Abstract
Background:
Lower health-related quality of life (HRQoL) has been shown to predict a higher risk of hospital readmission and mortality in patients with cardiovascular disease (CVD). Few studies have explored the associations between HRQoL and incident CVD. We explored the associations between baseline HRQoL and incident and fatal CVD in community-dwelling older people in Australia and the United States.
Methods:
Longitudinal study using ASPirin in Reducing Events in the Elderly (ASPREE) trial data. This includes 19,106 individuals aged 65–98 years, initially free of CVD, dementia, or disability, and followed between March 2010 and June 2017. The physical (PCS) and mental component scores (MCS) of HRQoL were assessed using the SF-12 questionnaire. Incident major adverse CVD events included fatal CVD (death due to atherothrombotic CVD), hospitalizations for heart failure, myocardial infarction or stroke. Analyses were performed using Cox proportional-hazard regression.
Results:
Over a median 4.7 follow-up years, there were 922 incident CVD events, 203 fatal CVD events, 171 hospitalizations for heart failure, 355 fatal or nonfatal myocardial infarction and 403 fatal or nonfatal strokes. After adjustment for sociodemographic, health-related behaviours and clinical measures, a 10-unit higher PCS, but not MCS, was associated with a 14% lower risk of incident CVD, 28% lower risk of hospitalization for heart failure and 15% lower risk of myocardial infarction. Neither PCS nor MCS was associated with fatal CVD events or stroke.
Conclusion:
Physical HRQoL can be used in combination with clinical data to identify the incident CVD risk among older individuals.
Keywords: Quality of life, Cardiovascular diseases, Incidence, Risk factors, Aged
1. Introduction
Cardiovascular disease (CVD) is the leading cause of death, contributing to one-third of global deaths per year [1]. CVD also represents the greatest cause of long-term disability claims for older people [2]. Globally, CVD-related direct healthcare costs exceed those of any other disease [3]. In 2010, the global cost of CVD was estimated at 863 billion USD and it is expected to reach 1044 billion USD by 2030 [4]. Hence, the global burden of CVD, combined with the rapid growth in the ageing population [5], adds pressure on healthcare systems and society to identify those most at risk of adverse outcomes [6].
Health-related quality of life (HRQoL) is the individual’s subjective perception of health status on their overall functioning and well-being, including physical, mental, emotional and social functioning domains [7]. Moreover, HRQoL is considered an important patient-reported outcome measure for interventions and treatments in patients with CVD [8,9]. Lower HRQoL has been shown to predict a higher risk of hospital readmission and mortality in patients with CVD such as heart failure or ischemic heart disease [9,10], and CVD mortality among various community-dwelling samples, including in later life [11–14].
Some studies have also demonstrated that lower HRQoL is associated with higher CVD incidence among community-dwelling, predominantly middle-aged adults [15–19]. However, almost all previous studies focused on specific CVD subtypes [15–17,19] and most included samples of less than 10,000 individuals [15,17,19]. Given that older people are at a higher risk of CVD due to plaque build-up along the artery walls, changes in blood vessels, and increased risk of atrial fibrillation [20], it is of interest to investigate the association between HRQoL and the risk of incident CVD among older people in the community. To our knowledge, no other study has explored these novel associations in a large community-dwelling cohort of healthy older people.
Therefore, the overarching aim of our study was to explore whether physical (PCS) and mental component scores (MCS) of HRQoL are associated or not with incident and fatal CVD events in community-dwelling older people who were initially free of overt CVD.
2. Methods
2.1. Study population
This prospective study used five-year longitudinal data from the ASPirin in Reducing Events in the Elderly (ASPREE) clinical trial. In brief, ASPREE was a double-blind, community-based randomized controlled trial of low dose aspirin for primary prevention of disability and dementia in initially ‘healthy’ older people from Australia and the United States (U.S.) [21]. ASPREE recruited a total of 19,114 people aged 65 years and older who were free of CVD, dementia or other major life-limiting diseases likely to be fatal within the next five years [22]. The sample recruitment in Australia was mainly through general practices, and the U.S. sample was recruited through clinical trial and academic centres [21,22]. The ASPREE cohort was recruited between March 2010 and December 2014, and followed prospectively until June 12, 2017 [21,22]. Of the entire cohort, 19,106 (99.96%) completed the HRQoL questionnaire at baseline and are included in the present study. Therefore, based upon a type I error of 0.05 (two-tailed), our study was able to detect a Hazard Ratio of 0.95 (i.e. 5% reduced risk) with 94% power.
The ASPREE study complies with the Declaration of Helsinki, and was approved by multiple Institutional Review Boards (www.aspree.org). All participants signed informed consent on participation.
2.2. Determinants: health-related quality of life
HRQoL was measured by using the validated Medical Outcomes Study 12-item short-form (SF-12, version-2) questionnaire at the baseline [23,24]. The SF-12 estimates two summary measures – the physical (PCS) and mental component scores (MCS) [23]. They were derived using a norm-based scoring algorithm with a mean of 50 and a standard deviation of 10, with higher scores indicating a better HRQoL [23].
2.3. Outcome measures
The main outcomes were (a) incident cardiovascular disease (CVD) and (b) fatal CVD; and the sub-outcomes included hospitalization for heart failure, fatal or nonfatal myocardial infarction (MI), and fatal or nonfatal stroke.
In Australia and the U.S., detailed medical records, and imaging documentation (computed tomographic scans or magnetic resonance images scans) were used to identify all incident CVD events [25]. For the adjudication of the underlying causes of CVD, ASPREE staff, blinded to treatment allocation, prepared case summaries using the clinical details from clinicians, hospitals, nursing homes, and hospices [25,26]. Information from the next of kin or family members and death certificates from the government state-based registries were also sourced [26]. The case summaries were then provided to blinded Australian or U.S. expert committees, two of whom conducted web-based adjudication [26]. A third adjudicator resolved discordant adjudications [26].
2.3.1. Incident cardiovascular events
Incident CVD was a prespecified composite secondary end-point of ASPREE [25], consisting of fatal coronary heart disease, nonfatal MI, fatal or nonfatal stroke, or hospitalization for heart failure. Fatal coronary heart disease included fatal MI, sudden cardiac death or any other death in which the underlying cause of death was coronary heart disease. Fatal stroke was defined as any death due to the rapid onset of a new neurological deficit attributed to obstruction or rupture in the intracranial or extra-cranial cerebral arterial system. Fatal CVD was defined as death due to an ischemic event, including MI, or other coronary heart diseases, sudden cardiac death, cardiac failure death (with coronary cause) or stroke.
