Physical activity is a key behavior for treating overweight and obesity [2]. However, weight loss resulting from physical activity when not coupled with a concurrent reduction in energy intake appears to be quite modest. The 2008 Physical Activity Guidelines Advisory Committee Report [8] concluded that 180–270 min/week of physical activity results in a weight loss of approximately 0.5–3.0 kg. This magnitude of weight loss resulting from an increase in physical activity alone is consistent with the conclusions drawn from other systematic literature reviews [9]. In studies of 3–6 months in duration, interventions that have focused exclusively on physical activity have resulted in weight loss of approximately 0.5–2.0 % of initial body weight [10, 11]. Jakicic et al. reported a similar degree of weight loss of approximately 2 % of initial body weight at 6 month and 1 % at 18 months in overweight adults prescribed home-based physical activity [12].

Despite reports of these modest effects of physical activity on weight loss, there may be a dose–response effect, with greater weight loss being achieved with higher doses of physical activity. A recent systematic review of the literature reported that while there does not appear to be a significant change in body weight in response to <150 min/week of physical activity, physical activity of >150 and 225–440 min/week is associated with weight loss of 2.0–3.0 and 5.0–7.5 kg, respectively [13]. A secondary analysis of an 18-month intervention study conducted by Jakicic et al. also found a dose–response relationship [12]. Moreover, it was reported that a higher level of physical activity (approximately 160 min/week above baseline levels) was observed in participants who lost >3 % of their initial weight (mean weight loss of approximately 8 %) compared to those who remained weight stable or gain weight over the 18 months of follow-up [12].

Clinical guidelines recommend the inclusion of both dietary modification and physical activity to maximize weight loss [2, 13, 14]. Physical activity adds approximately 0.5–3.0 kg of weight loss to what can be achieved with a dietary intervention alone [2, 9, 13, 15]. For example, over a 6-month intervention, Goodpaster et al. reported a weight loss of 8.2 kg in response to diet alone versus 10.9 kg in response to the combination of diet plus physical activity, a difference of 2.7 kg [15]. A similar pattern has been reported by others [10, 11]. Based on a systematic review, Curioni and Lourenco [16] concluded that there is a 20 % greater weight loss with diet combined with physical activity compared to diet alone.

Physical activity appears to be an important predictor of improved weight loss in interventions of ≥12 months duration, and close examination of these data suggest that relatively high doses of physical activity improve long-term weight loss outcomes [13]. Wadden et al. [17] report that a mean dose of 287 min/week of physical activity was associated with a mean 1-year weight loss of 11.9 % in the Look AHEAD Study, and physical activity was the strongest correlate of weight loss achieved at 1 year. Unick et al. have also shown MVPA to be the strongest correlate of the ability to achieve a ≥10 % weight loss within the context of an intervention that also include a prescribed reduce calorie diet [18]. Secondary data analysis from another study of obese adults that has shown that followed subjects for an additional 6 months after an initial 6-month weight loss intervention, those individuals who lost weight in response to the intervention increased objectively measured physical activity, represented as MVPA and steps per day, while those who gained weight reduced their levels of physical activity [19]. Furthermore, change in steps per day (r = −0.29, p < 0.007) and minutes of MVPA (r = −0.27, p < 0.01) were inversely correlated with weight change. These findings across studies support the importance of physical activity as an important lifestyle behavior that is associated with improved long-term weight loss and the prevention of weight regain.

There is a growing body of literature suggesting that a relatively high dose of physical activity is associated with improved long-term weight loss and prevention of weight regain following significant weight loss. Jakicic and colleagues have repeatedly shown that ≥250 min/week of MVPA (~2,000–2,500 kcal/week) is associated with the greatest long-term weight loss and prevention of weight regain [20–23]. Results of a randomized trial conducted by Jeffery et al. [24] also demonstrate that prescription of 2,500 kcal/week of physical activity resulted in greater weight loss than prescription of 1,000 kcal/week, when both doses of physical activity were combined with a reduced calorie diet. Tate et al. [25] has also reported that continued engagement in ≥2,500 kcal/week of physical activity is associated with improved long-term weight loss within the context of a comprehensive weight loss intervention program.

Additional evidence supporting the idea that relatively high amounts of physical activity promote long-term weight loss maintenance comes from the National Weight Control Registry, which is an observational study of individuals who have sustained a weight loss of ≥30 lb for at least 1 year. Klem et al. initially characterized this sample (N = 784) and observed that individuals in this registry were reporting >2,800 kcal/week of physical activity [26]. A follow-up analysis of individuals in this registry (N = 3,683) conducted by Catenacci et al. confirmed that relatively high levels of physical activity were associated with sustained weight loss [27]. While these findings were based on self-reported physical activity, a more recent report that objectively measured physical activity supports the hypothesis that participants in the National Weight Control Registry are engaging in more structured period of physical activity (approximately 41 min/day) when compared to normal weight adults (approximately 26 min/day). These results suggest that relatively high amounts of physical activity are needed to sustain weight loss and prevent weight regain. This dose of physical activity is also consistent with the 250–300 min/week of physical activity that is recommended by the American College of Sports Medicine to improve long-term weight loss and to prevent weight regain [13].

