How did the study come about?
In common with virtually all industrialized countries and many less developed nations,
Australia is facing rapid population ageing. Historical patterns of fertility and
migration, along with changes in life expectancy, mean that the over 65 age group
is likely to increase by around 50% in the next 15–20 years.
1
The further increase in the proportion of people in the very old age groups will result
in the ‘ageing of the aged’. The challenges presented by the ageing of the population
are far reaching. Discussions have tended to focus on its likely health and economic
consequences; however, few aspects of society will remain unaffected by the issue.
There is an urgent need for reliable evidence to inform policy to support healthy
ageing. The concept of healthy ageing encompasses traditional ideas relating to freedom
from disease, as well as broader considerations including independence, quality of
life, management of disability, participation in society and the workforce and productivity.
A wide range of factors are likely to affect health in later life, including socioeconomic,
environmental and cultural variables, cigarette smoking, alcohol consumption, diet,
physical activity, reproductive and hormonal factors, infections, availability of
healthcare and use of pharmaceutical agents, as well as individuals’ susceptibility
to disease. A comprehensive investigation of the determinants of healthy ageing must
incorporate assessment of disease risk, quality of life and other indices, in relation
to a very wide range of possible exposures, and with consideration of how these exposures
might interact with one another. Research needs to be of a sufficient scale to provide
specific information on the major diseases and health problems experienced in later
life. This is because reliable assessments of risk factor–disease relationships require
a substantial degree of pathological homogeneity of outcome and appropriate consideration
of confounding. At the same time, research needs to be able to assess the broad risks
and benefits of particular exposures, to allow meaningful conclusions to be reached
about suitable public health interventions. Finally, it needs to be large and long
term enough to track the impact of health interventions and policies at the population
level.
Australia has some unique characteristics that will impact on healthy ageing and provide
particular challenges in delivering health care. For example, it has: a relatively
heterogenous population with a large migrant community; an indigenous population with
an average life expectancy 17 years less than for non-indigenous Australians; some
remote and sparsely populated regions and a mixed health care system with responsibility
shared between the national and state governments and delivery in both the public
and private sectors. Excellent population-level databases relating to use of health
services and medications, and registers of cancers and deaths, are available for statistical
linkage with research data sets. There is therefore a need for research that addresses
issues specific to the Australian population and makes use of the unique features
of the Australian setting, giving the opportunity to provide insights of international
relevance.
The 45 and Up Study was conceived as a long-term collaborative resource to investigate
healthy ageing, in response to the gaps in existing knowledge and the needs of researchers.
Initial discussions among interested researchers resulted in the formation of a Scientific
Steering Group in 2003 to oversee the development of the Study. The Study is auspiced
by the Sax Institute, which also provided funding for its development. The Sax Institute
is an independent organization with core funding from the state government of New
South Wales, Australia's most populous state. Its mission is to improve health through
facilitating high-quality research and increasing the impact of this research on health
policy and services; it has membership from public health and health service research
centres and universities across New South Wales. Workshops, meetings and consultations
with individual researchers were held to identify research priorities for the Study
and researchers were invited to become 45 and Up Study Collaborators. There were 120
collaborators at the time of writing (see www.45andUp.org.au for details). The protocol
for the Study was completed in late 2003, the Scientific Director was appointed in
2003, the Study materials, methods and governance framework were developed in 2004
and 2005 and recruitment of participants started early in 2006.
Following international peer review of the Study protocol, funding for the Study infrastructure
was provided by Foundation Study Partners, The Cancer Council New South Wales, the
New South Wales Division of the National Heart Foundation of Australia and the New
South Wales Department of Health. These were later joined by beyondblue: the national
depression initiative, the New South Wales Department of Ageing, Disability and Home
Care, with additional support from the Macquarie Bank Foundation, the Baxter Charitable
Foundation, the Alma Hazel Eddy Trust and Freehills solicitors. Competitive funding
for a range of research projects has been secured and researchers using Study data
for their projects support the Study through contributions to cost recovery and Study
sustainability.
What does it cover?
The 45 and Up Study is a large-scale Australian cohort study of individuals aged 45
and over aiming to provide researchers with timely and reliable information on a wide
range of exposures and outcomes of public health importance for the ageing population.
