Original quantitative research – Prevalence and sequence of chronic conditions in older people with dementia: a multi-province, population-based cohort study

Abstract

Introduction: Comorbid chronic conditions contribute to increased health service use and poor outcomes for people with dementia, but there is little information about the prevalence of these conditions in this population.

Methods: We used linked administrative data from British Columbia (BC), Ontario (ON), Quebec (QC) and Prince Edward Island (PE) to identify a cohort of 287 453 individuals aged 65 years and older with prevalent dementia in April 2015, and followed this population until March 2020. We determined the prevalence of comorbid chronic conditions and ascertainment dates using Canadian Chronic Disease Surveillance System definitions, and used descriptive statistics to compare patterns across provinces.

Results: Sociodemographic characteristics were similar across provinces (mean age: 83.0 [PE]–84.3 [BC] years; female sex: 61.8% [BC]–66.2% [QC]; and long-term care facility residence: 39.5% [QC]–41.6% [BC]). People with dementia commonly experienced five or more comorbid conditions (38.8% [PE]–53.5% [ON]); the most prevalent were hypertension (76.4% [PE]–81.4% [ON]), mental illness and alcohol- or drug-induced disorders (44.4% [QC]–91.2% [BC]) and osteoarthritis (43.8% [PE]–60.4% [ON]). Hypertension, diabetes and stroke were frequently apparent before dementia ascertainment, whereas heart failure and traumatic brain injury were apparent almost as frequently after dementia ascertainment as before.

Conclusion: Patterns of comorbid chronic conditions were similar across provinces, with most present prior to dementia ascertainment. Health service planning strategies should be developed and shared across provinces to address the complex health care needs of people with dementia.

Keywords: dementia, chronic disease, comorbidity, epidemiology, hypertension, mental illness, osteoarthritis, pan-Canadian

Introduction

Alzheimer disease and related dementias (referred to as dementia in this article) are progressive neurodegenerative diseases that affected 6.4% of Canadians aged 65 years and older in fiscal year 2020 to 2021.^{Footnote 1} The prevalence of dementia in Canada is rising, partly due to longer lifespans, and the number of individuals with dementia is expected to double by 2050, to 1.3 million in Canada and 153 million globally.^{Footnote 2} The large burden of cognitive and behavioural symptoms frequently associated with dementia is also stressful for people with the disease and their care partners.^{Footnote 3} The direct health care costs for people living with dementia in Canada are projected to increase to CAD 18.2 billion by 2031.^{Footnote 4}

Because the number of chronic conditions increases with age and because specific conditions (such as diabetes and hypertension) are modifiable risk factors for dementia, people with dementia often have comorbid conditions.^{Footnote 5} Comorbidity can exacerbate functional decline^{Footnote 5} and accounts for significant additional health service use, fragmented care and poorer outcomes.^{Footnote 6} Management of comorbid chronic conditions requires an integrated, interdisciplinary approach to improve the quality of life of people with dementia.^{Footnote 7}

Public health in Canada focusses on promoting health, preventing disease and injuries, responding to public health threats and providing information to support decision-making. Canada released its first national dementia strategy in 2019.^{Footnote 8} One of the objectives of the strategy is prevention, which is supported by surveillance and data that provide a more accurate picture of the impact of dementia in Canada and ensure that preventive efforts are appropriately targeted.^{Footnote 8} Surveillance includes tracking modifiable risk and protective factors and bridging evidence gaps.^{Footnote 8}

Describing the prevalence of comorbid chronic conditions, and their sequence of occurrence in relation to dementia case ascertainment, would help decision makers understand the complexity of the care required in order to plan services, to ensure the availability of adequate health care personnel and to educate health care providers on the management of comorbidity.^{Footnote 9Footnote 10} However, there is limited information on the prevalence of chronic conditions among people with dementia across Canada. Our objective is to describe the prevalence of comorbid chronic conditions among older people with dementia, in four Canadian provinces, and to determine the temporal sequence of conditions relative to dementia ascertainment.

Methods

Study design and data sources

We conducted a series of population-based, retrospective cohort studies of adults aged 65 years and older living in the provinces of British Columbia (BC), Ontario (ON), Quebec (QC) and Prince Edward Island (PE). These four provinces, where more than 75% of the Canadian population reside,^{Footnote 11} participated in an initiative led by the Public Health Agency of Canada to report on the prevalence of comorbid chronic conditions among people with dementia. We used the health administrative databases available in each province to identify and characterize each cohort and to examine chronic conditions. These databases included provincial health insurance registries, the Canadian Institute for Health Information’s Discharge Abstract Database, physician billing claims and prescription drug dispensation databases (see Table 1). Database availability varied across provinces and some analyses used unique jurisdiction-specific datasets, such as the Ontario Mental Health Reporting System.

