Received: 30 March 2020

DOI: 10.1002/gps.5424

RESEARCH ARTICLE
- Accepted: 3 September 2020

A 10 year retrospective cohort study of inpatients with

younger‐onset dementia

Samantha M. Loi1,2 | Dhamidhu Eratne1,2 | Anita M. Y. Goh1,2,3 |

Pierre Wibawa1 | Sarah Farrand1 | Wendy Kelso1,4 | Andrew Evans5 |
Rosie Watson5,6 | Mark Walterfang1,2,7 | Dennis Velakoulis1,2

1
Neuropsychiatry, NorthWestern Mental
Health, Melbourne Health, Royal Melbourne Abstract
Hospital, Parkville, Australia
Objectives: Younger‐onset dementia (YOD) refers to a dementia where symptom
2
Department of Psychiatry, The University of
onset occurs when the patient is less than 65 years of age. YOD is far less common
Melbourne, Parkville, Australia
3
National Ageing Research Institute, Parkville,
than late‐onset dementia (occurring when patients are over 65 years old) and more
Australia challenging to diagnose due to its heterogeneous presentation. There have been
4
Department of Psychology, Monash relatively few studies describing demographic and diagnostic characteristics of
University, Clayton, Australia
5
patients with YOD in the community, particularly with follow‐up information.
Department of Medicine, Royal Melbourne
Hospital, Parkville, Australia Methods: A retrospective cohort study was performed of inpatients admitted to a
6
Population Health and Immunity Division, The tertiary neuropsychiatry service, located in metropolitan Victoria, Australia, from
Walter and Eliza Hall Institute of Medical
2009 to 2019. Inpatients with a YOD diagnosis were identified and data regarding
Research, Parkville, Australia
7 diagnosis, demographics and investigations were obtained.
Florey Institute of Neuroscience and Mental
Health, Parkville, Australia Results: There were 849 individual inpatients who were admitted to the service in
the 10‐year period and received comprehensive assessment. There were 306 in-
Correspondence
Samantha M. Loi, Melbourne Neuropsychiatry, dividuals who received a YOD diagnosis, using contemporaneous diagnostic criteria
NorthWestern Mental Health, Royal (frequency 36%). The most common diagnoses were Alzheimer's disease (24.2%),
Melbourne Hospital, John Cade Level 2,
Parkville, VIC 3050, Australia. frontotemporal dementia (23.1%), Huntington's disease (16.7%) and vascular de-
Email: [email protected] mentia (7.8%). More than half of these inpatients were followed up and 6.5% had a
diagnostic change when reviewed.
Funding information
National Health and Medical Research Conclusions: This study reports on the largest cohort of YOD to date, with diag-
Council, Grant/Award Number: GNT1138968;
nostic breakdown similar to previous retrospective file reviews. The neuropsychi-
Yulgilbar Alzheimer's Research Program
atry service is funded to follow‐up its patients, thus allowing re‐assessment and
continuity of care. While there are limitations in this study such as the lack of
neuropathological outcomes, the findings emphasise the strengths of follow‐up and
appropriate service provision for these patients.

KEYWORDS
dementia, neuropsychiatry, younger‐onset dementia

Int J Geriatr Psychiatry. 2020;1–8. wileyonlinelibrary.com/journal/gps © 2020 John Wiley & Sons Ltd.

- 1
2
- LOI ET AL.

1 | INTRODUCTION
Key points
There are approximately 27,247 people living with younger‐onset
� There was a 36% frequency of younger‐onset dementia
dementia (YOD) in Australia.1 YOD has a heterogeneous range of
presentations and aetiologies and is associated with significant delays over 10 years, with Alzheimer's disease and fronto-

in diagnosis.2,3 As symptoms occur before the age of 65, people with temporal dementia being the most common causes of

