussion.

In our population, FTLD was not as common as in some autopsy studies; however, it was somewhat more common at a younger age than suggested by previous epidemiologic studies.4,5⇓ The fact that all of our cases were women is probably due to chance, given the small number of cases. All cases were of the behavioral-dysexecutive type. There were no cases of primary progressive aphasia or semantic dementia. Although the clinical diagnosis of FTLD can be imperfect, our case definition was conservative and required imaging confirmation.

The prevalence of FTLD from a UK study was 15 per 100,000 in the 45- to 64-year-old population4; the prevalence from a Netherlands study was lower (3.6 per 100,000 among the 50- to 59-year-old and 9.4 per 100,000 among the 60- to 69-year-old population).5 Our incidence rates cannot be directly compared with prevalence figures. However, the duration of survival of our four patients at the time of the study was quite long (two patients died at 5 and 12 years, and two patients were censored alive at 10 and 12 years). If we assume a conservative median survival of 6 years from onset of symptoms,10 our incidence rates would yield a prevalence higher than in both prior studies (incidence × median survival ≈ prevalence; 3.3 × 6 = 19.8 cases per 100,000 for the 50- to 59-year-old and 8.9 × 6 = 53.4 cases per 100,000 for the 60- to 69-year-old population).

Most prior knowledge about the frequency of FTLD came largely from autopsy series. In a dementia brain bank in the Twin Cities, MN,2 FTLD was rare (about 6%); however, it represented 17% of dementia cases that were under age 70 at death. In a neuropathologic review of 382 cases from the State of Florida Brain Bank,3 FTLD constituted 5% of all dementia patients; however, the frequency was 8% for cases below 70 years of age.

The strengths of the current study include the high likelihood that dementia patients under age 70 were detected. Importantly, this includes both FTLD and other types of dementia. Another strength was the availability of autopsy confirmation of diagnoses in half of the cases with FTLD. The weaknesses of this study include the small geographic area examined and the consequent small number of cases of FTLD identified. If we assume that the population of Rochester is a random sample derived from a hypothetical reference population, the CI for the incidence rate of FTLD is quite large (overall incidence rate for the population aged 40 to 69 years = 4.1; 95% CI based on the Poisson distribution = 1.1 to 10.4). Another weakness is our reliance on medical record review to define cases rather than on a prospective assessment. Because the symptoms of FTLD may have been missed or misinterpreted in some cases by the treating physicians, our incidence rates may represent an underestimation of the occurrence of FTLD. On the other hand, of the 60 patients with onset of dementia before age 70, 24 were examined by behavioral neurologists, 12 by other neurologists, and 14 had an unequivocal cause of dementia. Only for 10 cases was the historical diagnosis imperfect, leaving room for incomplete recognition of FTLD symptoms.