At the same time as addressing specific research questions, the Study is also designed
as a framework for a range of future research activities. Priority areas for the Study
are:
Social and economic determinants of healthy ageing, including income, education, ethnicity,
work and retirement, social capital and rurality.
Health effects of obesity, overweight and physical activity.
Impact of environmental factors on healthy ageing.
Risk factors for, and the detection and management of, cancer, cardiovascular disease
and mental health problems, including depression.
Use of health services in relation to ageing, including the determinants of use of
residential aged care.
Health in people aged 80 years and over (the ‘old old’).
Who is in the sample?
The 45 and Up Study is in the process of recruiting 250 000 men and women aged 45
and over from the general population of the state of New South Wales, Australia (Figure
1). Individuals aged 45 and over and resident in New South Wales are randomly sampled
from the Medicare Australia enrolment database. This is the database through which
national healthcare is administered and includes all citizens and permanent residents
of Australia, as well as some temporary residents and refugees. Eligible individuals
are mailed an invitation to take part, an information leaflet, the study questionnaire
and consent form and a reply paid envelope (available at www.45andUp.org.au). The
data included in the baseline questionnaire are listed in Table 1. Participants join
the study by completing the questionnaire and consent form and mailing them to the
Study coordinating centre. Completed questionnaires are then scanned electronically
and stored as images and data are double-entered.
Figure 1
Location of the 45 and Up Study: New South Wales, Australia
Table 1
Data collected in the 45 and Up Study baseline questionnaire
Demographic and social characteristics
Personal health behaviours
General health-related data
Date of birth
Smoking
Disease and surgical history
Education
Alcohol
Family history of illness
Income
Physical activity
Medication
Marital status
Fruit and vegetable consumption
Functional capacity
Country of birth and ancestry
Other dietary information
Psychological distress
Retirement and work
Sleep habits
Cancer screening history
Social connectedness
Falls
Oral health
Skin pigmentation and response to sunlight
Reproductive history
Incontinence
Prostate symptoms and sexual functioning (in men)
The study over-samples, by a factor of two, individuals aged 80 years and over and
people resident in rural areas; all residents of remote areas are sampled. People
may also volunteer to join the study by calling the study helpline and requesting
an invitation pack. When complete, the 45 and Up Study will include ∼10.9% of the
general population in the target age range.
Recruitment into the 45 and Up Study commenced in February 2006; since then 36 645
men and women have joined the study. By the end of 2007, 100 000 participants will
have joined the study, with the full 250 000 recruited by the end of 2009. The current
overall response rate to the mailed invitations to join the study is estimated to
be 17.9% (95% CI 17.8–18.1), however, the exact response rate is difficult to specify
as some people may not have received the invitation if their address details were
incorrect in the Medicare Australia database. In addition to the 36 162 (98.7%) people
joining the Study by invitation, 483 (1.3%) people have joined by contacting the helpline
without receiving an invitation to take part.
The demographic characteristics of the first 36 645 members of the cohort are shown
in Table 2. Overall, 52% of participants are females. Participants range in age from
45 to 101 years, with a mean age of 64.1 years (SD 11.7) for men and 61.3 years (SD
10.8) for women. They show considerable diversity in terms of location of residence
(44% of men and 40% of women are resident in major cities, 54% of men and 57% of women
are resident in regional areas and 2% of men and 3% of women are resident in remote
areas), ethnicity (21% of all participants were born outside of Australia; 9% speak
a language other than English in the home; 0.9% are of Aboriginal and/or Torres Strait
Islander origin), level of education (14% of participants report having no educational
qualifications; 22% report having a university degree) and income (32% of participants
report a gross household income less than 30 000 Australian dollars per year; this
is ≤60% of the median gross household income in New South Wales
2
).