YOD are often employed with young families and providing financial dementia
� Repeat assessment of younger‐onset patients is impor-
support, and the diagnostic delay and uncertainty result in distress
and adverse emotional effects on the individual and their families. 4 tant in terms of reviewing diagnostic stability
� Despite comprehensive assessment, people were diag-
The prevalence and diagnostic make‐up of YOD in Australia re-
mains unclear with no Australia‐wide epidemiological data available. nosed with a dementia of unknown aetiology
5
The available data is restricted to local and international epidemio-
logical6,7 and retrospective file reviews. Most recently, Withall et al.5
reported a prevalence of 68.2 per 100,000 in the Eastern Sydney re- provides outpatient follow‐up through several YOD clinics (including
gion and alcohol‐related dementia was cited as the most frequent YOD specialist Huntington's disease and predictive genetic clinics).
(18.4%), followed by Alzheimer's disease (AD) (17.7%). Previous Patients receive comprehensive clinical assessments including
prevalence studies have reported a lower rate,6,7 and AD and vascular neuropsychiatry, neurology, allied health and nursing observations.
dementia (VaD) were the most commonly occurring YODs. Investigations typically include neuroimaging, blood and cerebrospi-
Retrospective file reviews contribute to the scientific evidence nal fluid (CSF) analysis. Multidisciplinary clinical reviews are held bi‐
base by evaluating the clinical disease characteristics and course over weekly and diagnoses are made using contemporaneous consensus
time.8 While not epidemiological samples, they have yielded crucial criteria. Formal feedback and discharge planning is provided to the
information on clinical assessment and frequency of the different patient and their family at the end of the assessments. A compre-
aetiologies of YOD. The diagnostic ‘mix’ of YOD has varied, with AD hensive summary is provided to referring clinicians.
and frontotemporal dementia (FTD) being the two most commonly The study was approved on 14 November 2017 by the Mel-
reported YOD diagnoses.9,10 As well as dementia aetiology, risk bourne Health hospital research ethics committee (2016.038).
factors such as family history of dementia and cerebrovascular risk
factors have been investigated.11
Internationally, there have been no further epidemiological or 2.1 | Data sources
file review studies in YOD since 2014. There is an urgent need to
revise the frequency of the different types of YOD, obtain de- 1. Comprehensive discharge summaries. An initial list of inpatient
mographic details and evaluate the need for services. In addition, the admissions who had a discharge diagnosis of dementia was ob-
studies which have included longitudinal follow‐up of patients re- tained by searching the electronic discharge summaries of all
ported on diagnostic change.9,11 This suggests that for the highly admissions admitted to the inpatient unit from January 1 2009 to
heterogeneous YOD group, close follow‐up of patients is crucial to September 7 2019 (n ¼ 1097). The Hospital International Clas-
monitor diagnostic stability as this can have significant implications sification of Diseases Adapted codes for discharge diagnoses of
on management.12 ‘dementia’ and for other neurodegenerative disorders were used.
The main aim of this paper is to describe the characteristics of These terms included ‘cognitive impairment’, ‘Alzheimer's dis-
YOD patients attending a neuropsychiatry service in metropolitan ease’, ‘frontotemporal dementia’ and ‘Huntington's disease’ (see
Victoria, Australia. Specifically, we report on the clinical experience Table S1).
of a dedicated YOD service in order to provide a contemporaneous 2. A database (maintained weekly) listing all the inpatients and their
description of the diagnostic and demographic characteristics of a discharge diagnoses. SL searched this database manually to
YOD cohort. A secondary aim was to provide information regarding ascertain whether the electronic search had missed any additional
longitudinal diagnostic changes. dementia diagnoses, obtaining the details of individual patients
who had their ‘index’ admission to the unit from the above dates.

2 | MATERIALS AND METHODS Detailed clinical information was collected.

The relevant contemporaneous consensus criteria for dementia
A retrospective file review of all inpatients who attended Neuro- was used to establish the diagnosis of dementia. This included for AD,
psychiatry, a tertiary specialist service in Melbourne, Australia, be- the NINCDS‐ADRA or NIA‐AA for AD13; consensus criteria for
tween 2009 and 2019 was undertaken. This service provides behavioural‐variant FTD14; the McKeith criteria for dementia with
diagnostic input to people with a range of neuropsychiatric pre- Lewy Bodies15 and NINDS‐AIREN criteria for VaD.16 Neuroimaging
sentations and comprises of a six‐bed diagnostic inpatient unit and details such as results of structural magnetic resonance imaging,
LOI ET AL.
- 3

