Information for researchers
Data Analysis and Research Network
Data Summary 2010
Data Book - April 2010
Data Book - August 2008
Data dictionary 2008
To help researchers using The 45 and Up Study baseline questionnaire, on our website have provided some useful SAS code. This syntax has been developed by The 45 and Up Study and two other researchers, and includes code to:
a) score widely used scales such as the K10 and MOS-PF
b) clean and calculate time and sessions spent in physical activity, and
c) derive single categorical variables from multi-item questions such as level of private health insurance and work status.
These have been checked and will work on the April 2010 dataset of the full cohort. Some of this code will not work directly on previous versions of the baseline dataset (noted below), users can contact the Data Users Group for a version that will work with older versions.
We’d like to thank the researchers who provided us with their useful syntax (Ian Robinson and Sanja Lujic), and we’d like to encourage you to share any useful tools you have developed for 45 and Up Study data (full acknowledgement given of course).
Run the formats.sas before any other program, the rest are designed to be run in a data step (the easiest way to do this is with the ‘%inc(\path to file\file.sas’ option). Please let us know if you have any questions about this SAS code.
The Kessler Psychological Distress Scale (K10) is a scale based on 10 items used to measure non-specific psychological distress ((Kessler et al 2003). It is widely used in Australia in state (e.g. NSW Population Health Survey) and national (e.g. ABS Health Surveys) surveys and is in the baseline questionnaire in The 45 and Up Study (Question 57). A common issue with multi-item scales like the K10 and MOS-PF (see below) is that there are relatively high rates of missing items which makes calculation of the score problematic. The syntax provided by Ian Robinson from The University of Newcastle performs logical imputations of missing values where there is a valid value for a similar but more severe item. For example, when the value for “..how often did you feel: depressed” is missing, then the value for “..how often did you feel: so depressed that nothing could cheer you up” is imputed to the less severe item. The average of all non-missing items is imputed for up to one missing item, and then the final score is calculated. The code provided can be used directly on any version of the baseline questionnaire data.
The Medical Outcomes Study – Physical Functioning scale (MOS-PF) is a widely used scale to measure physical functioning (e.g. on the SF-36). The 10 item version used in the 45 and Up study is considered to be a reliable measure of low levels of physical function but less useful for comparisons of higher levels of physical function (Hays et al 2007). The SAS syntax provided will calculate the MOS-PF as a score from 0 to100, with lower values representing lower levels of physical function. As suggested for SF-36 scales, a score is calculated where there are up to 5 missing items. This code will work on any version of the baseline questionnaire data.
The physical activity (PA) items on the 45 and Up questionnaire were modified between different versions of the questionnaire with respect to order of items and wording. This has resulted in systematic differences in the data between versions, particularly the rate of valid 0’s and missing values. The SAS syntax provided performs cleaning and some imputation of PA variables to reconcile the different versions of the questionnaire. The total number of PA time and sessions is then calculated (as well as specific PA type time and sessions) as standard and metabolically adjusted. This syntax was used in Banks et al (2010) to clean and calculate PA time and sessions. This code will work on any version of the baseline questionnaire data.
Participants were asked on the baseline questionnaire about which health insurance (if any) they held. A short piece of SAS syntax is provided to use the five private health insurance variables to create a single variable representing the more comprehensive level of cover held by the participant. This syntax was provided by Sanja Lujic from The University of Western Sydney and used in Banks et al (2009) to derive the main outcome variable of private health insurance status.
There are 13 items in Q47&Q49 of the baseline questionnaire concerning work status. The code provided uses the number of hours in paid work (Q47) and then the rest of the items in Q47 to derive a work status variable denoting full, part-time, or no current paid work. This has been used in Banks et al (2010) to derive a work status variable used in analysis of obesity and screen-time.
References
Banks et al (2009) Health, ageing and private health insurance: baseline results from the 45 and Up Study cohort. Australia and New Zealand Health Policy 6:16.
Banks et al (2010) Screen-time, obesity, ageing, and disability: findings from 91,266 participants in The 45 and Up study. Public Health Nutrition in press.
Hays et al (2007) Item response theory analyses of physical functioning items in the medical outcomes study. Med. Care 45(5 S1):S32-38.
Kessler, R.C, Barker, P.R, Colpe, L.J, et al.(2003). Screening for serious mental illness in the general population, Arch Gen Psychiatry, Vol 60.
The 45 and Up Study Research Network is an overarching group of researchers, policy makers and others with an interest in the 45 and Up Study. The Research Network replaces the system of theme committees that were essential in the early stages of developing the Study methods and materials.
The main purpose of the Research Network is to facilitate high quality practice- and policy-relevant research within the 45 and Up Study. Members of the 45 and Up Study Research Network receive:
- regular email updates about the 45 and Up Study through our E-News
- invitations to the annual Collaborators' Meeting, workshops and other Study events
- information on the research being carried out using data from the Study and invitations to collaborate in these where relevant
- technical information and advice on the Study for the preparation of funding applications or when undertaking analyses of the Study data
If you have any questions or comments, you would like to join the Research Network or you would like to update your contact details or membership status please send an online enquiry or call the Study Infoline on 1300 45 11 45 (Mon-Fri, 9am-5pm).