Australia is now the fattest nation on earth - with four million Australians classified as obese and another 5.4 million as overweight and experts say exercising is one way to reduce this frightening trend.

New research by Victoria University's Dr Erin Pearson shows that when it comes to changing people's exercise behaviour having the right messenger is as important as having the right message.

Dr Pearson said: "We hear all the time the message that exercise is important and yet many of us don't take any notice. This study has given us a clue about why that might be. What we have found is that the person delivering the message needs to be someone the audience identifies with and respects in order to bring about a desired change in behaviour."

The study investigated intervention strategies designed to get sedentary women exercising and found that interventions based on women's social identities were far more likely to be effective.

"Humans are social creatures and we form social identities in which we group ourselves with others in a social categorisation based either on personality-based self characteristics we share, such as 'politically minded' or 'creative' or role-based characteristics, such as 'mother' or 'professional'," said Dr Pearson.

From a pool of 214 female volunteers, Dr Pearson identified two groups whose social identities were based on the personality based self-characteristics - the 'strong' and 'independent' group and the 'spiritual' and 'caring' group.

A third control group was randomly selected. Women in all groups were aged 25 to 45 and spent less than 150 minutes per week exercising.

Before introducing the interventions, Dr Pearson recorded the women's rate of exercise. The average time spent exercising per week was 88 minutes for the 'strong independent' group, 109 minutes for the 'spiritual caring' group, and 128 minutes for the randomly selected group.

For the 'social identity' groups the intervention consisted of watching video footage of similar women, either 'strong independent' or spiritual caring', talking about the importance of regular physical exercise. The women were asked to focus on how exercise was an integral part of being either a 'strong independent' woman or 'spiritual caring' woman.

Meanwhile the control group was played footage about the importance of physical activity presented by health professionals and were asked to focus on the importance of exercise for them as individuals. Neither the message nor the messengers were chosen to appeal to these women's social identities.

After two weeks, Dr Pearson again evaluated the women's rate of exercise and all of the groups showed a significant increase. The 'strong independent' women were now doing 239 minutes, the 'spiritual caring' women 166 minutes, and the randomly selected group averaged 220 minutes per week.

However, at the final evaluation after three months only the women in the two social identity groups were found to have maintained the increased levels of exercise.

On average the 'strong independent' women were doing 264 minutes per week and the 'spiritual caring' women recorded an average of 170 minutes. Both the social identity groups recorded a slight increase in exercise levels since the two-week evaluation, while the level for the control group had dropped back to almost pre-intervention levels at 143 minutes.

"What appeared to have happened was that an 'internal change' had taken place in the women whose intervention was based on social identity," said Dr Pearson.

"By creating a mental association between a valued aspect of their social identity and the goal of physical activity it was possible to get the women to maintain an increase in their rate of exercise activity."

Dr Pearson said that in the future it should be quite easy for exercise providers to classify people according to their social identities and then design interventions that will continue to be effective over a long period.

The implications for this research extend beyond exercise, particularly to education and training, and Dr Pearson is aiming to conduct further research into the effectiveness of social identity based interventions.

Dr Erin Pearson is available for interview.