Nonfatal MI was defined according to the joint guidelines of the European Society of Cardiology and the American College of Cardiology [27]. Hospitalization for heart failure was defined as any unplanned overnight stay or longer in a hospital or similar facility with heart failure as the principal reason for admission. The criteria for the diagnosis of nonfatal stroke was identified by the World Health Organization (WHO) [28].
2.4. Covariates
Sociodemographic factors, health-related behaviours, and clinical measures which are known to be related to HRQoL or CVD, were considered as potential covariates in this study [1,17,18,29]. Sociodemographic factors at baseline included age in years, sex (male or female), years of education (<12-years or ≥ 12-years), living situation (living alone or with family/others) and country and ethno-racial group (white Australian, white American, black/African-American, or Hispanic/Latino/Asiatic/Other). Three baseline health-related behaviours were smoking status (never, former, or current), alcohol consumption (never, former, current low risk with ≤10 standard drinks per week / ≤4 standard drinks on any one day, or current high risk with >10 standard drinks per week / >4 standard drinks on any one day) [30] and a proxy of physical activity (i.e. ‘physical ability’, collected as the longest amount of time walking outside their home without any rest in the last two weeks; classified as none, ≤15 min, 16–30 min, or > 30 min). In an ASPREE subsample of 12,667 older Australians [31], individuals with a high level of ‘physical ability’ were more likely to engage in moderate or vigorous physical activities (age and sex-adjusted OR = 4.55, p-value <0.001).
The clinical measures at baseline consisted of (a) hypertension (yes or no, based on whether the average of three blood pressure measurements was systolic blood pressure ≥140 mmHg or diastolic blood pressure ≥90 mmHg and/or whether the participants were on treatment for high blood pressure); (b) diabetes (yes or no, based on self-report of diabetes or fasting glucose ≥126 mg/dL or on treatment for diabetes); (c) dyslipidaemia (yes or no, based on the use of cholesterol-lowering medications or low-density lipoprotein, LDL > 160 mg/dL (>4.1 mmol/L) or serum cholesterol ≥212 mg/dL (≥5.5 mmol/L) for Australians or ≥ 240 mg/dL (≥6.2 mmol/L) for U.S. participants) [22]; (d) plasma/serum creatinine level in mg/dL; (e) body mass index (BMI, based on measured body weight (kg) divided by height (m) squared); and (f) depressive symptoms measured by the Center for Epidemiologic Studies Depression scale (CESD) 10 [32].
2.5. Statistical analyses
Cox proportional-hazards regression models with time-to-event analysis were used to examine the associations between PCS/MCS (i.e. 10-units difference) and (a) the two main outcomes – incident and fatal CVD; and (b) three sub-outcomes – hospitalization for heart failure, fatal or nonfatal MI, and fatal or nonfatal stroke. Cause-specific hazards were estimated for outcomes in relation to PCS/MCS, with deaths from causes other than our interested outcomes treated as censoring events. The ‘entry date’ was the baseline HRQoL administered date when the participants were randomized in the ASPREE study. The ‘end date’ was the date of CVD end point (or the censored date June 12, 2017).
The crude results were adjusted for possible confounders using two levels of adjustment. The first model adjusted for age only, aiming to control for the strongest associated covariate with both HRQoL and CVD events [29]. The second model adjusted for age, sex, education level, living situation, ethno-racial group, smoking status, alcohol consumption, the longest amount of time walking outside home without any rest, clinical history of hypertension, diabetes or dyslipidaemia, creatinine levels, current BMI, and depressive symptoms score. Additionally, PCS and MCS were also examined in quartiles and then we plotted the crude cumulative incidence curves for each outcome, considering the competing risk of death from other causes rather than our interested outcome.
Given that sex plays a role in social consequences of health especially in health-care seeking behaviours and availability of family support [33] and CVD risk differs by sex [20], we conducted analyses stratified by sex. In addition, we repeated analyses stratified by country given the social, cultural and healthcare differences between Australia and the U.S. Furthermore, the interactions between PCS/MCS and age group (<75-years and ≥75-years) were tested by introducing multiplicative terms in the regression models.
Sensitivity firstly involved re-running the analyses with both PCS and MCS in the models. Secondly, to assess the possibility of reverse causality, the association between HRQoL and outcomes were repeated after excluding those with a positive outcome or censored within 12 months of follow-up. Thirdly, using PCS/MCS based on the weighted norms for SF-36 (version-2) in Australian population (Mancuso S. Personal communication with Freak-Poli R and Phyo AZZ. Sep 29–Oct 4, 2020) [34,35]. Fourthly, to investigate associations with a single SF-12 index measure using the SF-6D measure, (i.e. overall measure of HRQoL using the preference-based algorithm provided by the University of Sheffield, under license) [36] as the main exposure variable.
STATA version 16.0 was used for the statistical analyses (Stata-CorpLP, College Station, Texas, the U.S.).
3. Results
Of the eligible 19,106 participants, just over half were female (56.4%) and the median age at baseline was 74 years (interquartile range 71.6–77.7 years; Table 1). Individuals who reached an incident or fatal CVD event during follow-up had lower PCS scores, and were more likely to be older, male, have less education, living alone, former/current smoker, former/high-risk current alcohol drinker, and had less physical ability or more comorbidities at baseline (Table 1).
Table 1.