Table 1. Summary of health administrative databases available by province, Canada

Database	Characteristics used in this study	Availability by province
		BC	ON	QC	PE
Provincial health insurance registry	Age Sex Postal code (to report community size, neighbourhood income quintile and other area-based measures, where available via linkage to census data using Statistics Canada’s Postal Code Conversion File)	Yes	Yes	Yes	Yes
Canadian Institute for Health Information (CIHI) Discharge Abstract Database (DAD)	Acute care hospitalization within the 1 year prior to the index date Chronic conditions	Yes	Yes	Yes	Yes
Ontario Mental Health Reporting System (OMHRS)	Inpatient health service use for mental illnesses	NA	Yes	NA	NA
Physician billing claims	Outpatient visits to family physicians, dementia specialists and other specialists within the 1 year prior to the index date Chronic conditions	Yes	Yes	Yes	Yes
Prescription drug dispensations	Chronic conditions Note: Generally available for people aged ≥ 65 years and other select populations, e.g. long-term care residents, people receiving social assistance, people who do not have private drug coverage (QC)	Yes	Yes	Yes	Yes
Continuing Care Reporting System – Long-term care	Long-term care residents at the index date	Yes	Yes	NAFootnote a	NA
Canadian Index of Multiple Deprivation (CIMD)	Situational vulnerability and economic dependency quintiles	Yes	NA	NA	NA
Ontario Marginalization Index (ON-MARG)	Situational vulnerability and economic dependency quintiles	NA	Yes	NA	NA
Institut national de santé publique du Québec (INSPQ) Material and Social Deprivation Index	Situational vulnerability and economic dependency quintiles	NA	NA	Yes	Yes

Study population

We identified all adults aged 65 years and older with dementia on 1 April 2015 (the index date) using the Canadian Chronic Disease Surveillance System (CCDSS) dementia case ascertainment algorithm,^{Footnote 12} an adaptation of an algorithm previously validated in ON.^{Footnote 13} The algorithm defines dementia as one or more hospital separation records (generated each time a patient is discharged or transferred from a health care facility, signs out against medical advice or dies); three or more physician claims for dementia within 2 years (at least 30 days between each); or one or more cholinesterase inhibitor drug prescriptions. The algorithm has a sensitivity of 79.3% and a specificity of 99.1% among older adults when validated against family physician records.^{Footnote 13} An individual’s dementia ascertainment date was defined as the date they first met one of the criteria the algorithm uses.

We chose a study time period that excluded the COVID-19 pandemic in the follow-up period because of the changes that were occurring in health system use and health care service availability.

Chronic conditions

We identified 15 comorbid chronic conditions according to CCDSS algorithms based on hospital discharge abstracts, physician billing claims and dispensed medication claims (see Table 2 for the list of conditions and definitions).^{Footnote 12} We selected conditions identified by the Lancet Commission report on dementia prevention, intervention and care^{Footnote 14} as modifiable risk factors for dementia and by a 2017 study that used CCDSS data to examine multimorbidity,^{Footnote 15} and based on the conditions’ associations with age.^{Footnote 16} We also included traumatic brain injury based on a definition obtained from the literature,^{Footnote 17Footnote 18Footnote 19} but as emergency department visit data were not available for all the participating provinces, case ascertainment was based solely on hospital records.

Ascertainment for all conditions involved a look-back period, defined by data availability in each province and CCDSS case definitions,^{Footnote 12} of approximately two decades except for mental illness and alcohol- or drug-induced disorders for which the look-back period was 10 years. Individuals were followed until 31 March 2020 for the development of chronic conditions and were censored on their death date, if applicable, to allow for a minimum follow-up of 5 years after the study index date (1 April 2015). The unequal length of the look-back and follow-up periods reflect the natural progression of dementia, which tends to be ascertained later in life. The case ascertainment date for each condition was defined as the date individuals first met algorithm criteria.

Table 2. Case ascertainment algorithms for chronic conditions based on CCDSS definitions