positron emission tomography (PET) and single photon computed

tomography (HMPAO‐SPECT). Other investigations such as CSF 1097 total admissions
biomarkers such as α‐βeta 1‐42, total tau and phosphorylated tau
369 repeat admissions of 121 inpa!ents
levels and genetic testing.
Psychiatric diagnoses were made/confirmed/refuted by the 849 individual inpa!ents
treating psychiatrist (depending on what diagnosis had made previ-
Search for “demen!a”
ously) and according to the DSM‐IV or V criteria depending on the
year. It was noted if a ‘new’ diagnosis of a psychiatric disorder had 373 individual inpa!ents
with a demen!a diagnosis
occurred, with ‘new’ being defined as within 10 years of the eventual
dementia diagnosis.17 Cognitive testing using the Neuropsychiatry 52 with symptom onset >65
Unit Cognitive Assessment tool (NUCOG). This assesses the five
321 individual inpa!ents
major cognitive domains: attention, memory, visuospatial, executive
function and language. Scores are out of 100, with higher scores 15 with a psychiatric diagnosis
indicating better cognitive function.18 Function was assessed using
306 inpa!ents with YOD
the Global Assessment of Functioning (GAF), which rates function
19
from a scale of 1 to 100.
Neurological examination findings as undertaken by a neurolo- F I G U R E 1 Flow chart of 1097 inpatients from 1 Jan 2009–7
gist and the multidisciplinary services used during the inpatient September 2019
assessment such as allied health (neuropsychology, social work,
occupational therapy, speech therapy) were noted. In addition, de- repeat admissions of 121 individual inpatients. Of the 849 individual
mographic information and potential risk factors such as family his- inpatients, 373 had a discharge diagnosis of ‘dementia’. Fifty‐two of
tory of dementia and presence of vascular risk factors was collected. these 373 patients had symptom onset older than 65 years old and
In order to review diagnostic change, we reviewed the inpatients were excluded from further analysis, leaving 321 with a discharge
who were followed up by the service (this could have been another diagnosis of a YOD. After detailed examination of the records of
inpatient admission or through outpatients), who had their most recent these 321 patients, 15 were found to have been diagnosed with a
diagnosis reviewed to determine if there was a change in diagnosis. primary psychiatric diagnosis rather than a dementia and thus
Conversely, if the patients had previous inpatient admissions (i.e., prior excluded, resulting in 306 of the 849 (36%) individual patients
to 2009) or previous outpatient appointments (e.g., one patient had diagnosed with YOD (see Figure 1 flow chart).
their index inpatient admission in 2018 but had been seen in the
outpatient setting 7 years before this inpatient admission), their files
were also reviewed to ascertain whether they had a different diagnosis 3.1 | Demographics
rather than the one which was given during their index admission.
All identifying information was removed before usage and anal- The mean age at admission was 55.8 years old (SD ¼ 9.2, range 31–
ysis for this manuscript. 69 years). Of the 306 YOD inpatients, 51% were men and 75% came
from home prior to being admitted. Sixty percent obtained secondary
school education. Almost 60% lived in metropolitan Victoria (58%).
2.2 | Statistical analysis The majority of these inpatients were Caucasian (91.4%). The median
length of stay in the inpatient unit was 10 days (IQR 7.8–14 days,
Statistical Package for the Social Sciences (SPSS), version 24 (IBM range 3–109 days) (Table 1).
Corporation) was used for analyses. Chi‐squared analyses were used
to compare proportions of categorical variables, with t‐tests used for
differences in means. Kruskal–Wallis tests were used to compare all 3.2 | YOD diagnostic breakdown
continuous data such as age of onset. ANOVA was used to compare
continuous data for different groups. Descriptive analyses were used The most common YOD diagnosis was AD (24.2%), followed by FTD
for demographics. Normality was tested using Kolmogorov–Smirnov (23%), HD (16.7%) and VaD (10.8%). The mean age of symptom onset
test, and non‐parametric tests were used for measures which were was 51.0 years old (SD ¼ 13.4; range 20–65 years old), with no dif-
not normally distributed. ferences between men and women (p ¼ 0.412). The average duration
of diagnostic delay was 3.7 years (SD ¼ 2.6). Inpatients diagnosed
with Niemann–Pick type C (NPC) had the youngest age of onset (p <
3 | RESULTS 0.0001). The GAF score was most commonly rated as between 31
and 40 (25%) and 41–50 (31.3%), suggesting that approximately half
There were 1097 admissions to the neuropsychiatry inpatient unit the inpatients had ‘some/moderate impairment’ and ‘serious symp-
from 1 January 2009 to 7 September 2019. Of these, there were 369 toms’, respectively. Chi‐squared analysis did not reveal any
4
- LOI ET AL.

T A B L E 1 Demographics of 306 inpatients diagnosed with 3.3.4 | Risk factors for dementia
younger‐onset dementia

Number (%) Missing data (%) 26.7% of inpatients had a previous or current history of heavy alcohol
intake (n ¼ 88). Vascular risk factors (Table 2) were present in 63.4%
Men 156 (51.0) ‐
(n ¼ 194) of patients. People with VaD had the highest average
Location ‐ 3 (1.0)
number of risk factors, 2.8 (SD ¼ 1.3), which was significant (p <
Metropolitan 178 (58.2) ‐ 0.0001), F (10,205) ¼ 9.508, p < 0.0001.
Rural/regional 101 (33.0) ‐

Interstate 24 (7.8) ‐
3.3.5 | Family history
Location prior to admission ‐ 4.0 (1.3)

Home 229 (74.8) ‐ Similar proportions (65%) of inpatients had a family history of de-
Hospital 20 (6.5) ‐ mentia or psychiatric disorders. In terms of a ‘new’ psychiatric
diagnosis, 39% (n ¼ 120) had a new diagnosis of depression, 21.6% (n
Psychiatric inpatient unit 27 (8.8) ‐
¼ 66) a new diagnosis of psychosis and 3.6% (n ¼ 11) had a new
Residential facility 26 (8.5) ‐
diagnosis of mania or bipolar affective disorder.
Education ‐ 31 (10.1)

