The current study provides further evidence that behavioral intervention in combination with diet and exercise produces modest long term weight loss maintenance and improvements in physical and mental quality of life measures. Study participants completed a 6 month behavioral intervention focused on increasing physical activity and adopting a healthy diet. At the end of the 6 months, the participants were randomized into two extended care treatment arms that received mailed intervention materials. At the end of two years, the participants maintained a 3 kg weight loss and 30% of the sample that returned for testing retained at least a 5% weight loss.
Many studies have shown that increasing levels of overweight and obesity are associated with decrements in the HRQL [12, 22]. Although other studies have found decrements in HRQL across all of the subscales, the current study sample, although overweight and obese was only below the age-specific population norms in bodily pain, vitality, and mental health, and therefore may not be obese enough to have impairments across all aspects of HRQL. Kolotkin and Crosby's  examined HRQL by BMI level and did not find consistent differences until individuals had BMI's greater than 35 kg/m2. The mean BMI in the current study was 32 kg/m2. One exception has been physical functioning, which has been shown to be impaired at BMI levels greater than 27–30 kg/m2 (e.g. [23–25]). However, that finding was not replicated in the current study.
Both the mental and physical composite scores improved at the end of the 6 month intervention and this was driven by changes in the physical functioning, general health, vitality, and mental health subscales. This parallels the findings of many other studies that have examined modest weight loss. For example, Fontaine et al.  studied 38 adults in a 13-week weight loss treatment program. Study participants lost an average of 8.6 kg and they reported improvements in physical functioning, role-physical, general health, vitality, and mental health. In a 12-week study, Rippe and colleagues  reported improvements in physical functioning, role physical, and mental health in 30 participants that lost 6.1 kg. A prospective analysis of the Nurses Health Study  reported that women that lost weight improved their physical functioning vitality, and bodily pain. In a study of a 4-month very low energy dietary intervention, there were transient improvements in many of the SF-36 scales .
Taken together these results suggest that it is possible to improve health related quality of life using behavioral interventions. Previous studies have consistently found improvements in physical functioning and many have found improvements in mental health, vitality, and role physical. The current study supported the improvements in physical functioning and also found support for improving general health, vitality, and mental health at the end of the 6 month intervention in which there was a moderate weight loss.
Given the ability of weight loss interventions to improve HRQL, it is necessary to examine long terms changes and what happens after the weight loss intervention ends. In the current study, at 1 year the scores on the physical composite scale were not significantly different than baseline levels, however the mental composite scale and physical functioning, general health, vitality, and mental health subscales all remained above baseline levels. At the 24 month follow-up, participants retained their improvements above baseline in the mental composite scale and the physical functioning, mental health, and vitality subscales.
The results of the current study have many similarities to the only other 2 year follow-up study of which we are aware . Kaukua et al.  reported modest weight loss at 2 year follow up with 1/3 of patients maintaining ≥ 5% weight loss. In the current study, 30% of the study sample maintained a 5% weight loss at 24 months. There was a peak of improvements for many of the HRQL measures at the end of the 6 month intervention, followed by a gradual return towards baseline which mirrored the changes in weight. Kaukua et al.  reported a similar pattern, but only physical functioning remained improved over baseline levels at 2 years. The mental health subscale was the only exception, in that it increased over the entire 24 months of the study. Unfortunately we do not have data on changes in anti-depressant medications or enrollment in psychotherapy that might help explain this pattern.
In contrast to our study, Kaukua et al.  reported significant group differences when examining weight loss categories. In particular, they found that a 10% weight loss was necessary for improvements in physical functioning, physical role functioning, bodily pain, general health, vitality, and mental health. The results must be interpreted with some caution, as there were only 9 participants out of the 126 in the study that maintained a weight loss greater than 10% of their initial body weight. The current study used a cutoff of 5% weight loss or greater (30% of participants) and found no main or interaction effects. The lack of significance of the amount of weight loss on changes in HRQL has been previously reported. Kolotkin et al.  reported that only 14% of the changes in HRQL scales could be explained by weight loss. Similarly, Mathias et al.  reported that only 2 of 7 quality of life measures were different among individuals who lost greater than 5% of their weight compared to those that had stable weights (± 5%) and those that gained weight (> 5%). Fontaine et al.  also reported no difference among weight loss maintainers or regainers.
There is clearly a need to develop a better understanding of what is leading to improvements in HRQL among overweight and obese adults beyond weight loss. It is possible that behavioral factors such as exercising and changing diet can explain the improvements in HRQL . It is also possible that the social interaction and support of the weight loss intervention is responsible for some of the improvements in HRQL. There is also a need to understand how to maintain improvements after completion of the intervention. In the current study, despite participants regaining weight, there were still improvements in vitality, physical functioning, and mental health at 24 months. An understanding of what programmatic aspects influence HRQL may help in the development of interventions that can foster continued improvements even after the formal intervention is over.
The majority of studies on obesity and HRQL have been examined from the perspective of surgical and/or pharmacological treatment for the severely obese. This study adds to the growing literature on the effects of behavioral interventions in producing more modest changes in weight that also can positively impact participants' quality of life. Further research is needed to examine the differential effects of very low energy diets, low fat diets, and low-carbohydrate diets. As research begins to suggest that the different diets may result in similar long-term weight loss results , it is possible that there may be differential effects on quality of life that are impacted by participants feelings of food choice and caloric restriction. It is also possible that different exercise prescriptions, such as different intensities and formats, may have differential impacts upon HRQL outcomes.
In general, the results of the current study are consistent with the few existing long term studies on health related quality of life and weight loss. However, there are several limitations to the current study. The current study only used a generic measure of HRQL (SF-36). Our results may not be the same if we used the obesity specific scale, such as the Impact of Weight on Quality of Life scale. The study should be replicated using multiple measures, including obesity specific and general HRQL. The current study sample was primarily white, female, and well-educated. A limitation of our analyses was the need to have complete data across all four time points. Therefore we were only able to analyze 48% of the participants that were originally randomized into the trial. The individuals who participated in the study were volunteers; therefore, they may differ from general population on some important characteristics. There is some research to suggest that individuals who seek out clinical treatment for obesity are more likely to have HRQL impairments than those not seeking to lose weight , although the current sample was relatively similar to age norms for HRQL. Therefore, replication should be done using different samples to increase generalizability.