Chronic condition	Case definition	Case ascertainment date	ICD-9 codes	ICD-10 codes	Physician billing codes
Dementia	≥ 1 hospital separation records; or ≥ 3 physician claims within 2 years, with at least 30 days between qualifying claims; or ≥ 1 cholinesterase inhibitor drug prescription	Hospital separation record, last physician claim, or drug prescription date (whichever comes first)	046.1, 290.0, 290.1, 290.2, 290.3, 290.4, 294.1, 294.2, 331.0, 331.1, 331.5 (or 331.82 in ICD-9-CM)	G30, F00, F01, F02, F03	290, 331
Heart failure	≥ 1 hospital separation records or ≥ 2 physician claims within 1 year	Hospital separation record or last physician claim (whichever comes first)	428	I50	428
Hypertension (excluding gestational hypertension)	≥ 1 hospital separation records or ≥ 2 physician claims within 2 years	Hospital separation record or last physician claim (whichever comes first)	401, 402, 403, 404, 405	I10, I11, I12, I13, I15	401, 402, 403, 404, 405
Ischemic heart disease	≥ 1 hospital separation records or procedure codes, or ≥ 2 physician claims within 1 year	Hospital separation record or last physician claim (whichever comes first)	410, 411, 412, 413, 414Footnote a	I20, I21, I22, I23, I24, I25Footnote a	410, 411, 412, 413, 414
Diabetes (excluding gestational diabetes)	≥ 1 hospital separation records or ≥ 2 physician claims within 2 years	Hospital separation record or last physician claim (whichever came first)	250	E10, E11, E12, E13, E14	250
Stroke	≥ 1 hospital separation records or ≥ 2 physician claims within 1 year	Hospital separation record or last physician claim (whichever comes first)	325, 362.3x, 430, 431, 432.9, 433.x1, 434 (or 434.x1), 435.x, 436, 437.6Footnote b	G08, G45.x (excluding G45.4), H34.0, H34.1, I60.x, I61.x, I62.9, I63.x, I64, I67.6	325, 430, 431, 432.9, 434, 435, 436, 437.6
Traumatic brain injuryFootnote c	≥ 1 hospital separation records	Date of hospital discharge or emergency department registration date (whichever came first)	310.2, 800.1, 800.3, 801.1, 801.3, 802.6, 802.7, 803.1, 803.3, 804.1, 804.3, 850, 851, 852, 853, 854, 907.0, 907.1, 925	F07.2, S02.0, S02.1, S02.3, S02.7, S02.8, S02.9, S06.0-S06.9, S07.1, T90.2, T90.5	Not applicable
Parkinsonism	≥ 2 physician claims within 1 year, with at least 30 days between qualifying claims	Last physician claim	Not applicable	Not applicable	332
Epilepsy	≥ 1 hospital separation records, or ≥ 3 physician claims within 2 years, with at least 30 days between qualifying claims	Hospital separation record or physician claim (whichever comes first)	345.0, 345.1, 345.4, 345.5, 345.6, 345.7, 345.8, 345.9	G40	345
Mental illness and alcohol- or drug-induced disorders (10-year look-back period)	≥ 1 hospital separation records or ≥ 1 physician claims within 1 year	Hospital separation record or physician claim (whichever comes first)	290.8, 290.9, 291–293, 294.0, 294.8, 294.9, 295–319	F04–F99	291–319 (50B in BC only)
Schizophrenia	≥ 1 hospital separation records, or ≥ 2 physician claims within 2 years, with at least 30 days between qualifying claims	Hospital separation record or physician claim (whichever comes first)	295	F20, F21, F23, F25	295
Osteoarthritis	≥ 1 hospital separation records, or ≥ 2 physician claims (separated by at least 1 day) within 5 years	Hospital admission or physician claim (whichever comes first)	715	M15–M19	715
Rheumatoid arthritis	≥ 1 hospital separation records, or ≥ 2 physician claims (> 8 weeks apart) within 2 years, with exclusion criterionFootnote d	730 days after hospital separation record or last physician claim (whichever comes first)	714	M05–M06	714
Osteoporosis	≥ 1 hospital separation records or ≥ 1 physician claims	Hospital separation record or last physician claim (whichever comes first)	733	M80, M81	733
Osteoporosis-related fractures (hip, forearm, pelvic, humerus and spine)	Hip: ≥ 1 hospital separation records (6-month episode) Forearm, pelvic and humerus: ≥ 1 hospital separation records or ≥ 2 physician claims within 3 months (6-month episode) Spine: ≥ 1 hospital separation records or ≥ 1 physician claims (6-month episode)Footnote e	Hip: Hospital admission Forearm, pelvic and humerus: Hospital admission or last physician claim (whichever comes first) Spine: Hospital admission or physician claim (whichever comes first)	Hip: 820 Forearm, pelvic and humerus: 813, 808, 805.6, 805.7, 812 Spine: 805.2–805.5	Hip: S72.0, S72.1, S72.2 Forearm, pelvic and humerus: S52, S32.1, S32.3, S32.4, S32.5, S42.2, S42.3, S42.4 Spine: S22.0, S22.1, S32.0	Hip: Not applicable Forearm, pelvic and humerus: 813, 814Footnote f, 808, 812 Spine: 805
Asthma	≥ 1 hospital separation records or ≥ 2 physician claims within 2 years	Hospital separation record or last physician claim (whichever comes first)	493	J45, J46	493
COPD	≥ 1 hospital separation records or ≥ 1 physician claims	Hospital separation record or physician claim (whichever comes first)	491, 492, 496	J41, J42, J43, J44	491, 492, 496

Other characteristics

We used each province’s health insurance registry to obtain sociodemographic characteristics for individuals on the study index date, including their age, sex and postal code. The latter was linked to Statistics Canada’s Postal Code Conversion File to identify neighbourhood income quintiles and community size. The geocode, in turn, was linked to structural determinants of health indices that measure situational vulnerability and economic dependency. Neighbourhood income quintiles were defined based on the average household income of the neighbourhood (dissemination area) relative to the income level of the larger census regions.

We also identified health service utilization during the year preceding cohort index date. These included acute care hospitalizations, emergency department visits, family physician visits, dementia specialist visits (i.e. neurology, geriatrics and psychiatry) and other specialist visits.

Statistical analyses

We compared sociodemographic characteristics, prior health system use and prevalent comorbid chronic conditions across the provinces by reporting the mean (and standard deviation) for continuous variables (e.g. age) and counts and percentages for categorical variables (e.g. sex). To describe the temporal sequence of chronic conditions relative to dementia onset, we compared the number and percentage of chronic conditions ascertained before versus after dementia case ascertainment.

Ethics approval

In BC, this work was part of a comprehensive population health surveillance and research initiative of the Office of the Provincial Health Officer and was reviewed by the University of British Columbia Research Ethics Board (#H22-01818).