Sample distribution and incident and fatal cardiovascular disease according to baseline characteristics (n = 19,106).
| Baseline characteristics | Total sample distribution |
Incident CVD a |
Fatal CVD b |
||||
|---|---|---|---|---|---|---|---|
| (n = 19,106) |
Yes (n = 922) |
No (n = 18,184) |
p-value |
Yes (n = 203) |
No (n = 18,903) |
p-value |
|
| n (%) | n (%) | n (%) | n (%) | n (%) | |||
|
| |||||||
| Sociodemographic factors | |||||||
| Age in years | |||||||
| Median (IQR) | 74.0 (71.6–77.7) | 76.3(72.7–81.0) | 73.9 (71.6–77.5) | <0.001 | 79.1 (73.4–83.5) | 74.0 (71.6–77.6) | <0.001 |
| Sex | |||||||
| Male | 8329 (43.6%) | 508 (55.1%) | 7821 (43.0%) | <0.001 | 124 (61.1%) | 8205 (43.4%) | <0.001 |
| Female | 10,777 (56.4%) | 414 (44.9%) | 10,363 (57.0%) | 79 (38.9%) | 10,698 (56.6%) | ||
| Years of education | |||||||
| Under 12 years | 8634 (45.2%) | 446 (48.4%) | 8188 (45.0%) | 0.05 | 109 (53.7%) | 8525 (45.1%) | 0.01 |
| 12 years and above | 10,471 (54.8%) | 476 (51.6%) | 9995 (55.0%) | 94 (46.3%) | 10,377 (54.9%) | ||
| Living situation | |||||||
| At home alone | 6249 (32.7%) | 358 (38.8%) | 5891 (32.4%) | <0.001 | 83 (40.9%) | 6166 (32.6%) | 0.01 |
| With family or others | 12,857 (67.3%) | 564 (61.2%) | 12,293 (67.6%) | 120 (59.1%) | 12,737 (67.4%) | ||
| Ethno-racial group | |||||||
| White Australian | 16,355 (85.6%) | 784 (85.0%) | 15,571 (85.6%) | 0.60 | 163 (80.3%) | 16,192 (85.7%) | 0.01 |
| White American | 1088 (5.7%) | 59 (6.4%) | 1029 (5.7%) | 10 (4.9%) | 1078 (5.7%) | ||
| Black/African-American | 900 (4.7%) | 47 (5.1%) | 853 (4.7%) | 20 (9.9%) | 880 (4.7%) | ||
| Hispanic/Latino/Asiatic/Otherc | 763 (4.0%) | 32 (3.5%) | 731 (4.0%) | 10 (4.9%) | 753 (4.0%) | ||
| Health-related behaviours | |||||||
| Smoking status | |||||||
| Never | 10,575 (55.4%) | 434 (47.1%) | 10,141 (55.8%) | <0.001 | 93 (45.8%) | 10,482 (55.5%) | 0.003 |
| Former | 7796 (40.8%) | 438 (47.5%) | 7358 (40.5%) | 95 (46.8%) | 7701 (40.7%) | ||
| Current | 735 (3.9%) | 50 (5.4%) | 685 (3.8%) | 15 (7.4%) | 720 (3.8%) | ||
| Alcohol consumption | |||||||
| Never | 3333 (17.4%) | 165 (17.9%) | 3168 (17.4%) | 0.06 | 42 (20.7%) | 3291 (17.4%) | 0.01 |
| Former | 1135 (5.9%) | 71 (7.7%) | 1064 (5.9%) | 21 (10.3%) | 1114 (5.9%) | ||
| Current (low risk) | 10,094 (52.8%) | 458 (49.7%) | 9636 (53.0%) | 90 (44.3%) | 10,004 (52.9%) | ||
| Current (high risk) | 4544 (23.8%) | 228 (24.7%) | 4316 (23.7%) | 50 (24.6%) | 4494 (23.8%) | ||
| Average longest amount of walking time outside home without any rest (last 2 weeks) | |||||||
| None | 839 (4.4%) | 50 (5.4%) | 789 (4.4%) | <0.001 | 15 (7.4%) | 824 (4.4%) | <0.001 |
| ≤15 min | 2341 (12.3%) | 169 (18.4%) | 2172 (12.0%) | 40 (19.8%) | 2301 (12.2%) | ||
| 16–30 min | 4144 (21.7%) | 229 (24.9%) | 3915 (21.6%) | 60 (29.7%) | 4084 (21.7%) | ||
| >30 min | 11,741 (61.6%) | 472 (51.3%) | 11,269 (62.1%) | 87 (43.1%) | 11,654 (61.8%) | ||
| Clinical measures | |||||||
| Hypertensiond | |||||||
| Yes | 14,191 (74.3%) | 753 (81.7%) | 13,438 (73.9%) | <0.001 | 166 (81.8%) | 14,025 (74.2%) | 0.01 |
| No | 4915 (25.7%) | 169 (18.3%) | 4746 (26.1%) | 37 (18.2%) | 4878 (25.8%) | ||
| Diabetese | |||||||
| Yes | 2044 (10.7%) | 109 (11.8%) | 1935 (10.6%) | 0.26 | 27 (13.3%) | 2017 (10.7%) | 0.23 |
| No | 17,062 (89.3%) | 813 (88.2%) | 16,249 (89.4%) | 176 (86.7%) | 16,886 (89.3%) | ||
| Dyslipidaemia f | |||||||
| Yes | 12,464 (65.2%) | 588 (63.8%) | 11,876 (65.3%) | 0.34 | 122 (60.1%) | 12,342 (65.3%) | 0.12 |
| No | 6642 (34.8%) | 334 (36.2%) | 6308 (34.7%) | 81 (39.9%) | 6561 (34.7%) | ||
| Creatinine level (mg/DL) | |||||||
| Mean (SD) | 0.9 (0.2) | 1.0 (0.3) | 0.9 (0.2) | <0.001 | 1.0 (0.3) | 0.9 (0.2) | <0.001 |
| Body mass index (kg/m2) | |||||||
| Mean (SD) | 28.1 (4.7) | 28.2 (5.1) | 28.1 (4.7) | 0.65 | 27.6 (5.7) | 28.1 (4.7) | 0.11 |
| CES-D-10 score | |||||||
| Mean (SD) | 3.2 (3.3) | 3.3 (3.5) | 3.2 (3.3) | 0.31 | 3.6 (3.5) | 3.2 (3.3) | 0.10 |
| Health-related quality of life | |||||||
| Physical component score | |||||||
| Mean (SD) | 48.3 (8.8) | 46.3 (9.3) | 48.4 (8.7) | <0.001 | 45.6 (10.2) | 48.4 (8.7) | <0.001 |
| Mental component score | |||||||
| Mean (SD) | 55.7 (7.1) | 55.8 (7.7) | 55.7 (7.1) | 0.54 | 55.6 (7.5) | 55.7 (7.1) | 0.91 |
CES-D-10 score, Center for Epidemiologic Studies Depression score; CVD, cardiovascular disease; IQR, interquartile range; SD, standard deviation.
