Published: Monday 23 November 2020

A new paper - a collaboration between Neuroscience Research Australia, University of New South Wales and Stanford University - explores what measures are most useful and effective for research into the genetics of wellbeing.  

Dr Justine Gatt and her team at NeuRA explain the aims and key findings of their latest study involving identical and fraternal twins. Read the full paper here   

What did you research and why?

Wellbeing is an important aspect of mental health which is not simply the absence of symptoms of mental illness. Wellbeing is moderately heritable, and independent studies employ a variety of instruments to measure and quantify different components of wellbeing, including either subjective wellbeing or psychological wellbeing. These instruments can include simple single-item questions (e.g. how satisfied are you with your life?) to more elaborate multi-item questionnaires (e.g. COMPAS-W: see diagram of this scale below). It is not yet clear which measurement is more useful or effective for research in genetics of wellbeing.

We compared different instruments for measuring wellbeing in regards to their heritability and stability over time. We also investigated how wellbeing polygenic scores (which are a summary measure of genetic burden across many thousands of genetic sequence variants in each person’s genome) can predict the wellbeing scores measured using the different instruments. We wanted to contrast our wellbeing scale, the COMPAS-W, which measures both subjective and psychological wellbeing, to other commonly used instruments. We also wanted to test if there is a difference between single item questions and multi-item questionnaires in their heritability and predictability using polygenic scores of wellbeing.

What are the previous discoveries that have led up to your current work?

There have been worldwide efforts to identify genetic factors which contribute to wellbeing, which have made great inroads into our understanding of the polygenic nature of this trait. It has been shown that meta-analysing the results of genome wide association studies on closely related phenotypes (e.g. positive affect, depressive symptoms, life satisfaction) significantly increases the power of gene discovery. Here we employed an alternative and complementary approach to test whether a composite phenotype of wellbeing comprising a more wholistic measure of both subjective and psychological wellbeing can further augment existing research approaches.

Who participated in your research?

The sample is drawn from the TWIN‐E study, an ongoing longitudinal prospective study of 1660 individuals sourced from Twins Research Australia (TRA). The sample comprised of 1660 individuals aged 18–62 years (mean (M) = 39.64, standard deviation (SD) = 12.73; 680 males), including 484 monozygotic (MZ) twin‐pairs, 267 dizygotic (DZ) twin‐pairs and 158 singletons from twin‐pairs. The participants provided responses to a series of online questionnaires, neurocognitive tasks and saliva samples for genetic material. A subset of the twins also came into our lab to get EEG and MRI brain imaging tests. The study originally started in 2009, with a 12-month follow-up already conducted. The team is currently in the midst of a 10-year follow-up of all twins, supported by a NHMRC Project Grant led by Dr Justine Gatt (NeuRA, UNSW).

What did your research involve?

We estimated the heritability of each instrument using univariate twin modelling (comparing the similarity of MZ twins and DZ twins) and constructed a polygenic score of wellbeing for each participant using their genetic data and the information we got from the most recent GWAS on wellbeing. Then we compared these estimates for different instruments.

What were your key findings?

We found that multi-item questionnaires tend to be more heritable and stable over time compared to single-item questions that measure wellbeing. Our instrument, COMPAS-W had the highest heritability and was one of the most stable estimates over the 12-month period. We found that the polygenic scores of wellbeing can significantly predict the scores of most instruments, but this was variable for different instruments and varied over different time points with a pretty stable estimate for COMPAS-W.

What impact or difference will these findings make?

Our findings show that using a reliable and multi-item measure of wellbeing which encompasses a broad spectrum of wellbeing, including both subjective and psychological components, can enhance the power of studies in wellbeing research. The result of this study can support future research on the genetics of wellbeing in selecting the appropriate instruments for measurement.

Will you be undertaking further research in this field?

In terms of our genetic work, we are currently validating our approach in a larger database of participants for whom both genetic data and questionnaire data are available. We hope then to expand our work to examine relationships between genetic signatures and brain structural changes which relate to wellbeing and resilience.

More broadly, our research team are currently re-testing the same twin cohort for a 10-year follow-up to identify risk versus resilience trajectories over time, and the corresponding neurocognitive and genetic markers that differentiate those profiles.


The COMPAS-W Wellbeing Scale (Gatt et al., 2014) is a 26-item scale that measures both subjective
and psychological wellbeing.



Twins Research Australia

Address: 3/207 Bouverie St
Carlton, Vic 3010


ABN: 84 002 705 224

Twins Research Australia has received continuous funding from the National Health and Medical Research Council (NHMRC) since 1981, most recently through a Centre of Research Excellence Grant (2015-2022). TRA is administered by the University of Melbourne.

Privacy Policy


Follow or ‘like’ us