In ON, these datasets were linked using unique encoded identifiers and analyzed at ICES. ICES is an independent, nonprofit research institute whose legal status under Ontario’s health information privacy law allows it to collect and analyze health care and demographic data, without consent, for health system evaluation and improvement. The use of the data in this project is authorized under section 45 of Ontario’s Personal Health Information Protection Act and does not require review by a research ethics board.

In QC, this work is part of the ongoing chronic disease surveillance mandate assigned to the Institut national de santé publique du Québec (INSPQ) by the Minister of Health and Social Services. All surveillance activities under this mandate are approved by the provincial Public Health Ethics Committee. No informed consent was required.

In PE, this work is part of ongoing and systematic chronic disease surveillance conducted by the Chief Public Health Office, under the authority of the Minister of Health and Wellness through sections 3 and 58 of the PE Public Health Act, and does not require review by a research ethics board.

Results

We identified 287453 people aged 65 years and older with dementia living in the four participating provinces on 1 April 2015, from 1390 in PE to 152816 in ON. (For a flow diagram showing the process of identification, see Figure 1.)

Figure 1 : Descriptive text

Flow diagram showing identification of people (≥ 65 years) with dementia in participating provinces, Canada

• Step 1 of 5:
  • People with dementia on the index date^{Footnote a} identified using the CCDSS dementia case ascertainment algorithm:
    • BC = 53 623
    • ON = 181 678
    • QC = 315 636
    • PE = 5459
• Step 2 of 5: people removed due to being out of scope
  • Age < 65 or > 110 years:
    • BC^{Footnote b} = Not shown
    • ON = 14 417
    • QC^{Footnote c} = 4171
    • PE^{Footnote b} = Not shown
  • Death date prior to or on the study index date:
    • BC^{Footnote b} = Not shown
    • ON = 8424
    • QC = 214 909
    • PE^{Footnote b} = Not shown
• Step 3 of 5:
  • BC = 44 637
  • ON = 158 837
  • QC = 96 556
  • PE = 1560
• Step 4 of 5: people removed due to ineligibility
  • Not eligible for provincial health insurance in the 6 months prior to or on the index date:
    • BC = 0
    • ON = 421
    • QC^{Footnote d} = 7587
    • PE = 170
  • No contact with the health system for 5 years prior to the index date:
    • BC = 359
    • ON = 5600
    • QC = 0
    • PE = 0
• Step 5 of 5:
  • Final cohort of adults ≥ 65 years with dementia:
    • BC = 44 278
    • ON = 152 816
    • QC = 88 969
    • PE = 1390

Abbreviations: BC, British Columbia; CCDSS, Canadian Chronic Disease Surveillance System; ON, Ontario; PE, Prince Edward Island; QC, Quebec.
^a The study index date was 1 April 2015.
^b Exact values not shown due to small cell size suppression.
^c Exclusion for age < 65 years in Quebec only.
^d Exclusion for health insurance eligibility at index date only in Quebec.

Baseline demographics

The mean age of people with dementia was similar across the four provinces (from 83.0 years in PE to 84.3 years in BC) (Table 3). There were more females than males with the diagnosis in each province (from 61.8% in BC to 66.2% in QC). A large percentage of individuals were living in long-term care homes (from 39.5% in QC to 41.6% in BC) and in lower neighbourhood income quintiles 1 and 2 (from 47.7% in ON to 55.9% in PE). Fewer people with dementia were living in rural areas than in larger communities (from 10.5% in BC to 35.3% in PE). About one-fifth lived in areas with high situational vulnerability (from 15.1% in PE to 20.6% in ON).

Table 3. Baseline demographic and health service use characteristics of adults (≥ 65 years) with prevalent dementia on 1 April 2015, by province, Canada