All p-values are from chi-squared or t-test or Wilcoxon rank-sum test comparison of participants with incident CVD or fatal CVD versus those without incident CVD or fatal CVD. For categorical variables, it represents the incidence of the outcome between categories of the variables (eonfounders). For numerical variables (including body mass index, creatinine, health-related quality of life), this is about differences in the mean of the exposure variable between those with a positive or negative outcome.
Incident CVD includes fatal coronary heart disease (death from myocardial infarction, sudden cardiac death, or any other death in which the underlying cause was coronary heart disease), nonfatal myocardial infarction, fatal or nonfatal stroke (including haemorrhagic stroke), or hospitalization for heart failure.
Fatal CVD includes death from myocardial infarction, sudden cardiac death, stroke or any other death in which the underlying cause was coronary heart disease.
Other included Aboriginal and Torres Strait Islander, Native Hawaiian, other Pacific Islander, Maori, American Indian, or more than one race.
Hypertension = systolic blood pressure ≥ 140 mmHg or diastolic blood pressure ≥ 90 mmHg and/or on treatment for high blood pressure.
Diabetes mellitus = self-report of diabetes or fasting glucose ≥126 mg/dL or on treatment for diabetes.
Dyslipidaemia = cholesterol-lowering medications or serum cholesterol ≥212 mg/dL (≥5.5 mmol/L; Australia) and ≥ 240 mg/dL (≥6.2 mmol/L; U.S.) or low-density lipoprotein >160 mg/dL (>4.1 mmol/L).
Over the median 4.7-years follow-up (interquartile range: 3.6 to 5.7 years), there were 922 total incident CVD events (11.0 events per 1000 person-years), including 203 fatal CVD (2.3 events per 1000 person-years), 171 hospitalizations for heart failure (2.0 events per 1000 person-years), 355 fatal or nonfatal MI (4.2 events per 1000 person-years), and 403 fatal or nonfatal strokes (4.7 events per 1000 person-years) (Tables 2 and 3).
Table 2.
Baseline physical component score (PCS) as a predictor of cardiovascular disease events (n = 19,106).
| Cardiovascular disease events | Events | Crude |
Model 1 |
Model 2 |
|||
|---|---|---|---|---|---|---|---|
| HR a | 95% CI | HR a | 95% CI | HR a | 95% CI | ||
|
| |||||||
| Incident cardiovascular disease | 922 | 0.76 | (0.71–0.81)* | 0.82 | (0.76–0.87)* | 0.86 | (0.79–0.92)* |
| Fatal cardiovascular disease | 203 | 0.71 | (0.62–0.82)* | 0.82 | (0.71–0.95)* | 0.86 | (0.73–1.01) |
| CVD subtypes | |||||||
| Hospitalization for heart failure | 171 | 0.58 | (0.50–0.67)* | 0.65 | (0.56–0.75)* | 0.72 | (0.60–0.85)* |
| Fatal or nonfatal myocardial infarction | 355 | 0.78 | (0.70–0.86)* | 0.81 | (0.73–0.91)* | 0.85 | (0.75–0.96)* |
| Fatal or nonfatal stroke | 403 | 0.82 | (0.74–0.92)* | 0.89 | (0.80–0.99)* | 0.90 | (0.80–1.02) |
Model 1 adjusted for age only; Model 2 additionally adjusted for sex, education, living situation, ethno-racial group, smoking status, alcohol consumption, average longest amount of walking time outside home without any rest (last 2 weeks), hypertension, diabetes, dyslipidaemia, creatinine level, body mass index, and depressive symptom score.
Incident cardiovascular disease providing 83,997.9 person-years of observation (mean: 4.4 ± SD1.4 years, median: 4.5 years, IQR: 3.4 to 5.6 years, range: 0 and 7.3 years of observation).
Fatal cardiovascular disease providing 88,345.0 person-years of observation (mean: 4.6 ± SD1.3 years, median: 4.7 years, IQR: 3.6 to 5.7 years, range: 0 and 7.3 years of observation).
Hospitalization for heart failure providing 85,279.0 person-years of observation (mean: 4.5 ± SD1.4 years, median: 4.6 years, IQR: 3.5 to 5.6 years, range: 0 and 7.3 years of observation).
Fatal or nonfatal myocardial infarction providing 84,875.2 person-years of observation (mean: 4.4 ± SD1.4 years, median: 4.6 years, IQR: 3.4 to 5.6 years, range: 0 and 7.3 years of observation).
Fatal or nonfatal stroke providing 84,890.5 person-years of observation (mean: 4.4 ± SD1.4 years, median: 4.6 years, IQR: 3.4 to 5.6 years, range: 0 and 7.3 years of observation).
p-value <0.05.
For every 10-unit increase in PCS.
Table 3.
Baseline mental component score (MCS) as a predictor of cardiovascular disease events (n = 19,106).
| Cardiovascular disease events | Events | Crude |
Model 1 |
Model 2 |
|||
|---|---|---|---|---|---|---|---|
| HR a | 95% CI | HR a | 95% CI | HR a | 95% CI | ||
|
| |||||||
| Incident cardiovascular disease | 922 | 1.01 | (0.92–1.11) | 1.00 | (0.92–1.10) | 1.03 | (0.93–1.14) |
| Fatal cardiovascular disease | 203 | 0.98 | (0.80–1.18) | 0.96 | (0.80–1.16) | 1.07 | (0.86–1.32) |
| CVD subtypes | |||||||
| Hospitalization for heart failure | 171 | 1.18 | (0.94–1.49) | 1.16 | (0.93–1.45) | 1.25 | (0.98–1.59) |
| Fatal or nonfatal myocardial infarction | 355 | 1.00 | (0.86–1.16) | 0.99 | (0.86–1.15) | 0.98 | (0.83–1.16) |
| Fatal or nonfatal stroke | 403 | 0.96 | (0.84–1.11) | 0.96 | (0.84–1.10) | 0.99 | (0.85–1.15) |
Model 1 adjusted for age only; Model 2 additionally adjusted for sex, education, living situation, ethno-racial group, smoking status, alcohol consumption, average longest amount of walking time outside home without any rest (last 2 weeks), hypertension, diabetes, dyslipidaemia, creatinine level, body mass index, and depressive symptom score.
Incident cardiovascular disease providing 83,997.9 person-years of observation (mean: 4.4 ± SD1.4 years, median: 4.5 years, IQR: 3.4 to 5.6 years, range: 0 and 7.3 years of observation).