Primary 2 (0.7) ‐

Secondary 182 (59.5) ‐ 3.4 | Follow‐up and diagnostic change

Tertiary 90 (29.4) ‐
More than half (56%, n ¼ 170) of the 306 inpatients were followed up in
Other (special school) 1.0 (0.3) ‐
the neuropsychiatry outpatient clinic (end date 7 September 2019)
Almost one third of these patients had additional follow‐up by other
services such as a private specialist or community mental health team.
significant differences between the GAF scores for the different YOD There was a median of four follow‐up appointments (IQR 2‐9) over a
groups χ(60) ¼ 77.945, p ¼ 0.060 (Table 2). mean period of 366 days (range 1–3184 days). For patients who were
not followed up by the service, similar proportions were followed up by
the community mental health team (16.6%) or private specialists, such
3.3 | Clinical assessments as a private psychiatrist or private neurologist (15.8%).
There were 20 inpatients (20/306 ¼ 6.5%) who had a diagnostic
3.3.1 | Clinical reviews change during the 10‐year period (Table 3). Seventeen were re‐
diagnosed from one subtype of dementia to a different subtype of
All patients were reviewed by a neuropsychiatrist, 96.8% had dementia and three were re‐diagnosed from a psychiatric disorder
neurology review and 75.5% had a neuropsychological assessment. to dementia. Eleven of the 20 had a change in dementia diagnosis
Other allied health involvement included occupational therapy due to repeat assessments of neuropsychiatry, neuropsychology,
(82.4%), social work (66.7%) and speech therapy (32.7%). allied health and structural and functional imaging. There were
three inpatients whose dementia diagnosis was changed to a psy-
chiatric condition after the repeat assessment. With the addition of
3.3.2 | Cognitive screening genetics, four inpatients received a definitive diagnosis (C9orf72,
MAPT, HD and SPG1 spastic paraplegia). Two patients had repeat
Using the NUCOG, this was moderate, with an average total score of assessments, including a lumbar puncture for CSF Alzheimer's pro-
68.9 (SD ¼ 17.9). There was a significant difference between the total teins, had a subsequent change in diagnosis to AD. Conversely, CSF
score of inpatients with language‐variant FTD (mean score 49.2, SD negative for proteins excluded the diagnosis of AD in two inpatients.
¼ 21.2, p ¼ 0.0001), with these patients having the most severe One patient was to be included in a trial for behavioural‐variant
cognitive impairment on the NUCOG. FTD and had an amyloid PET scan as part of the trial. This scan
came back positive for amyloid and he was subsequently diagnosed
with frontal AD.
3.3.3 | Investigations

Ninety‐nine patients (32.4%) had a lumbar puncture for CSF analysis. 4 | DISCUSSION
One hundred and two (33.3%) had genetic testing of which 72 (HD ¼ 51,
NPC ¼ 11) (71.3%) yielded a positive result. The majority of inpatients This retrospective cohort study aimed to report on the diagnoses of
had structural imaging (92.7%) and functional imaging (90.1%). YOD over a 10‐year period. We found that 36% of all inpatients
LOI ET AL.
- 5

TABLE 2 Aetiology of YOD in 306 inpatients

No. males NUCOG mean total Mean no. vascular risk

No. (% of total YOD) Genetic findings no. (%) (% of total) score (SD)a factors (SD)

Alzheimer's disease 74 (24.2) ‐ 31 60.5 (16.7) 1.5 (0.8)

AD 55 (18.0) PSEN1 1 (0.6) 21 (38.2) ‐ ‐

AD PCA 17 (5.6) ‐ 8 (47) ‐ ‐

Frontal variant AD 2 (0.7) ‐ 2 (100) ‐ ‐

Frontotemporal degeneration 71 (23) 6 36 (50.7) ‐ ‐

Behavioural‐variant FTD 46 (15.0) HSAN1 1 (0.6) 22 (47.8) 65.5 (19.3) 1.6 (0.8)

FTD MND 14 (4.5) C9orf72 5 (1.6) 8 (57.1) 67.5 (15.6) 1.5 (1.0)

FTD PNFA 5 (1.6) ‐ 1 (20) 49.2 (21.2) 2.1 (1.7)

FTD semantic 6 (2.0) ‐ 5 (83.3) ‐ ‐

Huntington's disease 51 (16.7) ‐ 19 (37.2) 68 (14.1) 0.7 (0.6)

Vascular dementia 27 (10.8) Notch3 1 (0.6) 17 (62.9) 69 (15.7) 2.8 (1.3)