Characteristics	BC N = 44 278		ON N = 152 816		QC N = 88 969		PE N = 1390
	n	(SD) or %	n	(SD) or %	n	(SD) or %	n	(SD) or %
Time since dementia case ascertainment, mean years (SD)	3.6	(3.2)	4.3	(3.7)	3.8	(3.2)	2.8	(2.8)
Mean age at index date,Footnote a years (SD)	84.3	(7.5)	83.3	(7.7)	83.4	(7.4)	83.0	(7.5)
Method of dementia case ascertainment, n (%)
Hospitalization criterion	9087	20.5	6280	4.1	23 333	26.2	363	26.1
Physician claims criterion	17 543	39.6	44 964	29.4	18 452	20.7	381	27.4
Drug prescription criterion	17 648	39.9	101 572	66.5	47 184	53.0	646	46.3
Categorized age at index date, years
65–69	1764	4.0	8309	5.4	3840	4.3	59	4.2
70–74	3736	8.4	13 949	9.1	7884	8.9	141	10.1
75–79	6558	14.8	23 176	15.2	13 542	15.2	241	17.3
80–84	10 300	23.3	35 208	23.0	21 761	24.5	324	23.3
85–89	11 416	25.8	38 312	25.1	23 311	26.2	355	25.5
≥ 90	10 504	23.7	33 862	22.2	18 631	20.9	270	19.4
Sex, n (%)
Male	16 903	38.2	55 829	36.5	30 039	33.8	502	36.1
Female	27 375	61.8	96 987	63.5	58 930	66.2	888	63.9
Living in long-term careFootnote a	18 441	41.6	61 320	40.1	35 175	39.5	NA	NA
Neighbourhood income quintile
Missing	284	0.6	816	0.5	NA	NA	6	0.4
Q1 (lowest)	12 846	29.0	39 295	25.7	NA	NA	360	25.9
Q2	8954	20.2	33 624	22.0	NA	NA	417	30.0
Q3	9199	20.8	28 705	18.8	NA	NA	272	19.6
Q4	6757	15.3	25 696	16.8	NA	NA	169	12.2
Q5 (highest)	6238	14.1	24 680	16.2	NA	NA	166	11.9
Community size, n
Missing	281	0.6	779	0.5	252	0.3	6	0.4
≥ 1 500 000	21 358	48.2	59 490	38.9	41 942	47.1	0	NA
500 000–1 499 999	0	NA	25 753	16.9	17 927Footnote b	20.1	0	NA
100 000–499 999	11 431	25.8	35 222	23.0	17 927Footnote b	20.1	0	NA
10 000–99 999	6542	14.8	15 331	10.0	11 765	13.2	893	64.2
< 10 000 (rural)	4666	10.5	16 241	10.6	17 083	19.2	491	35.3
Situational vulnerabilityFootnote c
Missing	88	0.2	1742	1.1	25 738	28.9	265	19.1
Q1 (least deprived)	9146	20.7	30 252	19.8	9399	10.6	153	11.0
Q2	9207	20.8	30 215	19.8	11 341	12.7	236	16.9
Q3	8448	19.1	29 154	19.1	11 869	13.3	188	13.5
Q4	9511	21.5	29 929	19.6	14 336	16.1	338	24.3
Q5 (most deprived)	7878	17.8	31 524	20.6	16 286	18.3	210	15.1
Economic dependencyFootnote c
Missing	88	0.2	1742	1.1	25 738	28.9	265	19.1
Q1 (least deprived)	5595	12.6	15 989	10.5	10 726	12.1	219	15.8
Q2	7008	15.8	20 005	13.1	10 582	11.9	117	8.4
Q3	7020	15.9	22 610	14.8	13 090	14.7	131	9.4
Q4	9322	21.1	27 395	17.9	14 052	15.8	343	24.7
Q5 (most deprived)	15 245	34.4	65 075	42.6	14 781	16.6	315	22.7
Previous health service use (at least once in the previous year)
Acute care hospitalization	9294	21.0	37 322	24.4	23 140	26.0	415	29.9
Family physician visit	43 046	97.2	142 922	93.5	78 112	87.8	1153	82.9
Dementia specialist visit (including neurology, geriatrics and psychiatry)	10 580	23.9	44 296	29.0	23 307	26.2	465	33.5
Specialist visit, other	27 839	62.9	83 749	54.8	61 158	68.7	656	47.2

Case ascertainment, health system use and comorbidity

Time since dementia case ascertainment was similar across the four provinces, from 2.8 years in PE to 4.3 years in ON (Table 3). Use of cholinesterase inhibitor drug prescriptions was the most common data source for dementia case ascertainment across the provinces, although the percentage of cases identified by this data source varied from 39.9% in BC to 66.5% in ON.

The proportion of individuals with at least one acute care hospitalization in the year prior to dementia ascertainment was similar across the provinces, from 21.0% in BC to 29.9% in PE (Table 3). Family physician visits in the year prior to the index date were least prevalent in PE (82.9%) and most prevalent in BC (97.2%). In contrast, visits to dementia specialists were least prevalent in BC (23.9%) and most prevalent in PE (33.5%).

Hypertension, mental illness and alcohol- or drug-induced disorders, and osteoarthritis were the most prevalent comorbid conditions (Figure 2; Table 4). The prevalence of mental illness and alcohol- or drug-induced disorders varied the most across the provinces, from 44.4% in QC to 91.2% in BC. The prevalence of ischemic heart disease, stroke, osteoporosis-related fractures, osteoarthritis, rheumatoid arthritis and osteoporosis also varied, but the relative ranking of prevalence was similar. Between 38.8% (PE) and 53.3% (ON) people with dementia had five or more comorbid chronic conditions.

Figure 2 : Descriptive text

Prevalent chronic conditions among adults (≥ 65 years) with prevalent dementia on 1 April 2015, by province, Canada

Chronic condition	British Columbia	Ontario	Quebec	Prince Edward Island
Heart failure	23.6%	19.2%	16.1%	17.3%
Hypertension	78.5%	81.4%	78.2%	76.4%
Ischemic heart disease	37.9%	44.3%	42.3%	33.7%
Diabetes	31.5%	33.7%	30.5%	28.9%
Stroke	25.5%	28.5%	20.5%	28.9%
Traumatic brain injuryFootnote a	4%	3.3%	3.4%	2.4%
Parkinsonism	5.9%	5.5%	4.7%	4.2%
Epilepsy	2.9%	3.3%	3.5%	2.5%
Mental illness and alcohol/drug-induced disordersFootnote b	91.2%	77.9%	44.4%	87.6%
Schizophrenia	4.2%	3.8%	3.2%	3%
Osteoarthritis	47.8%	60.4%	44.5%	43.8%
Rheumatoid arthritis	5.1%	0.5%	3.4%	2%
Osteoporosis	30%	36.6%	40.8%	20.2%
Osteoporosis-related fractures	27%	17.9%	22.9%	18.9%
Asthma	10.7%	14.3%	11.1%	9.7%
COPD	28.1%	29.3%	27.3%	27.3%

Abbreviations: CCDSS, Canadian Chronic Disease Surveillance System; COPD, chronic obstructive pulmonary disease.
^a Traumatic brain injury data reported using hospitalization data alone (not a CCDSS case definition).
^b A 10-year look-back period from 1 April 2015 (the index date) was used to report use of health care services for mental illness and alcohol- or drug-induced disorders.