Fatal cardiovascular disease providing 88,345.0 person-years of observation (mean: 4.6 ± SD1.3 years, median: 4.7 years, IQR: 3.6 to 5.7 years, range: 0 and 7.3 years of observation).
Hospitalization for heart failure providing 85,279.0 person-years of observation (mean: 4.5 ± SD1.4 years, median: 4.6 years, IQR: 3.5 to 5.6 years, range: 0 and 7.3 years of observation).
Fatal or nonfatal myocardial infarction providing 84,875.2 person-years of observation (mean: 4.4 ± SD1.4 years, median: 4.6 years, IQR: 3.4 to 5.6 years, range: 0 and 7.3 years of observation).
Fatal or nonfatal stroke providing 84,890.5 person-years of observation (mean: 4.4 ± SD1.4 years, median: 4.6 years, IQR: 3.4 to 5.6 years, range: 0 and 7.3 years of observation).
For every 10-unit increase in MCS.
An inverse association was found between PCS and the risk of incident CVD (Table 2 and Fig. 1). The cumulative incidence at the median follow-up time indicated a dose-response relationship, with the lowest PCS quartile had twice the incidence compared to the highest PCS quartile (6% and 3% respectively; Fig. 1). For heart failure, this ratio appeared to be even higher, but basically, all outcomes showed a similar pattern (Fig. 1). After adjustment for all potential confounders, a 10-unit higher PCS was associated with a lower risk of incident outcomes: 14% for incident CVD, 28% for hospitalization for heart failure, and 15% for fatal or nonfatal MI (Table 2). With respect to fatal CVD and stroke, the associations were no longer present after adjustment for all potential confounders (Table 2). However, when ‘physical ability’ was removed from the adjusted models, a higher HRQoL was associated with a lower risk of all CVD events (18% each for incident CVD, fatal CVD and fatal or nonfatal MI, 31% for hospitalization for heart failure, and 12% for stroke; data not shown). MCS was not associated with CVD events (Table 3 and Supplementary Fig. S1). There was no evidence that sex, country or age group modified any of the associations (p-values for interaction >0.05 in all cases). Sex and country stratified analyses are shown in Supplementary Tables S1–S4.
Fig. 1.
Cumulative incidence of cardiovascular disease, fatal cardiovascular disease, hospitalization for heart failure, fatal or nonfatal myocardial infarction, and fatal or nonfatal stroke according to the physical component score (PCS) in quartiles. PCS, physical component score; Q1, quartile 1; Q2, quartile 2; Q3, quartile 3; Q4, quartile 4. Quartile 1 = lowest, Quartile 4 = highest.
3.1. Sensitivity analyses
The observed associations remained consistent (a) when both PCS and MCS were included in the models (Supplementary Table S5), (b) after exclusion of events occurring in the first year of study follow-up (Supplementary Table S6) and (c) when PCS and MCS were calculated based on Australian population weights [34] (Supplementary Table S7). When SF-6D was examined [36], the magnitude of associations reported in Supplementary Table S8 were between the main findings for PCS and MCS reported in Tables 2 and 3, as expected. The SF-6D was associated with incident CVD and fatal or nonfatal MI (Supplementary Table S8). Hence, the SF-6D represents a mix of our PCS and MCS findings.
4. Discussion
Our study provides some of the first evidence of a longitudinal association between baseline PCS and incident CVD in relatively ‘healthy’ older people living in the community. In general, those in the lowest PCS quartile had twice the cumulative incidence of CVD compared to those in the highest PCS quartile at a median follow-up time of 4.7 years. Higher PCS, but not MCS was associated with a lower risk of incident CVD. These associations did not vary by sex, age or country.
Our findings align with previous studies using the same SF-12 instrument but predominantly conducted in community-based samples of middle-aged adults [15,17]. In both Scottish adults aged ≥20 years and Spanish individuals aged 35 to 74 years, individuals with the lowest PCS had a two-fold incident CVD risk compared to those with a better HRQoL over an average follow-up time of 7.6-years and 6.3-years, respectively [15,17]. In both of these studies, MCS was not associated with incident CVD risk [15,17]. In a US population-based study with 22,229 adults aged ≥45 years, a 10-unit higher PCS or MCS was associated with an 11% and 8% lower risk of incident CVD, respectively, over a median 8.4-years follow-up [18]. However, when PCS and MCS were treated as categorical variables in that study, lower MCS was only associated with CVD risk when PCS was also low [18]. Furthermore, a Swedish study with 1001 people aged 45–69 years also showed that mental health and role limitations due to emotional problems subscales of the SF-36 were not associated with 13-year coronary heart disease (CHD) incidence [19]. In contrast, most of the sub-scales contributed heavier weights in the PCS calculation, such as bodily pain and role limitations due to physical problems predicted CHD (84% and 67% higher CHD risks for the group with the lowest compared to that with the highest scores) [19]. Moreover, given our cohort was relatively healthy, HRQoL estimates in our study are consistently higher than those of other population-based studies [29]. However, our findings are consistent with previous evidence [15,17–19], highlighting that PCS can be used for CVD risk stratification across different populations.
PCS is the perception an individual has about their physical functioning, and the limitations of their functioning [23]. From the SF-12 questionnaire, it is calculated with a heavier weighting on questions about general physical function, limitations due to physical problem, and bodily pain [23]. The exact biological mechanisms underlying these associations is not clear. However, lower PCS is likely a marker of overall poorer physical health and thus could indicate a profile of CVD risk beyond the established risk factors which were considered in this analysis. Aligning with our findings, previous studies showed that lower heart rate variability was correlated with lower physical HRQoL [37], and low heart rate variability was also associated with an increased risk of CVD [38].
Of note, we found that older individuals who reached an incident CVD event during follow-up scored on average 2–3 points lower on baseline PCS than individuals without an event. Such differences determined cross-sectionally are less likely to reflect the clinically meaningful change [39]. However, it can be considered as the minimally important difference threshold between individuals with or without an incident CVD event [39]. Furthermore, we observed that the five-year risk of a CVD event (especially hospitalization for heart failure) among people with lower PCS is twice the risk of those with a better score. This is similar to the CVD risk of having diabetes, metabolic syndrome, or laboratory markers [40,41]. Therefore, the self-reported PCS HRQoL can be used in combination with clinical data to improve the detection of those at a higher risk of CVD events and predict further complications.