Αsynucleinopathyb 16 (5.2) ‐ 8 (50) 72.2 (13.9) 1.5 (0.6)

Alcohol‐related 13 (4.2) ‐ 11 (84.6) 75.4 (14.3) 1.3 (0.5)

dementia

Niemann–Pick type C 11 (3.6) ‐ 5 (45.4) 60.1 (20.2) 0.4 (0.5)

Cerebella degeneration 4 (1.3) ‐ 3 (75) ‐ ‐

CBS 3 (1.0) ‐ 2 (66.7) ‐ ‐

PSP 2 (0.7) ‐ 0 0 ‐ ‐
c
Other 17 (5.6) CSF1R, SPG1 13 (76.5) ‐ ‐

Dementia NOS 17 (5.6) ‐ 11 (64.7) 66.1 (17.4) 1.5 (1.1)

Abbreviations: AD, Alzheimer's disease; CBS, corticobasal syndrome; FTD, frontotemporal dementia; MND, motor neuron disease; NOS, not otherwise
specified; NUCOG, Neuropsychiatry Unit Cognitive Assessment tool; PCA, posterior cortical atrophy; PNFA, progressive non‐fluent aphasia; PSP,
progressive supranuclear palsy; VaD, vascular dementia; YOD, younger‐onset dementia.
a
Includes Parkinson's disease dementia (n ¼ 12) and dementia with Lewy bodies (n ¼ 4).
b
Mean NUCOG scores not calculated for cerebellar degeneration, CBS and PSP due to small numbers.
c
Includes adrenoleukodystrophy (n ¼ 2) multiple sclerosis (n ¼ 2) and HIV‐related dementias (n ¼ 2).

referred to a neuropsychiatry service received a diagnosis of a YOD. disorder but was a common aetiology in this study, reflecting this
This is less than reported by Fujihara et al.20 and Panygeres et al.11 Neuropsychiatry's specialist expertise in the assessment, diagnosis
who reported frequencies of YOD at 45.3% and 49.5%, respectively. and treatment of HD, with established relationships with Hunting-
Shinagawa et al.21 compared YOD and older‐onset dementia pre- ton's Victoria, the peak consumer organisation for HD.
senting to their memory clinic and found a lower frequency of YOD We found age of symptom onset to be 51 years old (SD ¼ 13.4),
at 27.7%. The differences may be due to the types of patients with Ferran et al.9 and Fujihara et al.20 also reporting mean age of
referred and the service models of the clinics. Fujihara et al. (2004) onset in the early 50s. A prevalence study of YOD in Eastern Sydney
reviewed consecutive files for 6 years of patients attending a found age of onset to be 55 years old (SD ¼ 9.5),5 and a Japanese
cognitive clinic, whereas the clinic where Panygeres et al. (2007) and study reported age of onset of YOD to be 53.4 years old (SD ¼ 7.9).7
9
Ferran et al. reviewed files of patients who attended clinics run by None of the previous studies reported on the age of symptom onset
neurologists. The neuropsychiatry service is a psychiatry‐led clinic of the different types/diagnoses of YOD.
and thus receives many more referrals of patients who have psy- An almost 50:50 gender split between men and women diag-
chiatric conditions with possible cognitive and behavioural change. nosed with YOD was found in this study, similar to others10,11 and
Psychiatric conditions such as schizophrenia can have cognitive the Eastern Sydney prevalence study,5 but was slightly less compared
22
change, but this may not be diagnosed as a dementia per se. to Ferran et al.9 and Fujihara et al.20 These two studies have gender
With regards to YOD aetiology, similar to previous studies (see results consistent with epidemiological reports that more men are
Table 4), we found the three most common to be AD, FTD (behav- affected by YOD, compared to more women are affected by older‐
ioural‐ and language‐variants) and VaD. HD is a relatively rare onset dementia.23
6
- LOI ET AL.

TABLE 3 Reallocation of diagnoses from initial to final diagnoses

Follow‐up diagnoses

Initial diagnoses (n) AD VaD bv‐FTD Other FTDa Genetic DLB, PDD, MSA Other dementiab Dementia NOS