Table 4. Prevalent chronic conditions among older adults (≥ 65 years) with prevalent dementia on 1 April 2015, by province, Canada

Characteristics, n (%)	BC N = 44 278		ON N = 152 816		QC N = 88 969		PE N = 1390
	n	%	n	%	n	%	n	%
Number of chronic conditions
≥ 5	21 900	49.5	81 492	53.3	42 440	47.7	539	38.8
Specific chronic conditions
Heart failure	10 440	23.6	29 323	19.2	14 282	16.1	241	17.3
Hypertension	34 757	78.5	124 376	81.4	69 589	78.2	1062	76.4
Ischemic heart disease	16 801	37.9	67 660	44.3	37 620	42.3	468	33.7
Diabetes	13 965	31.5	51 560	33.7	27 114	30.5	402	28.9
Stroke	11 286	25.5	43 599	28.5	18 209	20.5	402	28.9
Traumatic brain injury	1766	4	5094	3.3	3016	3.4	33	2.4
Parkinsonism	2609	5.9	8428	5.5	4140	4.7	59	4.2
Epilepsy	1305	2.9	5018	3.3	3151	3.5	35	2.5
Mental illness and alcohol- or drug-induced disorders	40 382	91.2	119 083	77.9	39 532	44.4	1218	87.6
Schizophrenia	1842	4.2	5795	3.8	2824	3.2	42	3
Osteoarthritis	21 170	47.8	92 368	60.4	39 613	44.5	609	43.8
Rheumatoid arthritis	2258	5.1	753	0.5	3055	3.4	28	2
Osteoporosis	13 293	30	55 967	36.6	36 275	40.8	281	20.2
Osteoporosis-related fractures (hip, forearm, pelvic, humerus and spine)	11 957	27	27 342	17.9	20 369	22.9	262	18.9
Asthma	4728	10.7	21 842	14.3	9876	11.1	135	9.7
COPD	12 447	28.1	44 815	29.3	24 321	27.3	380	27.3

A larger percentage of chronic conditions were apparent prior to rather than after dementia case ascertainment (Figure 3). Vascular-related conditions such as hypertension, ischemic heart disease, diabetes and stroke as well as mental illness and alcohol- or drug-induced disorders, osteoarthritis, rheumatoid arthritis, osteoporosis and asthma were often apparent prior to dementia ascertainment. Heart failure, traumatic brain injury, parkinsonism, osteoporosis-related fractures and epilepsy became apparent almost as often after dementia ascertainment as before. These patterns were generally consistent across provinces, with some variation in prevalence estimates.

Figure 3 : Descriptive text

Sequence of chronic conditions apparent before and after dementia case ascertainment of adults (≥ 65 years) with prevalent dementia on 1 April 2015, by province, Canada

Chronic condition	British Columbia - Before dementia case ascertainment	British Columbia - After dementia case ascertainment	Ontario - Before dementia case ascertainment	Ontario - After dementia case ascertainment	Quebec - Before dementia case ascertainment	Quebec - After dementia case ascertainment	Prince Edward Island - Before dementia case ascertainment	Prince Edward Island - After dementia case ascertainment
Heart failure	51.9%	48.1%	47.3%	52.7%	70.6%	29.4%	58.5%	41.5%
Hypertension	91.0%	9.0%	89.5%	10.5%	90.7%	9.3%	89.7%	10.3%
Ischemic heart disease	82.0%	18.0%	79.8%	20.2%	86.8%	13.2%	84.4%	15.6%
Diabetes	82.4%	17.6%	78.5%	21.5%	86.3%	13.7%	85.8%	14.2%
Stroke	66.6%	33.4%	63.0%	37.0%	79.2%	20.8%	68.2%	31.8%
Traumatic brain injury	52.4%	47.6%	44.3%	55.7%	62.5%	37.5%	48.5%	51.5%
Parkinsonism	62.3%	37.7%	56.4%	43.6%	69.3%	30.7%	78.0%	22.0%
Epilepsy	56.9%	43.1%	55.4%	44.7%	66.7%	33.3%	60.0%	40.0%
Mental illness and alcohol/drug induced disordersFootnote a	87.7%	12.3%	72.6%	27.5%	83.5%	16.5%	78.9%	21.1%
Schizophrenia	66.4%	33.6%	46.0%	54.0%	79.0%	21.0%	71.4%	28.6%
Osteoarthritis	83.3%	16.7%	88.0%	12.0%	87.0%	13.0%	85.6%	14.4%
Rheumatoid arthritisFootnote b	82.8%	17.2%	82.8%	17.3%	79.3%	20.7%	84.0%	16.0%
Osteoporosis	77.9%	22.1%	81.0%	19.0%	83.1%	16.9%	77.6%	22.4%
Osteoporosis-related fractures	55.5%	44.5%	58.5%	41.5%	67.6%	32.4%	58.0%	42.0%
Asthma	85.3%	14.7%	85.6%	14.4%	86.6%	13.4%	88.2%	11.8%
COPD	69.2%	30.8%	72.4%	27.6%	81.1%	18.9%	70.3%	29.7%