Our findings contrast with previous studies from Australia, Spain and the U.K. (predominantly middle-aged people, average follow-up of 6.5 to 14 years) that showed higher PCS was associated with a lower risk of stroke [16] or fatal CVD [11–14]. This discrepancy may be explained by residual confounding, as only one of these studies adjusted for physical activity in their analysis [16]. Physical activity is associated with both, CVD risk [42] and HRQoL [43], and it has been reported in the literature that higher levels of physical activity can attenuate the impact of poor HRQoL on CVD mortality [11]. Indeed, in our analysis, we adjusted for a proxy of physical activity (i.e. ‘physical ability’ – the longest amount of time walking outside their home without any rest). Interestingly, when we removed ‘physical ability’ from the models, a higher HRQoL was associated with a lower risk of stroke or fatal CVD, which is consistent with the conclusions from previous studies [11–14,16]. Additionally, the strengths of the associations between HRQoL and the other investigated CVD events became stronger after removing this covariate. In addition, given that different aetiology and pathophysiological pathways play a role in the development of different CVD diagnoses with a different impact at different ages [1,44–46], the beneficial impact of physical activity on these conditions may also vary. This, in turn, would partly explain our finding that PCS is associated with other CVD outcomes different from fatal CVD or stroke.
4.1. Strengths and limitations
To the best of our knowledge, our study is the largest to explore the longitudinal association between HRQoL and a range of incident CVD sub-types in community-dwelling older people. Additionally, our unique and large sample represents initially relatively ‘healthy’ ageing populations from two countries with a different health system and sociocultural context. Moreover, our study outcomes were adjudicated by expert outcomes committees. This, in turn, enhances the quality of our study. Another strength of this study is the availability of adjustment for a variety of possible confounders associated with CVD and all-cause mortality [47], but often not included as potential covariates. Another strength is that we ascertained the vital status of all ASPREE 19,106 participants who completed the baseline HRQoL questionnaire, thus with no loss to follow-up over a median follow-up of 4.7 years. This study however, has some limitations. First, our study exclusively focuses on healthy older people; and the incident rate of CVD events was thus lower than in some other studies. This limits to some extent the generalisability of our findings as our results cannot be extrapolated to older individuals with pre-existing CVD. However, our results were similar to those of prior research [15,17–19]. Furthermore, there was strong country imbalance in our sample (13% U.S.), but our U.S. sample is representative of minority ethnic groups [21,22] and we found no evidence that country modified our investigated associations.
4.2. Clinical implications
Given our findings, the PCS SF-12 measure may help clinical decision-making for individuals, especially for those who may be approaching the treatment threshold with current risk stratification. Furthermore, this measure can be used in combination with clinical information to improve the early detection of older people at a high risk of CVD [48]. Given that the SF-12 measure consists of only 12 questions, uses easy-to-understand language, and takes an average of 2–3 min to complete [23], it could be implemented without significant burden to the respondent or health staff. In addition, it has been demonstrated that SF-12 promotes effective communication between patients and healthcare providers and predicts not only health outcomes but also medical expenditure [49]. Hence, the valuable, comprehensive, and complementary insight from SF-12 could identify high-risk individuals for age-related diseases, which in turn could help to reduce the global socio-economic burden, and promote patient-centred healthcare. Particularly, our finding supports the decision of the Australian Commission on Safety and Quality in Health Care to incorporate the SF-12 into the annual collection of patient-reported outcome measures [50].
5. Conclusion
This large, population-based longitudinal study over a median follow-up of 4.7 years provides some of the first evidence that higher PCS, but not MCS is associated with a lower risk of incident CVD in ‘healthy’ older people aged ≥65 years. Our results provide additional evidence that PCS of SF-12 can be used in combination with clinical information to improve the early detection of older people at a high CVD risk. Further research is needed to explore how potential mediating factors explain the association between PCS and CVD risk, and whether intervention focusing on increasing PCS will reduce the incidence of CVD.
Supplementary Material
Acknowledgements
We would like to thank the ASPREE participants who volunteered for this study, the general practitioners and staff of the medical clinics who support the study participants, and the trial staff and management team of the ASPREE study in Australia and the United States (www.aspree.org).
Funding
This was mainly supported by grants from the National Institute on Ageing and the National Cancer Institute at the U.S. National Institutes of Health (grant number U01AG029824 and U19AG062682); the National Health and Medical Research Council of Australia (grant numbers 334047 and 1127060); Monash University (Australia) and the Victorian Cancer Agency (Australia). Other funding resources and collaborating organizations of the ASPREE study are listed on http://www.aspree.org. AZZP is supported by Monash International Tuition Scholarship (Medicine, Nursing, and Health Sciences) and Monash Graduate Scholarship (30072360). RFP is supported by a National Heart Foundation of Australia Postdoctoral Fellowship (101927). JR and CMR are supported by a National Health and Medical Research Council Dementia Research Leader Fellowship (APP 1135727) and Principal Research Fellowship (APP1136372) respectively. Funders played no role in the design of the study, in the collection, analysis, and interpretation of data and in the writing of the manuscript.
Footnotes
Declaration of Competing Interest
All authors: No conflict of interest to declare.
Ethical considerations
The data of the present secondary data-analysis study was from a five-year ASPREE clinical trial (Trial Registration: International Standard Randomized Controlled Trial Number Register (ISRCTN 83772183) and clinicaltrials.gov (NCT 01038583)). The current secondary data analysis has been approved by the Monash University Human Research Ethics Committee (project ID: 21714). The ASPREE trial was approved by multiple Institutional Review Boards in Australia and the U.S. (www.aspree.org).
Informed consent
All individual participants of the ASPREE clinical trial signed informed consent on participation.
Research involving human participants
ASPREE is being conducted in accordance with the Declaration of Helsinki 1964 as revised in 2008, the NHMRC Guidelines on Human Experimentation, the federal patient privacy (HIPAA) law and ICH-GCP guidelines and the International Conference of Harmonization Guidelines for Good Clinical Practice. ASPREE also follows the Code of Federal Regulations as it relates to areas of clinical research. The overall management and conduct of the ASPREE clinical trial is the responsibility of the ASPREE Steering Committee.
Author statement
All authors take responsibility for all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.org/10.1016/j.ijcard.2021.07.004.