Table 2
Socio-demographic characteristics of the first 36 645 participants joining the 45
and Up Study
Men
Women
Total
N = 17 443 (%)
N = 19 202 (%)
N = 36 645 (%)
Age (years)
45–49
12.4
16.0
14.3
50–59
29.9
36.2
33.2
60–69
26.2
25.8
26.0
70–79
18.5
14.9
16.6
80–89
11.9
6.5
9.1
≥90
1.0
0.6
0.8
Missing/invalid
0.0
0.0
0.0
Aboriginal or Torres Strait Islander orgin
Neither Aboriginal nor Torres Strait Islander
97.3
97.1
97.2
Aboriginal only
0.7
0.8
0.7
Torres Strait Islander only
0.1
0.1
0.1
Aboriginal and Torres Strait Islander
0.1
0.0
0.1
Missing/invalid
1.8
2.1
1.9
Language other than English spoken at home
Yes
9.3
8.2
8.7
No
90.7
91.8
91.3
Missing/invalid
0.0
0.0
0.0
Marital status
a
Single
7.3
6.4
6.8
Married
75.0
65.4
69.9
In a de facto relationship
5.4
5.1
5.3
Widowed
5.7
13.1
9.6
Divorced
6.2
9.6
8.0
Separated
3.1
3.2
3.1
Highest educational qualification
No school certificate
12.9
14.5
13.7
School/intermediate certificate
16.2
28.8
22.8
Higher school certificate
10.0
9.8
9.9
Trade/apprenticeship
18.8
4.3
11.2
Certificate/diploma
18.9
21.9
20.5
University degree
23.2
20.6
21.9
Missing/invalid
0.0
0.0
0.0
Household annual pre-tax income
b
<$10 000
6.4
7.6
7.1
$10 000–$19 999
14.0
14.2
14.1
$20 000–$29 999
11.3
9.6
10.4
$30 000–$39 999
8.5
7.5
8.0
$40 000–$49 999
7.6
6.5
7.0
$50 000–$69 999
11.2
9.4
10.2
≥$70 000
23.1
16.7
19.7
Prefer not to answer
13.7
22.1
18.1
Missing/invalid
4.3
6.4
5.4
aMultiple responses permitted.
bIn Australian dollars.
Based on self-reported height and weight, 4% of individuals have a body-mass index
(BMI) of <20 kgm−2, 33% have a BMI from 20 to 24.9 kgm−2 and 56% have a BMI of 25
kgm−2 or over, meaning they are overweight or obese (Table 3). Overall, 8% of participants
report they are current smokers and 65% report drinking alcohol at least weekly; 35%
of participants report engaging in at least 30 minutes of vigorous physical activity
at least weekly and 37% report no weekly vigorous physical activity.
Table 3
Lifestyle and social factors in the first 36 645 participants joining the 45 and Up
Study
Men
Women
Total
N = 17 443 (%)
N = 19 202 (%)
N = 36 645 (%)
Tobacco smoking
Current smoker
8.4
7.5
7.9
Former smoker
44.9
27.6
35.9
Never smoker
46.7
64.9
56.2
Missing/invalid
0.0
0.0
0.0
Alcoholic drinks per week
0
23.3
40.2
32.2
1–4
18.7
22.7
20.8
5–7
14.5
14.5
14.5
8–14
18.7
14.0
16.3
≥15
22.9
5.9
14.0
Missing/invalid
1.9
2.7
2.3
Moderate or vigorous physical activity (minutes per week)
None
11.4
10.9
11.2
1–30
6.6
6.2
6.4
30–60
6.6
5.7
6.1
60–90
9.1
9.1
9.1
90–120
5.8
5.7
5.7
≥120
55.7
57.0
56.4
Missing/invalid
4.8
5.3
5.1
Body mass index
<20
2.3
5.5
4.0
20–22.4
8.3
14.4
11.5
22.5–24.9
20.1
21.3
20.7
25.0–29.9
43.6
30.4
36.7
≥30
20.2
21.8
21.0
Missing/invalid
5.4
6.6
6.0
Current work status
a
Full time paid work
26.7
18.7
22.5
Part time paid work
6.1
18.6
12.7
Doing unpaid work
4.3
6.9
5.7
Self employed
16.6
9.2
12.7
Partially retired
7.0
4.5
5.7
Compeletly retired
41.7
37.5
39.5
Studying
0.9
2.0
1.5
Looking after home/family
3.1
18.3
11.1
Disabled/sick
4.9
3.6
4.2
Unemployed
2.3
2.1
2.2
Other
1.0
1.9
1.5
Times per week with family or friends
None
13.0
8.0
10.4
1–3
48.5
46.1
47.3
4–6
22.9
28.9
26.0
≥7
11.9
13.7
12.8
Missing/invalid
3.8
3.2
3.5
aMultiple responses permitted.