Abbreviations: COPD, chronic obstructive pulmonary disease; PE, Prince Edward Island.
^a Mental illness and alcohol/drug-induced disorders.
^b A range of values is provided for rheumatoid arthritis in PE because of small cell size suppression for values of 1–5.

Discussion

We compared the prevalence of comorbid chronic conditions among older adults with dementia across four Canadian provinces and examined their sequence of occurrence relative to dementia case ascertainment.

The COVID-19 pandemic revealed many issues in collecting, sharing, presenting and interpreting health data nationally.^{Footnote 20} Federal and provincial governments are collaborating to develop a pan-Canadian strategy on health data management^{Footnote 21} to which cross-provincial research contributes by consolidating the evidence base and refining methods to inform surveillance initiatives. Our methods may be used for other chronic conditions and in other jurisdictions.

Sociodemographic characteristics, prevalence of chronic conditions and sequence of condition occurrence were generally comparable across provinces, although we observed some important differences in both the prevalence of conditions and use of health services, such as services for mental health and addictions.

Several studies have reported that between 18% and 35% of people with dementia have five or more comorbid chronic conditions.^{Footnote 6Footnote 22Footnote 23Footnote 24} We found this prevalence to be higher (38.8%–53.3%) across the four provinces, which may be driven in part by the most prevalent conditions, that is, mental illness and alcohol- or drug-induced disorders, hypertension and osteoarthritis. However, it is difficult to compare estimates across studies as there is no single definition of comorbidity and various conditions are represented, including many we did not examine (such as retinal disorders, liver disease, thyroid disease, cardiac arrhythmia, prostatic hypertrophy and insomnia).^{Footnote 25Footnote 26Footnote 27}

To our knowledge, few studies have examined the timing of occurrence of a broad set of chronic conditions relative to dementia. Most studies examine a narrower set of conditions and focus on vascular-related conditions as risk factors for incident dementia such as hypertension in mid-life,^{Footnote 28} diabetes^{Footnote 29} and stroke.^{Footnote 30} Recent studies have noted that neurodegenerative conditions^{Footnote 31} and hospitalization for traumatic brain injury^{Footnote 32} are associated with incident dementia. We observed similar patterns of sequencing for vascular-related conditions across the provinces (most were present before dementia was ascertained). However, we found traumatic brain injury and congestive heart failure to be detected nearly as often after dementia ascertainment as before. Traumatic brain injuries increase the risk of dementia mainly when they occur in mid-life,^{Footnote 14} which precedes the look-back period in our study and may explain this finding. Traumatic brain injury is also associated with falls by older people, including after the onset of dementia when mobility and coordination may be impaired.

As with other conditions measured, we could only capture data on mental illness (including depression), an established risk factor for dementia, through proxy measurement of health service use, which does not represent a clinical diagnosis. Parkinsonism, including Parkinson disease, was also more frequently ascertained prior to dementia. This is an expected pattern as Lewy body dementia can develop as Parkinson disease progresses. The increased burden of chronic conditions among people with dementia creates complex challenges in simultaneously managing health care needs for health care providers, patients and care partners.

Strengths and limitations

Our study presents insights from population-based data on persons living with dementia in four Canadian provinces. These data can serve as robust comparators to data from other regions.

This study has certain limitations. At a broad level, differences in the availability and structure of provincial administrative databases make comparative research across provinces and territories difficult. However, we leveraged standardized case definitions used in national disease surveillance to enhance comparisons.^{Footnote 1}

More specifically, while we focussed on a comprehensive set of chronic conditions with standardized definitions, we did not include other important conditions and risk factors such as smoking, alcohol consumption and physical inactivity. Second, variability in methods and database availability and completeness (e.g. differences in coding, spaces for multiple diagnoses on health claims and unobtainable emergency department data in some provinces) may have contributed to the differences observed between provinces. In addition, differences in the organization of health and social care services across provinces may explain some of the differences in prevalence of chronic conditions observed, for example, psychosocial care in QC may be more frequently provided outside of medical settings, potentially resulting in our underestimating mental illness and alcohol- or drug-induced disorders in this study. Data availability also impacted case definitions for some conditions, for example, the prevalence of traumatic brain injury may have been underreported because of a lack of available emergency department data in the participating provinces.

Third, we were limited to the case ascertainment date according to administrative databases, but administrative data case definitions do not represent formal diagnoses, and individuals may have first experienced the condition much earlier than identified. Fourth, the dementia case ascertainment algorithm, as with other health administrative data case definitions, is imperfect and may misclassify some individuals. For example, those with earlier stage dementia may be missed because of a lack of a formal diagnosis. Individuals who have not presented to the health system will also not be captured. Future studies should explore differences in comorbid chronic conditions among people with dementia across age, sex, indicators of socioeconomic status and rurality.