AD (5) ‐ ‐ ‐ 1 ‐ 2 1 1

bv‐FTD (6) 2 1 ‐ 1 (SPG1, C9orf72) ‐ 1 1

DLB, PDD, MSA (1) ‐ ‐ 1 ‐ (MAPT) ‐ ‐ ‐

c
Other dementia (2) ‐ ‐ 1 ‐ ‐ 1 ‐ ‐

Dementia NOS (3) 1 1 ‐ ‐ ‐ ‐ 1 ‐

d
Psychiatric (3) ‐ ‐ 1 1 ‐ ‐ 1 ‐

Abbreviations: AD, Alzheimer's disease; bv‐FTD, behavioural‐variant FTD; CBS, corticobasal syndrome; DLB, dementia with Lewy bodies; FTD,
frontotemporal dementia; MSA, multiple system atrophy; NOS, not otherwise specified; PDD, Parkinson's disease dementia; PSP, progressive
supranuclear palsy; VaD, vascular dementia.
a
Includes FTD semantic (n ¼ 1) and FTD motor neuron disease (n ¼ 2).
b
Includes spastic paraplegia (n ¼ 1), vasculitis (n ¼ 1), Huntington's disease (n ¼ 1) and CBS (n ¼ 1).
c
Includes CBS (n ¼ 1) and PSP (n ¼ 1).
d
Includes psychotic depression (n ¼ 1) and depression (n ¼ 2).

The second aim was to report on diagnostic stability resulting methods in YOD in order to create a national YOD registry. This will
from follow‐up of these inpatients. The changes in diagnoses over the yield much more rich data about YOD in order to more easily, timely
10‐year period presents an interesting scenario. Panegyres et al. and accurately diagnose YOD and to monitor epidemiology in order to
(2007) reported on five changes; three who had an initial diagnosis of facilitate best practice guidelines. The utility of new investigative
FTD were shown to have a psychiatric disorder and one vice versa, processes such as amyloid PET and neurofilament light chain is likely
and one patient initially diagnosed with AD who was later re‐diag- to assist in more definitive diagnoses of YOD and accurately dis-
nosed with a psychiatric disorder. Ferran et al. (1996) described 183 tinguishing YOD from non‐neurodegenerative and psychiatric disor-
patients (out of their total of 200) who were re‐assessed after 12 ders.25 Following these patients over time may also provide
months. There were changes in dementia diagnoses, with 22 patients information regarding mortality rates in the different types of YOD.
who had AD, VaD and alcohol‐related dementia, re‐diagnosed with a The strengths of this study included the use of updated
different type of dementia and depression. There were also 42 pa- consensus criteria and a high follow‐up rate of the inpatients; how-
tients who had an ‘unknown’ diagnosis and after 12 months, two of ever, there were methodological limitations. The search process was
these continued to have the ‘unknown’ diagnosis. dependent on the dementia diagnosis which was annotated and while
Similarly, our study reiterates the importance of follow‐up as- we attempted to examine as many admissions as possible, some may
sessments for diagnostic stability. We had 6.5% of inpatients who had a have been missed. Similar to previous studies, there were small
diagnostic change, mostly from one type of dementia to another, numbers people with YOD from different ethnicities, so there con-
demonstrating that repeat assessment of neuropsychiatry, cognition tinues to be a lack of information in YOD in culturally and linguistic
and neuroimaging can be helpful in confirmation of or reformulating diverse populations. In addition, this study was based in a neuro-
diagnosis, which has implications for patients, families, and ongoing psychiatry service and referral bias may affect the proportions of
clinical management and research options. For example, patients who diagnoses. Although all patients received comprehensive multidisci-
received a re‐diagnosis of AD were started on a cholinesterase inhib- plinary assessments, multimodal investigations and diagnoses were
itor and were referred to a relevant AD clinical trial. We also found 15 based on established diagnostic criteria, other limitations include the
patients who initially had a dementia diagnosis but were re‐assessed to lack of follow‐up data for the remainder of inpatients we diagnosed
have a psychiatric disorder (subsequently excluded from analysis). This with YOD and thus the potential change of diagnosis may be under‐
emphasises the importance of appropriate diagnosis and service pro- or over‐estimated. We also lacked a definitive diagnosis from
vision in order to be able to follow‐up patients over time. neuropathological confirmation. Additionally, in a retrospective
Further adding weight to the need for longitudinal repeat as- study, not all relevant risk factors have been identified and recorded,
sessments is that our study still had 17 inpatients (5.6%) with a diag- potentially affecting any reported associations between risk factor
nosis of dementia NOS, similar to Ferran et al. (1996) who reported six and outcomes. Finally, another limitation is that many health pro-
patients (6%) who had unspecified dementia. More longitudinal fessionals have been involved in patient care, making the measure-
studies of YOD with rigorous follow‐up of neuropsychiatry, neuro- ment of risk factors and outcomes less consistent than that achieved
imaging, biomarkers and cognition will shed more light on the pro- with a prospective study design.
gression and prognosis of YOD has been recommended (Loi et al. This study reports on the aetiology of YOD and demographic de-
2020).24 Further research would include harmonisation of assessment tails, with the largest number of YOD cases in the literature. We used
 LOI
ET AL.

