This community based lifestyle intervention is feasible based on the reported population reach, attendance, retention and resources required to deliver the program. Population reach is demonstrated by successfully recruiting a representative sample of the target population, that is, women with young children. We delivered all components of the program and achieved high attendance and retention rates. One trained facilitator was able to deliver all intervention components.
The primary school as a setting for recruitment and delivery was successful, addressing a number of barriers to participation by women in health programs. the school is an existing trusted community resource with strong social connections often formed between parents. This social support has been found to have positive influence on physical activity participation. The school is accessible to most women, often within walking distance, and is familiar. Using the school we successfully attracted working-mothers, single-mothers and women from culturally diverse backgrounds, all income levels and education levels. School principals strongly supported the intervention and actively assisted the recruitment. As recruitment rates at each school varied, characteristics of the parent body may affect participation and warrants further investigation.
Overall approximately 11% of invited participants responded. In the context of population reach this is still a significant proportion. Improving the health of 11% of the population would equate to an important public health campaign. This setting gave us access to all women with children who were unselected in terms of health risks.
As expected, we attracted those who were overweight and obese, yet still attracted many leaner women. This is encouraging for future interventions aiming to prevent rather than treat obesity. Almost all of the participants (90%), reported they would like to weigh less, confirming women are generally dissatisfied with their body weight and shape. More than a third of women wished to weigh 1–5 kg less, suggesting weight gain to date is small in many women and potentially reversible. These women are therefore ideal candidates for interventions to prevent obesity, as risk to health increases with an adult weight gain of 5 kg or from a BMI of 22 kg/m2.
The majority of women reported they were actively trying to prevent weight gain over the past year, yet few were successful. This lack of success is also demonstrated by an ongoing increase in the prevalence of obesity in women in Australia and in many other countries. The data here suggests many women are motivated to self-manage weight, but make largely unsuccessful changes to behavior. Interventions should be designed to enhance effectiveness of self-management strategies currently used by women and target those women who are within the healthy weight range but have begun experiencing small steady annual weight gain.
Attendance was voluntary and we may have attracted a more motivated group than is found in the population in general. As most women wished to weigh less, it is possible women participated in this lifestyle intervention with the intention of controlling their weight rather than for specific health improvements associated with diet and physical activity change. This implies body weight is a strong motivator for attendance at such programs. The weight management practices described in our study is comparable to that described in large population studies. Therefore weight issues may motivate many women to seek assistance for changing behaviors, and overall this may be a stronger influence than for single component behavior change, such as changing diet composition or physical activity levels alone.
Defining the minimum level of contact and support to achieve desired outcomes in interventions is critical. Intensive programs increase the burden on participants and are costly, and low intensity mail based interventions have shown poorer outcomes but have higher population reach. In addition women who nominate a preference for group contact do not always attend as planned,  which has implications for delivery. In this study, attendance progressively declined over the three group sessions. We propose three face to face sessions are the limit for mothers who need to re-arrange usual activities and work commitments to attend. Non attendance may also occur because of a lack of engagement, inability of the program to meet individual expectations, or factors related to the facilitator. Retention rates were high at 4 months and women reported the sessions to be very helpful, suggesting the program quality met expectations.
Intervention strategies such as 'delivery by a health professional' and 'sessions held at school' were highly valued. Similar programs should be delivered by health professionals, a trusted source of health information, in local settings. 'Getting a walking group started at school' was considered least helpful. It was anticipated the school setting would provide strong social support for participants, and social support has been shown to be strongly correlated with physical activity in women.  However the unstructured format of the walking groups was not sustainable. Future research could focus on how women successfully acquire and maintain social support for physical activity.
Children have been nominated as barriers to the adoption of healthy lifestyles in women. Lack of time, lack of childcare to assist exercise opportunities and preparing meals that children prefer, have been reported elsewhere. We would expect the diet of women with children will differ from childless women, although there is no comparative nutrient intake data available. Ball et al report women who live with children are less likely to meet dietary recommendations for fat intake and 'extra' foods than women living alone. In our sample, the total dietary fat and saturated fat intake were found to be similar to population data, and higher than recommended by successful weight loss and chronic disease prevention interventions. The Diabetes Prevention Program aimed for a total fat intake less than 30% of energy, the Women's Healthy Lifestyle Project, less than 25% and the Women's Health Initiative, 20%.[8, 39, 40] A low fat intake is frequently recommended for weight management. Self reported dietary intake is often underestimated, so actual intake in this group may be even higher than reported. These results are in contrast to almost half of the participants claiming to have made changes toward a low fat diet. Overall this is pointing to a discrepancy between perceived dietary fat intake and actual fat intake and confirms prior studies. The difference may reflect a lack of knowledge on sources of fat in the diet or inaccuracies associated with self-reported intake. This has important implications for weight gain prevention strategies suggesting a role for further education on the appropriate quantity and quality of dietary fat intake in this population.
Participants were more confident in changing diet behaviors than physical activity. Women with children are likely to perceive a change in diet is within their personal control and skill level. They may also perceive fewer barriers to dietary change than for physical activity change, such as, arranging child care, re-arranging schedules, cost, weather and a lack of places to be active.
Reported average steps in healthy adults range from 7,000–13,000 per day  and our sample fell within that range. Reduced activity associated with having children has been reported previously  and step rates were expected to be lower than that observed. Much of the activity in this group appears to be low intensity presumably associated with home duties and caring for children. Women are likely to find it difficult to increase physical activity levels when already fatigued through extensive low intensity activity. Specific advice on participation in all levels of activity, low, moderate, vigorous and sitting time might improve overall activity in this group. The pedometers were sealed to eliminate feedback potentially motivating participants to be more active than usual. It is possible that simply wearing the pedometer may increase motivation to increase step counts in some women.
Limitations included recruitment of a slightly more educated sample than the general population. Education level may affect rate of weight gain. Also we attracted few full time working mothers, as the intervention was held primarily during the day. Women with children are likely to find attending programs out of work hours difficult. Workplace interventions may be more appropriate for this group.
We have reached a group of women who are dissatisfied with their current weight, are attempting to control their weight, and are more confident of changing their diet than physical activity. The lifestyle behaviors exhibited by this population are not robust enough to prevent weight gain and are possibly contributing to an increased risk of cardiovascular disease. Dyslipidemia when combined with other risk factors such as overweight, low activity, and poor diet increases the risk of CVD occurring at an earlier age. The findings highlight the need for increased risk awareness and lifestyle change at earlier life stages to improve long term health outcomes in women.
Overall the results support the need for small adjustments in diet and physical activity behaviors in women which is likely to have important consequences for body weight and health. This study is novel as it delivers the intervention in a school setting aimed at the mothers. The strengths of this study are the randomized controlled design, the large representative sample of community based women and data collection using both self reported and objective methods. If successful long term, the intervention could be adapted for use in various community settings involving women, and would be a valuable addition to interventions aimed at children in a school setting.