5. Behavioral characteristics

We explored the role of eating frequency, sleep duration, physical activity, and smoking on AO. We also draw a parallel between gender differences and these associations.

#### 5.1. Eating frequency

Since the 1960s, studies have suggested an inverse association between the consumption of a greater number of meals per day and body weight maintenance [57]. Since then, the media, health professionals, and guidelines for health and weight management have followed this recommendation [58]. However, studies that have attempted to determine the effects of eating frequency on weight and body composition have reached different conclusions, and some scientists have called attention to the lack of solid evidence to justify such recommendation.

Thus, several review studies have been recently conducted in order to determine whether there is association between eating frequency and body weight and body composition [59– 62]. However, a few studies that investigated AO or WC or waist-to-hip ratio as outcome were retrieved in these reviews [63].

A negative association between eating frequency was found in several observational studies [64–68]. Most of them found that having three or fewer meals/day is associated with excessive abdominal fat, when compared with having more than three meals/day; however, this association can only be seen among men [64–67]. In general, increased eating frequency has been postulated to increase metabolism [69], appetite control, and food intake and to improve glucose and insulin control [70–72]. However, this association only among men could be due to the fact that men who have a high eating frequency also have a healthier lifestyle, including the practice of physical activity and healthier eating habits, which results in reduced body fat and WC [73].

On the other hand, a recent cross-sectional study used data from the NHANES 2003–2012 and found that eating frequency was positively associated with central obesity in men and women. Compared with the lowest eating frequency (≤3 times/d), the odds for AO in the highest category (≥5 times/d) were OR 1.42 (95% CI 1.15;1.75) in men and OR 1.29 (95% CI 1.05;1.59) in women [74]. This finding seems plausible, given the observed positive association between eating frequency and energy intake [67, 75, 76]. In addition, the authors called attention to the apparent inverse relation between eating frequency and adiposity measures, such as an artifact that in part can be attributed to the underreporting of eating frequency concomitant with the underreporting of energy intake by obese or overweight subjects [74].

Regarding experimental evidence, a meta-analysis study published in 2015 evaluated experimental research on meal frequency with respect to changes in fat mass and lean mass, and its results suggest a small potential benefit of increased feeding frequencies for fat mass and body fat percentage. In studies retrieved in this meta-analysis, only the Arciero's study measured the outcome as abdominal fat. This randomized trial investigated the effects of consuming traditional (15% of total energy) versus higher [35% of total energy] protein intakes as three or six meals/day on abdominal fat. Six meals per day in a high-protein condition were superior to three meals per day with a high-protein or traditional protein intake [15%] for decreasing abdominal fat in the absence of significant differences in total energy intake or expenditure. However, the study design did not rule out the possibility of an interaction between the two [eating frequency and protein intake] playing a role in mediating these responses [63].

Although observational studies suggest that men may benefit from a potential protective effect of a high eating frequency, women may not experience this protective effect. However, there is no sufficient evidence to establish a clear and strong association between eating frequency and AO, and it is necessary to conduct more observational and experimental studies.

#### 5.2. Sleep duration

neighborhoods compared to the residents of high education areas. In the same group, the

Thus, the findings suggest that the association between SES and AO is similar to the association with overall obesity. Separate studies of individual nations showed that high status people tend to have a smaller WC than others, in HIC, and this association is the same for women, in LMIC. On the other hand, among men from LMIC, the association between SES and AO is positive. First, this reversal in the relationship of SES and AO across levels of economic development highlights the importance of the national socioeconomic context of AO. Second, the high prevalence of AO among low SES women, in LMIC, shows an important health inequality. In some LMIC countries, low SES groups may now have sufficient access to cheap, calorie-dense, and processed food, as a consequence of the globalization of the fast food

We explored the role of eating frequency, sleep duration, physical activity, and smoking on

Since the 1960s, studies have suggested an inverse association between the consumption of a greater number of meals per day and body weight maintenance [57]. Since then, the media, health professionals, and guidelines for health and weight management have followed this recommendation [58]. However, studies that have attempted to determine the effects of eating frequency on weight and body composition have reached different conclusions, and some scientists have called attention to the lack of solid evidence to justify such recommendation.