TABLE 4 Comparison of aetiologies and demographics to other YOD file reviews

Panegyres & Frencham Shinagawa et al. Fujihara et al. McMurtray et al. Ferran et al. Kelley et al.
This studya (2007)a (2006) (2004) (2006) (1996)a,b (2008)

Location Australia Australia Japan Brazil United States United Kingdom United States

Number 306 112 185 141 278 200 235

% YOD 36.0 49.5 27.7 45.3 29.3 ‐ 31.1

Age range 33–69 <65 <65 21–65 <65 25–71 17–45

Age onset (mean, SD) 51.0 (13.4) NR NR 51 (11.4) 51.5 (10.8) 52.6 (9.4) 34.7 (7.9)

Age at admission (mean, SD) 55.8 (9.2) NR 58.3 (11.0) 53.3 (12.1) 56.5 (9.8) 56 (9.3) 36.7 (7.8)

Aetiology, raw no. (%)

AD 74 (24.3)c 29 (25.9)c (38.5) 31 (21.3) 48 (17.3) 54 (27.0) 4 (1.7)

FTD (bv þ lang) 62 (20.2) 43 (38.4) (21.4) 7 (5.0) 7 (2.5) 7 (4.0) 31 (13.2)

VaD 24 (7.8) 7 (6.3) (12.6) 52 (36.9) 80 (28.8) 33 (17.0) 3 (1.3), 14 (5.9)d

HD 51 (16.7) Excluded NR NR 4 (1.4) NR 18 (7.7)

ARD 13 (5.6) 6 (5.4) NR 7 (5.0) 15 (5.4) 12 (6.0) 1 (0.4)

DLB 4 (1.3) 1 (0.9) (0.5) NR NR 4 (2.0) 1 (0.4)

Abbreviations: AD Alzheimer's disease; ARD, alcohol‐related dementia; DLB, dementia with Lewy bodies; FTD, frontotemporal dementia; HD, Huntington's disease; NR not reported; VaD, vascular
dementia; YOD, younger‐onset dementia.
a
Studies which followed up some patients with YOD.
b
Reviewed 200 consecutive files.
c
Includes posterior cortical atrophy and frontal AD.
d
Three patients with VaD and total of 14 had vascular disease.
-
7
8
- LOI ET AL.