Conclusion

Using national, standardized definitions for chronic conditions and a minimum common dataset approach, we compared the prevalence and sequence of occurrence of comorbid chronic conditions among people aged 65 years and older with dementia in four Canadian provinces. Our study found generally similar patterns of comorbid conditions across the provinces, suggesting that strategies for care management, resource planning and health system access could be shared across regions.

Acknowledgements

This project was made possible through the collaboration of the Public Health Agency of Canada, ICES and the provincial governments of British Columbia, Ontario, Quebec and Prince Edward Island. Parts of this material are based on data or information compiled and provided by the Canadian Institute for Health Information (CIHI), Statistics Canada, IQVIA Solutions Canada Inc., the Ontario Ministry of Health, the Ontario Ministry of Long-Term Care and Health PEI. This study used data adapted from the Statistics Canada Postal Code^OM Conversion File, which is based on data licensed from Canada Post Corporation, and/or data adapted from the Ontario Ministry of Health Postal Code Conversion File, which contains data copied under licence from Canada Post Corporation and Statistics Canada.

The authors thank Henry Ngo and Andrea Olmstead for providing analytical support. We thank IQVIA Solutions Canada Inc. for the use of their drug information file. We thank the Toronto Community Health Profiles Partnership for providing access to the Ontario Marginalization Index.

The analyses, conclusions, opinions and statements expressed herein are solely those of the authors and do not reflect those of the funding organizations or data sources; no endorsement by the Public Health Agency of Canada, ICES, the Ontario Ministry of Health, the Ontario Ministry of Long-Term Care, Health PEI, Canadian Institute for Health Information (CIHI), Statistics Canada, IQVIA Solutions Canada Inc., the provincial governments of British Columbia, Ontario, Quebec and Prince Edward Island or the Government of Canada is intended or should be inferred.

Funding

This study received funding from the Public Health Agency of Canada (British Columbia: 6D02303002, Ontario: 6D02303001, Quebec: 4500413802, Prince Edward Island: 4500413866). This study was also supported by ICES, which is funded by an annual grant from the Ontario Ministry of Health and the Ministry of Long-Term Care.

Conflicts of interest

Susan E. Bronskill receives funding from the Public Health Agency of Canada (PHAC), the Canadian Institutes of Health Research (CIHR) and the Ontario Brain Institute (OBI) and support from ICES, which is funded by the Ontario Ministry of Health and the Ontario Ministry of Long-Term Care.

J. Michael Paterson receives funding from the CIHR, funding from PHAC, in support of the Canadian Chronic Disease Surveillance System (CCDSS), and support from ICES, which is funded by the Ontario Ministry of Health and the Ontario Ministry of Long-Term Care.

Andrea Gruneir receives grants from the CIHR.

Karen A.M. Phillips and Rasaq Ojasanya receive funding from PHAC via the Department of Health and Wellness, Government of Prince Edward Island.

Xibiao Ye receives funding from PHAC via the British Columbia Ministry of Health.

Kayla McLean receives funding from PHAC via the British Columbia Ministry of Health.

Victoria Massamba receives funding from PHAC via the Institut national de santé publique du Québec (INSPQ) and the Fonds de recherche du Québec – Santé and grants from Health Canada’s Substance Use and Addictions Program.

Isabelle Vedel receives funding from PHAC via the Institut national de santé publique du Québec (INSPQ) and the Fonds de recherche du Québec – Santé and grants from the CIHR.

Colleen J. Maxwell receives grants from the CIHR.

Authors’ contributions and statement

• SB: Conceptualization, methodology, validation, resources, writing – original draft, visualization, supervision, project administration, funding acquisition.
• AA: Conceptualization, methodology, validation, writing – original draft, visualization.
• LCM: Conceptualization, methodology, validation, writing – original draft, visualization.
• XW: Methodology, software, validation, formal analysis, data curation, writing – review & editing.
• HC: Methodology, validation, writing – original draft, visualization.
• JMP: Conceptualization, methodology, validation, writing – review & editing.
• AG: Conceptualization, methodology, validation, writing – review & editing.
• KAP: Methodology, software, validation, formal analysis, data curation, writing – review & editing.
• RO: Methodology, software, validation, formal analysis, data curation, writing – review & editing.
• XY: Methodology, software, validation, formal analysis, data curation, writing – review & editing.
• KM: Methodology, software, validation, formal analysis, data curation, writing – review & editing.
• FE: Methodology, software, validation, formal analysis, data curation, writing – review & editing.
• CGS: Methodology, validation, writing – review & editing.
• VM: Methodology, validation, writing – review & editing.
• LR: Methodology, software, validation, formal analysis, data curation, writing – review & editing.
• IV: Methodology, validation, writing – review & editing.
• LS: Conceptualization, methodology, validation, writing – review & editing.
• CP: Conceptualization, methodology, validation, writing – review & editing.
• CM: Conceptualization, methodology, validation, writing – review & editing.

The content and views expressed in this article are those of the authors and do not necessarily reflect those of the Government of Canada.

Table of Contents | Next