A previous diagnosis of non-melanoma skin cancer was reported by 26% of participants,
with a history of melanoma, heart disease and diabetes being reported by 6, 12 and
9% of participants, respectively (Table 4). Prostate cancer was reported by 6% of
men and breast cancer by 5% of women. In terms of functional status, 5% of the cohort
report needing help with day to day activities and 4% report that they are caring
full time for a sick or disabled person.
Table 4
Medical and surgical history and functional capacity of the first 36 645 participants
joining the 45 and Up Study
Men
Women
Total
N = 17 443 (%)
N = 19 202 (%)
N = 36 645 (%)
History of
a
Skin cancer(excluding melanoma)
29.5
23.2
26.2
Melanoma
6.9
4.2
5.5
Prostate cancer
6.3
Breast cancer
0.0
5.4
2.8
Other cancer
6.6
6.6
6.6
Heart disease
16.4
8.2
12.1
High blood pressure
b
35.9
33.0
34.4
Stroke
4.7
2.7
3.7
Diabetes
10.7
7.1
8.8
Thrombosis
3.9
5.8
4.9
Enlarged prostate
16.3
Asthma or hayfever
14.1
20.9
17.7
Parkinson's disease
1.1
0.7
0.8
Past operations
Removal of skin cancer
31.4
23.3
27.1
Vasectomy
22.7
Part of prostate removed
6.6
Whole of prostate removed
3.7
Hysterectomy
29.2
Bilateral oophorectomy
10.4
Tubal ligation
28.0
Repair of prolapsed bladder, womb or bowel
12.1
Knee replacement
4.4
3.4
3.9
Hip replacement
3.8
2.9
3.3
Gallbladder removed
6.3
14.1
10.4
Coronary artery bypass surgery
c
9.7
2.6
6.0
Current need for assistance
Need help with daily tasks
5.3
5.3
5.3
Do not need help with daily tasks
94.7
94.7
94.7
Missing/invalid
0.0
0.0
0.0
aself reported conditions.
bexcluding pregnancy related hypertension.
cincludes grafts, balloons and stents.
Further details of the responses to the 45 and Up Study baseline questionnaire are
available in the study preliminary data book, at www.saxinstitute.org.au.
What has been measured?
The data that will be collected for the 45 and Up Study include a baseline questionnaire,
5-yearly repeat questionnaires, linkage to routinely collected health data, biological
samples and more intensive measures from sub-studies conducted within the cohort.
The baseline questionnaire for the study broadly includes: measures of health status
and past medical and surgical history; known risk factors for major causes of morbidity
and mortality; likely confounding factors; potential mediators of risk and some novel
factors (Table 1). A pilot study conducted in July 2005 established the feasibility
of the recruitment methods and allowed the refinement of a number of questions. Repeat
questionnaires to all cohort members to update exposure data and health status are
planned at 5-yearly intervals.
Questionnaire data from study participants are linked with routinely collected data
from a range of population databases and registers. These include information on use
of prescription medication, use of general practice services and updated address details,
through Medicare Australia, and routine linkage to deaths, cancer registrations and
hospitalizations, through the New South Wales Centre for Health Record Linkage (see
www.cherel.org.au). Linkage is both retrospective and prospective, with the time period
covered dependent on the properties of the specific data set. A range of additional
data sets (e.g. use of community-based and residential aged care services, screening
records and transplant details) will become available for specific projects as the
study progresses.
A number of more detailed sub-studies are already underway within the framework of
the 45 and Up Study cohort, enhancing the information available on selected participants.
For example, the first 100 000 participants will receive an additional questionnaire
to provide more information on social, environmental and economic factors and health
status.
It is planned that biological samples, including blood for genotyping and examination
of a range of analytes, will be collected on as many participants as possible as the
Study progresses and funding becomes available. Blood samples will be centrifuged
and aliquotted, and stored long term in ultra low temperature freezers, with procedures
in place for retrieval of samples for specific projects. The protocol for collection
of samples is currently being developed and collection is planned to commence in 2008.