Data sharing statement
All individual participant data (re-identifiable) that underlie the results reported in this manuscript, are available upon request to qualified researchers without limit of time, subject to approval of the analyses by the Principal Investigators and a standard data sharing agreement. Details regarding requests to access the data will be available through the web site (www.ASPREE.org). The data will then be made available through a web-based data portal safe haven at Monash University, Australia.
References
- [1].World Health Organization, Cardiovascular Diseases 2017. [cited 2020 July 7]. Available from: https://www.who.int/en/news-room/fact-sheets/detail/cardiovascular-diseases-(cvds).
- [2].Murray CJL, Vos T, Lozano R, et al. , Disability-adjusted life years (DALYs) for 291 diseases and injuries in 21 regions, 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010, Lancet 380 (9859) (2012) 2197–2223. [DOI] [PubMed] [Google Scholar]
- [3].Tarride J-E, Lim M, Desmeules M, et al. , A review of the cost of cardiovascular disease, Can. J. Cardiol. 25 (6) (2009) e195–e202. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [4].Bloom DE, Cafiero ET, Jané-Llopis E, et al. , The Global Economic Burden of Noncommunicable Diseases, World Economic Forum, Geneva, 2011. [Google Scholar]
- [5].World Health Organization, Aging and Health 2018. [cited 2020 July 10]. Available from: https://www.who.int/news-room/fact-sheets/detail/ageing-and-health.
- [6].Roth GA, Johnson C, Abajobir A, et al. , Global, regional, and national burden of cardiovascular diseases for 10 causes, 1990 to 2015, J. Am. Coll. Cardiol. 70 (1) (2017) 1–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [7].Gandek B, Ware JE, Aaronson NK, et al. , Cross-validation of item selection and scoring for the SF-12 health survey in nine countries: results from the IQOLA project, J. Clin. Epidemiol. 51 (11) (1998) 1171–1178. [DOI] [PubMed] [Google Scholar]
- [8].Dyer MT, Goldsmith KA, Sharples LS, Buxton MJ, A review of health utilities using the EQ-5D in studies of cardiovascular disease, Health Qual. Life Outcomes 8 (1) (2010) 13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].Westin L, Nilstun T, Carlsson R, Erhardt L, Patients with ischemic heart disease: quality of life predicts long-term mortality, Scand. Cardiovasc. J. 39 (1–2) (2005) 50–54. [DOI] [PubMed] [Google Scholar]
- [10].Rodríguez-Artalejo F, Guallar-Castillón P, Pascual CR, et al. , Health-related quality of life as a predictor of hospital readmission and death among patients with heart failure, Arch. Intern. Med. 165 (11) (2005) 1274–1279. [DOI] [PubMed] [Google Scholar]
- [11].Higueras-Fresnillo S, Cabanas-Sánchez V, García-Esquinas E, Rodríguez-Artalejo F, Martinez-Gomez D, Physical activity attenuates the impact of poor physical, mental, and social health on total and cardiovascular mortality in older adults: a population-based prospective cohort study, Qual. Life Res. 27 (12) (2018) 3293–3302. [DOI] [PubMed] [Google Scholar]
- [12].Myint PK, Luben RN, Surtees PG, et al. , Relation between self-reported physical functional health and chronic disease mortality in men and women in the European Prospective Investigation into Cancer (EPIC–Norfolk): a prospective population study, Ann. Epidemiol. 16 (6) (2006) 492–500. [DOI] [PubMed] [Google Scholar]
- [13].Williams ED, Rawal L, Oldenburg BF, Renwick C, Shaw JE, Tapp RJ, Risk of cardiovascular and all-cause mortality: impact of impaired health-related functioning and diabetes: the Australian Diabetes, Obesity and Lifestyle (AusDiab) study, Diabetes Care 35 (5) (2012) 1067. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Myint PK, Smith RD, Luben RN, et al. , The short-form six-dimension utility index predicted mortality in the European Prospective Investigation into Cancer-Norfolk prospective population-based study, J. Clin. Epidemiol. 63 (2) (2010) 192–198. [DOI] [PubMed] [Google Scholar]
- [15].Ul-Haq Z, Mackay DF, Pell JP, Association between physical and mental health-related quality of life and adverse outcomes a retrospective cohort study of 5,272 Scottish adults, BMC Public Health 14 (2014) 1197. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [16].Myint KP, Surtees GP, Wainwright WJN, et al. , Physical health-related quality of life predicts stroke in the EPIC-Norfolk, Neurology 69 (24) (2007) 2243–2248. [DOI] [PubMed] [Google Scholar]
- [17].Muñoz M-A, Subirana I, Elosua R, et al. , Utility of a short quality of life questionnaire to predict cardiovascular events, Int. J. Cardiol. 151 (3) (2011) 392–394. [DOI] [PubMed] [Google Scholar]
- [18].Pinheiro L, Reshetnyak E, Sterling M, Richman J, Kern L, Safford M, Using health-related quality of life to predict cardiovascular disease events, Qual. Life Res. 28 (6) (2019) 1465–1475. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [19].Nilsson E, Festin K, Lowén M, Kristenson M, SF-36 predicts 13-year CHD incidence in a middle-aged Swedish general population, Qual. Life Res. 29 (4) (2019) 971–975. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [20].Rodgers JL, Jones J, Bolleddu SI, et al. , Cardiovascular risks associated with gender and aging, J. Cardiovasc. Dev. Dis. 6 (2) (2019) 19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [21].Aspree Investigator Group, Study design of ASPirin in Reducing Events in the Elderly (ASPREE): a randomized, controlled trial, Contemp. Clin. Trials 36 (2) (2013) 555–564. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [22].McNeil JJ, Woods RL, Nelson MR, et al. , Baseline characteristics of participants in the ASPREE (ASPirin in Reducing Events in the Elderly) study, J. Gerontol. A Biol. Sci. Med. Sci. 72 (11) (2017) 1586–1593. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [23].Ware JE Jr., Kosinski M, Turner-Bowker DM, Gandek B, User’s Manual for the SF-12v2® Health Survey (with a Supplement Documenting SF-12® Health Survey), QualityMetric Incorporated, Lincoln, RI, 2002. [Google Scholar]
- [24].Ware JE Jr., Kosinski M, Keller SD, A 12-item short-form health survey: construction of scales and preliminary tests of reliability and validity, Med. Care 34 (3) (1996) 220–233. [DOI] [PubMed] [Google Scholar]