comprehensive evaluation, which included neuropsychiatric examina- 11. Panegyres PK, Frencham K. Course and causes of suspected de-
tions, cognitive testing, biomarkers and neuroimaging and current mentia in young adults: a longitudinal study. Am J Alzheimers Dis
Other Demen. 2007;22(1):48‐56.
diagnostic criteria. The study also has the additional strength of follow‐
12. Lawrence E, Vegvari C, Ower A, Hadjichrysanthou C, De Wolf F,
up which few have reported on. Our findings highlight the multiple Anderson RM. A systematic review of longitudinal studies which
types of YOD and the extensive assessment required, using a multi- measure Alzheimer's disease biomarkers. J Alzheimers Dis.
disciplinary team, and the need for follow‐up. Findings also support the 2017;59(4):1359‐1379.
13. McKhann GM, Knopman DS, Chertkow H, et al. The diagnosis of
need for improvement of specialised service provision and policy for
dementia due to Alzheimer's disease: recommendations from the
people with a suspected YOD to access investigations (including im- National Institute on Aging‐Alzheimer's Association workgroups on
aging and CSF biomarkers), with the ability to repeat assessments to diagnostic guidelines for Alzheimer's disease. Alzheimers Dement.
review diagnosis and provide information about progression of disease. 2011;7(3):263‐269.
14. Rascovsky K, Hodges JR, Knopman D, et al. Sensitivity of revised
diagnostic criteria for the behavioural variant of frontotemporal
A CK N O WL E D G E ME N T S dementia. Brain. 2011;134(Pt 9):2456‐2477.
We gratefully acknowledge the neuropsychiatry clinical team and all 15. McKeith IG, Boeve BF, Dickson DW, et al. Diagnosis and manage-
the patients and families. SM Loi is funded by a National Health and ment of dementia with Lewy bodies: fourth consensus report of the
Medical Council Early Career Fellowship (GNT1138968) and the DLB Consortium. Neurology. 2017;89(1):88‐100.
16. Erkinjuntti T, Inzitari D, Pantoni L, et al. Research criteria for
Yulgilbar Alzheimer's Research Program.
subcortical vascular dementia in clinical trials. J Neural Transm Suppl.
2000;59:23‐30.
C ON F L I C T O F IN T E R ES T 17. Woolley JD, Khan BK, Murthy NK, Miller BL, Rankin KP. The diag-
There are no conflicts of interest to declare. nostic challenge of psychiatric symptoms in neurodegenerative dis-
ease: rates of and risk factors for prior psychiatric diagnosis in
patients with early neurodegenerative disease. J Clin Psychiatry.
D A T A AV AI LA B IL I T Y S T A T E M EN T 2011;72(2):126‐133.
The data that support the findings of this study are available from the 18. Walterfang M, Siu R, Velakoulis D. The NUCOG: validity and reli-
corresponding author upon reasonable request. ability of a brief cognitive screening tool in neuropsychiatric pa-
tients. Aust N Z J Psychiatry. 2006;40(11‐12):995‐1002.
19. Endicott J, Spitzer RL, Fleiss JL, Cohen J. The global assessment
ORCID scale. A procedure for measuring overall severity of psychiatric
Samantha M. Loi https://orcid.org/0000-0002-4953-4500 disturbance. Arch Gen Psychiatry. 1976;33(6):766‐771.
20. Fujihara S, Brucki SM, Rocha MS, Carvalho AA, Piccolo AC. Preva-
lence of presenile dementia in a tertiary outpatient clinic. Arq Neu-
R E F E RE N C E S ropsiquiatr. 2004;62(3A):592‐595.
21. Shinagawa S, Ikeda M, Toyota Y, et al. Frequency and clinical char-
1. Deloitte Access Economics. Dementia across Australia: 2011‐2050.
acteristics of early‐onset dementia in consecutive patients in a
Australian Capital Territory: Deloitte Access Eoncomics; 2011.
memory clinic. Dement Geriatr Cogn Disord. 2007;24(1):42‐47.
https://www.dementia.org.au/sites/default/files/20111014_Nat_Ac-
22. Trivedi JK. Cognitive deficits in psychiatric disorders: current status.
cess_DemAcrossAust.pdf.
Indian J Psychiatry. 2006;48(1):10‐20.
2. Draper B, Cations M, White F, et al. Time to diagnosis in young‐
23. Beam CR, Kaneshiro C, Jang JY, Reynolds CA, Pedersen NL, Gatz M.
onset dementia and its determinants: the INSPIRED study. Int J
Differences between women and men in incidence rates of dementia
Geriatr Psychiatry. 2016;31(11):1217‐1224.
and Alzheimer's disease. J Alzheimers Dis. 2018;64(4):1077‐1083.
3. Pasquier F, Richard F, Lebert F. Natural history of frontotemporal
24. Loi SM, Goh AMY, Walterfang M, Velakoulis D. Clinical application
dementia: comparison with Alzheimer's disease. Dement Geriatr Cogn
of findings from longitudinal studies of younger‐onset dementia:
Disord. 2004;17(4):253‐257.
rapid review and recommendations. The Journal of Neuropsychiatry
4. van Vliet D, de Vugt ME, Bakker C, et al. Caregivers' perspectives on
and Clinical Neurosciences. 2020. http://dx.doi.org/10.1176/appi.
the pre‐diagnostic period in early onset dementia: a long and
neuropsych.19110238.
winding road. Int Psychogeriatr/IPA. 2011;23(9):1393‐1404.
25. Eratne D, Loi SM, Walia N, et al. A pilot study of the utility of
5. Withall A, Draper B, Seeher K, Brodaty H. The prevalence and
cerebrospinal fluid neurofilament light chain in differentiating
causes of younger onset dementia in Eastern Sydney, Australia. Int
neurodegenerative from psychiatric disorders: a ‘C‐reactive pro-
Psychogeriatr. 2014;26(12):1955‐1965.
tein’ for psychiatrists and neurologists?. Aust N Z J Psychiatry.
6. Harvey RJ, Skelton‐Robinson M, Rossor MN. The prevalence and
2020;54(1):57‐67.
causes of dementia in people under the age of 65 years. J Neurol
Neurosurg Psychiatry. 2003;74(9):1206‐1209.
7. Ikejima C, Yasuno F, Mizukami K, Sasaki M, Tanimukai S, Asada T.
S U P P O R T I N G I NF O R M A TI O N
Prevalence and causes of early‐onset dementia in Japan: a popula-
tion‐based study. Stroke. 2009;40(8):2709‐2714. Additional supporting information may be found online in the Sup-
8. Sarkar S, Seshadri D. Conducting record review studies in clinical porting Information section at the end of this article.
practice. J Clin Diagn Res. 2014;8(9):JG01‐JG04.
9. Ferran J, Wilson K, Doran M, et al. The early onset dementias: a How to cite this article: Loi SM, Eratne D, Goh AMY, et al.
study of clinical characteristics and service use. Int J Geriatr Psychi-
A 10 year retrospective cohort study of inpatients with
atry. 1996;11:863‐869.
10. Kelley BJ, Boeve BF, Josephs KA. Young‐onset dementia: de- younger‐onset dementia. Int J Geriatr Psychiatry. 2020;1–8.
mographic and etiologic characteristics of 235 patients. Arch Neurol. https://doi.org/10.1002/gps.5424
2008;65(11):1502‐1508.