Following initial analyses of cross-sectional data, the main means of investigating
relationships between exposure and outcome will be through cohort analyses and nested
case–control studies. For conditions yielding at least 3500 incident cases or deaths
during the first 10 years of follow-up (e.g. diabetes mellitus, myocardial infarction,
stroke, colorectal cancer, prostate cancer, breast cancer and hip fracture), the study
will be able to detect a minimum relative risk of 1.3 for exposures affecting 10–90%
of controls and 1.4 for exposures affecting 5–95% of controls, with 95% power, 1%
significance and four controls per case. With the same power, significance and ratio
of cases to controls, for outcomes with 1000 or more events (e.g. ischaemic heart
and cerebrovascular disease deaths and incident Parkinson's disease, non-Hodgkin's
lymphoma and lung cancer) the study should be able to detect minimum relative risks
of 1.4–1.5 for exposures affecting 20–80% and 1.6 for exposures affecting 10–90% of
the population.
How is the Study optimizing use of the data?
The 45 and Up Study has sought to establish an approach that will make optimal use
of the data. Data from the study are accessible to local researchers for approved
research projects.
First, we have sought to encourage local researchers to use data from the Study by
involving a large number of leading researchers, from across a range of institutions,
in its management and design from the outset. A policy framework has been established
for the Study that addresses issues including use of and access to the data, intellectual
property, protecting the privacy and confidentiality of participants and charges for
data use. We have convened a Community and Ethical Oversight Committee, which includes
community leaders among its members. At the time of writing, 18 applications had been
made to the Scientific Advisory Committee for access to Study data for research projects
and funding applications. Six projects, spanning cancer risk factors, oral health,
health economics, functional capacity and psychological distress, are underway.
Second, to encourage use of the Study for research that will directly inform the planning
and provision of services, we have convened a policy roundtable. This includes representation
from major government and non-government agencies and private sector organizations
that have responsibility for providing health and aged care services. The roundtable
serves to generate ideas and opportunities for new policy-relevant research using
the Study. It has been funded by the MBF Foundation (a non-profit charitable institution).
What are the main strengths and weaknesses?
When recruitment is complete, the 45 and Up Study will be the largest population-based
cohort study in Australia and, to our knowledge, in the Southern Hemisphere. It is
unique among large-scale cohorts internationally, in terms of the scope of its ongoing
linkage, even for those participants who are otherwise lost to follow-up, with a large
number of routinely collected databases and registers that provide virtually complete
capture of use of a wide range of health services, as well as cancer registration
and deaths. The combination of large numbers of participants with individual prospective
information on exposures and additional linked exposure and follow-up information
will provide a valuable resource for the investigation of many different causes of
morbidity, mortality and of patterns of use of health and aged care services. The
addition of biological samples to the study over time will further increase its usefulness.
The study does not at present have the levels of funding required to collect biological
samples on participants, so this is being introduced gradually, as funding is secured.
A limitation of the study is the lack of large-scale clinical data on measures such
as blood pressure, spirometry, anthropometry and cognition; it is intended that many
of these will be gathered as part of sub-studies within the cohort.
The study population is relatively heterogeneous, with a good spread of responses
across most variables. The oversampling of people 80 years and over (with no upper
age limit to participation) and from rural areas will enable a particular focus on
these groups. The Study will provide a good basis for the investigation of aetiological
research questions, based on internal comparisons within the cohort. Although derived
from the general population, the relatively low response rate means that the cohort
is unlikely to be directly representative of the general population. Formal comparisons
of the cohort with the general population will be conducted in due course. The questionnaire
is currently only available in English, limiting the participation of people with
insufficient literacy in English. Having said this, it is likely to be one of the
most inclusive large-scale cohort studies conducted to date and the results from cohort
studies based on internal comparisons remain generalizable even when the cohort is
from a selected group.
3–5
Additional strengths of the study include: the accessibility of data to collaborating
researchers; its governance structure and community accountability and its partnerships
with policy agencies.
Can I get hold of the data? Where can I find out more?
The Study is accessible to local and collaborating researchers for high quality policy-relevant
research that is in the public interest. Details of the data access policy and procedures
are available at www.saxinstitute.org.au.