Unpublished data from a recently conducted study by Dr. Jakicic and colleagues used objective methods to assess physical activity and its association with long-term weight loss [22]. The use of objective measurement of physical activity has allowed for improved understanding of the patterns of physical activity that may be associated with successful weight loss, and allows for the examination of the association between light-intensity physical activity (1.5 to <3.0 metabolic equivalents [METS]) and weight loss. The preliminary findings of this study demonstrate that ~250 min/week of objectively measured MVPA that was accumulated in bouts of ≥10 min was associated with the greatest weight loss averaging ~15 % of initial body weight at 12 and 18 months of the intervention. These findings also show that participants who had the greatest weight loss and the greatest increase in MVPA also had the greatest increase in light-intensity physical activity, suggesting that this may have also contributed to improved weight loss. While not definitive, these preliminary findings suggest that while weight loss interventions should focus primarily on increasing MVPA in overweight and obese adults, recommendations to also increase light-intensity physical activity within the context of one’s occupation, household, or other lifestyle activities may also be important to maximize long-term weight loss.

The inverse of physical activity is sedentary behavior, and there is increasing interest in the role that sedentary behavior plays in both the development and treatment of obesity. A consistent association has been observed between measures of sedentary behavior and increased risk of developing obesity [28]. Ball et al. [29] reported an association between hours of sitting and risk of gaining ≥5 % of initial body weight over a period of 4 years. Using television viewing as a proxy for sedentary behavior, Hu et al. [30] reported that television viewing time was associated with an increased risk of becoming obese over a 6-year period. Moreover, each additional 2 h/day of television viewing was associated with a 23 % increased risk of becoming obese. In contrast, Ekelund et al. reported that sedentary behavior was not predictive of future weight gain and obesity in adults [31]. Thus, additional research is necessary to determine the degree to which sedentary behavior contributes to weight gain and obesity.

Sedentary behavior may be linked to weight gain and the development of obesity due to a number of factors [28]. For example, sedentary behavior may substitute for engagement in other more active behaviors, which results in an overall reduction in energy expenditure. Engagement in sedentary behavior may also be linked to more frequent eating which results in increased energy intake and weight gain. These two factors may also work in combination, resulting in both a decrease in energy expenditure coupled with an increase in energy intake.

In addition to physical activity influencing body weight by increasing energy expenditure, it may also influence body weight indirectly by affecting energy intake and eating behaviors. Jakicic and colleagues have shown that overweight and obese adults who are compliant with engagement in higher amounts of physical activity also appear to have higher compliance with dietary change [18, 32]. Based on secondary data analysis, change in physical activity was significantly correlated with weight loss (r = 0.33), reductions in energy intake (r = 0.20), and improvements in eating behaviors associated with weight loss (r = 0.24; p < 0.05). Change in physical activity remained a significant predictor of weight loss after controlling for changes in energy intake and weight loss eating behaviors. These results suggest that physical activity has a direct influence on body weight; however, it may also indirectly affect body weight by influencing energy intake and eating behaviors. Unick et al. has also reported that physical activity was the most significant predictor of the ability of subjects to achieve ≥10 % weight loss, with physical activity outweighing engagement in various dietary behaviors that are associated with weight loss [18]. DeLany et al. [19] found an association between both increased physical activity and lower energy intake with weight change over a 6-month intervention period. These findings suggest that a significant increase in physical activity during a weight loss intervention may be accompanied by a greater reduction in energy intake, which may contribute to the greater magnitude of weight loss. However, it is unclear if physical activity has a direct influence on energy intake or whether physical activity and diet operate as independent lifestyle behaviors for weight loss.

Unpublished observations from studies conducted in our research laboratory at the University of Pittsburgh have shown that in response to an exercise alone intervention in overweight adults, there is a significant (p < 0.001) increase in dietary restraint, even when the intervention did not include recommendations for dietary restriction. Moreover, we have been interested in whether the type of dietary restraint, based on subscales of rigid (strict approaches to dietary change) or flexible (moderate approach to dietary change) restraint, is associated with changes in physical activity that may mediate the observed change in body weight over 6–18 months. We examined data from overweight adults (BMI = 27 ± 1.73 kg/m²; age = 45.5 ± 7.69 years) and found that both total dietary restraint and the flexible dietary restraint subscale partially mediated the relationship between physical activity and change in body weight. We also retrospectively grouped subjects based on having lost >3 % of baseline weight (−7.16 ± 3.43 kg), remaining weight stable defined as within ±3 % of baseline weight (−0.061 ± 1.73 kg), or gaining >3 % of baseline weight (+5.07 ± 2.15 kg) in response to a physical activity only intervention. This analysis showed a significant increase in flexible dietary restraint in those who lost weight in response to a physical activity intervention compared to those who remained weight stable or those who gained weight. However, there were no differences for change in rigid dietary restraint between these three weight change groups in response to physical activity. Thus, it appears that dietary restraint, and in particular flexible dietary restraint, may partially mediate the association between physical activity and weight change in overweight adults. These findings may suggest a pathway by which physical activity affects dietary intake and eating patterns; however, whether this varies by the dose of physical activity is unclear.

It has been shown that positive affect increases in response to acute physical activity, with no change in negative affect [33, 34]. Moreover, when grouped by whether there was an increase or decrease in positive affect in response to physical activity, it was reported that subjects who had an increase in positive affect in response to exercise ate fewer calories when presented with an ad libitum meal compared to subjects who had a decrease in positive affect with physical activity. There was no association between change in negative affect in response to physical activity and calories consumed when presented with an ad libitum meal. Thus, it appears that changes in positive affect in response to physical activity may be associated with energy intake. This may suggest that physical activity modes and doses that result in an increase in positive affect may be preferred when prescribing physical activity to overweight and obese adults. However, this may need to be based on the individual preferences of the participant.