- [25].McNeil JJ, Wolfe R, Woods RL, et al. , Effect of aspirin on cardiovascular events and bleeding in the healthy elderly, N. Engl. J. Med. 379 (16) (2018) 1509. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [26].McNeil J, Nelson M, Woods R, et al. , Effect of aspirin on all-cause mortality in the healthy elderly, N. Engl. J. Med. 379 (16) (2018) 1519–1528. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [27].Antman E, Bassand J-P, Klein W, et al. , Myocardial infarction redefined—a consensus document of the Joint European Society of Cardiology/American College of Cardiology committee for the redefinition of myocardial infarction, J. Am. Coll. Cardiol. 36 (3) (2000) 959–969. [DOI] [PubMed] [Google Scholar]
- [28].World Health Organization, Stroke–1989. Recommendations on stroke prevention, diagnosis, and therapy. Report of the WHO task force on stroke and other cerebrovascular disorders, Stroke 20 (10) (1989) 1407–1431. [DOI] [PubMed] [Google Scholar]
- [29].Stocks N, González-Chica D, Woods R, et al. , Quality of Life for 19,114 participants in the ASPREE (ASPirin in Reducing Events in the Elderly) study and their association with sociodemographic and modifiable lifestyle risk factors, Qual. Life Res. 28 (4) (2019) 935–946. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [30].Australian Government National Health and Medical Research Council, Draft Australian Guidelines to Reduce Health Risks from Drinking Alcohol, National Health and Medical Research Council, Canberra, 2019. Available from: https://www.nhmrc.gov.au/health-advice/alcohol. [Google Scholar]
- [31].McNeil JJ, Woods RL, Ward SA, et al. , Cohort profile: the ASPREE longitudinal study of older persons (ALSOP), Int. J. Epidemiol. 48 (4) (2019), 1048–9h. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [32].Radloff LS, The CES-D scale: a self-report depression scale for research in the general population, Appl. Psychol. Meas. 1 (3) (1977) 385–401. [Google Scholar]
- [33].Vlassoff C, Gender differences in determinants and consequences of health and illness, J. Health Popul. Nutr. 25 (1) (2007) 47. [PMC free article] [PubMed] [Google Scholar]
- [34].Hawthorne G, Osborne R, Taylor A, Sansoni J, The SF36 version 2: critical analyses of population weights, scoring algorithms and population norms, Qual. Life Res. 16 (4) (2007) 661–673. [DOI] [PubMed] [Google Scholar]
- [35].Mancuso S, SF-12 v2 Scoring Using Australian Population Weights (Sam Mancuso in Scoring Algorithm, Self-Report Measures) [cited 2020 October 5]. Available from: https://sammancuso.com/2017/11/08/sf-12-v2-scoring-using-australia-population-weights/, 2017.
- [36].Brazier EJ, Roberts EJ, The estimation of a preference-based measure of health from the SF-12, Med. Care 42 (9) (2004) 851–859. [DOI] [PubMed] [Google Scholar]
- [37].Lu W-C, Tzeng N-S, Kao Y-C, et al. , Correlation between health-related quality of life in the physical domain and heart rate variability in asymptomatic adults, Health Qual. Life Outcomes 14 (1) (2016) 149. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [38].Hillebrand S, Gast KB, de Mutsert R, et al. , Heart rate variability and first cardiovascular event in populations without known cardiovascular disease: meta-analysis and dose-response meta-regression, Europace 15 (5) (2013) 742–749. [DOI] [PubMed] [Google Scholar]
- [39].Crosby RD, Kolotkin RL, Williams GR, Defining clinically meaningful change in health-related quality of life, J. Clin. Epidemiol. 56 (5) (2003) 395–407. [DOI] [PubMed] [Google Scholar]
- [40].Mottillo S, Filion KB, Genest J, et al. , The metabolic syndrome and cardiovascular risk a systematic review and meta-analysis, J. Am. Coll. Cardiol. 56 (14) (2010) 1113–1132. [DOI] [PubMed] [Google Scholar]
- [41].van Holten TC, Waanders LF, de Groot PG, et al. , Circulating biomarkers for predicting cardiovascular disease risk: a systematic review and comprehensive overview of meta-analyses, PLoS One 8 (4) (2013), e62080. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [42].Myers J, Exercise and cardiovascular health, Circulation 107 (1) (2003) e2–e5. [DOI] [PubMed] [Google Scholar]
- [43].Anokye NK, Trueman P, Green C, Pavey TG, Taylor RS, Physical activity and health related quality of life, BMC Public Health 12 (1) (2012) 624. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [44].Mohr JP, Albers GW, Amarenco P, et al. , Etiology of stroke, Stroke 28 (7) (1997) 1501–1506. [DOI] [PubMed] [Google Scholar]
- [45].Kemp CD, Conte JV, The pathophysiology of heart failure, Cardiovasc. Pathol. 21 (5) (2012) 365–371. [DOI] [PubMed] [Google Scholar]
- [46].Thygesen SK, Alpert DJ, White DH, Universal definition of myocardial infarction, Circulation 116 (22) (2007) 2634–2653. [DOI] [PubMed] [Google Scholar]
- [47].Hare DL, Toukhsati SR, Johansson P, Jaarsma T, Depression and cardiovascular disease: a clinical review, Eur. Heart J. 35 (21) (2014) 1365–1372. [DOI] [PubMed] [Google Scholar]
- [48].Gupta R, Stocks NP, Broadbent J, Cardiovascular risk assessment in Australian general practice, Aust. Fam. Physician 38 (5) (2009) 364–368. [PubMed] [Google Scholar]
- [49].Fleishman AJ, Cohen WJ, Manning GW, Kosinski GM, Using the SF-12 health status measure to improve predictions of medical expenditures, Med. Care 44 (5 suppl) (2006). I-54–I-63. [DOI] [PubMed] [Google Scholar]
- [50].Williams K, Sansoni J, Morris D, Grootemaat P, Thompson C, Patient-Reported Outcome Measures: Literature Review, Australian Commission on Safety and Quality in Health Care (ACSQHC), Sydney, 2016. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
All individual participant data (re-identifiable) that underlie the results reported in this manuscript, are available upon request to qualified researchers without limit of time, subject to approval of the analyses by the Principal Investigators and a standard data sharing agreement. Details regarding requests to access the data will be available through the web site (www.ASPREE.org). The data will then be made available through a web-based data portal safe haven at Monash University, Australia.