Thus, several review studies have been recently conducted in order to determine whether there is association between eating frequency and body weight and body composition [59– 62]. However, a few studies that investigated AO or WC or waist-to-hip ratio as outcome were

A negative association between eating frequency was found in several observational studies [64–68]. Most of them found that having three or fewer meals/day is associated with excessive abdominal fat, when compared with having more than three meals/day; however, this association can only be seen among men [64–67]. In general, increased eating frequency has been postulated to increase metabolism [69], appetite control, and food intake and to improve glucose and insulin control [70–72]. However, this association only among men could be due to the fact that men who have a high eating frequency also have a healthier lifestyle, including the practice of physical activity and healthier eating habits, which results in reduced body fat

On the other hand, a recent cross-sectional study used data from the NHANES 2003–2012 and found that eating frequency was positively associated with central obesity in men and women.

AO. We also draw a parallel between gender differences and these associations.

chance of AO was 2.05 times higher [56].

78 Adiposity - Epidemiology and Treatment Modalities

industry and agriculture.

5.1. Eating frequency

retrieved in these reviews [63].

and WC [73].

5. Behavioral characteristics

Over the past several decades, the prevalence of chronic sleep deprivation has grown to epidemic proportions. According to the National Sleep Foundation, by 1998 only 38% of American adults were obtaining 8 h of sleep and that number had fallen to 28% by 2009 [77]. Evidence has grown over the past decades supporting the role of sleep duration as a notable risk factor for overall obesity [78, 79]. More recently, studies have suggested that measures of central adiposity, such as WC, are also associated [80].

Heorell-Haglöw's [80] cross-sectional study investigated the association between sleep duration and WC in 6,461 Swiss women. The study showed a U-shaped association: Both short sleeping (<5 h) and long sleeping (≥10 h) women had greater means of WC compared to normal sleepers (7–8 h) in younger women (<50 years old). Other cross-sectional study involving men and women found a similar U-shaped association, considering short sleepers (<7 h) and long sleepers (>8 h) [81]. However, several studies found that only short sleep, and not long sleep duration, is a risk factor for central obesity [82–84].

A few longitudinal studies have investigated the association between sleep duration and AO. A prospective study with a 10-year follow-up in women confirmed the U-shaped association found in some cross-sectional studies. Among younger women (aged <40 years), both habitual short (<6 h) and long sleep durations (≥9 h) were risk factors for central obesity. In addition, decreased sleep duration from normal to short duration during the follow-up was a risk factor for increased WC [85]. Another prospective cohort investigated the association of short sleep duration (≤5 h) among women in the first year postpartum with adiposity status at 3 years postpartum. The authors demonstrated that postpartum sleep of ≤5 h/day was associated with a higher WC at 3 years postpartum, compared to sleeping >5 h [86].

In order to assess the magnitude and consistency of the relation of insufficient sleep and WC, a meta-analysis of cross-sectional studies was conducted in 2014. The analyses included 21 studies (total of 56,259 participants); the results confirmed a significant negative relation between sleep duration and WC, and longer sleep duration was not investigated [86]. However, What is short sleep and long sleep? Grandner et al. summarized the laboratory findings and epidemiological studies that investigated the consequences of acute reductions in sleep time, and they proposed to describe short sleep duration as <6 h/day [87]. However, there is no robust definition of long sleep duration, although several studies have defined long sleepers as those who have a habitual sleep duration of more than 8 h.

Finally, clinical trials that measured AO as an outcome are sparse, probably because it is difficult to maintain individuals on sleep deprivation for a long period of time. In a randomized study of adults, sleep curtailment was shown to undermine dietary efforts to reduce adiposity, after 14 days of moderate caloric restriction with 8.5- or 5.5-h nighttime sleep opportunity. It reduced the fraction of weight loss by 55% and increased the loss of fat-free body mass by 60% [88]. The hormonal regulation that occurs during sleep and its multiple peripheral effects depend on sleep duration and quality, indicating that sleep deprivation has deleterious health effects. In a randomized study, acute sleep deprivation in adults has been associated with changes in thermoregulation and, consequently, reductions in total energy expenditure in humans. Sleep-deprived individuals may also increase their total caloric intake due to the impact of sleep deprivation on peripheral regulators of satiety. Studies have associated sleep deprivation with lower leptin and higher ghrelin levels, with consequent increases in appetite and weight gain [89–92].

Furthermore, reduced sleep duration is known to have consequences on individuals' social routines and lifestyles, particularly regarding dietary habits and level of physical activity. In this sense, short sleep duration is associated with irregular eating habits, such as a dominance of snacks over meals, increased intake of fat and sweets, increased energy intake and, on the other hand, reduced intake of fruits and vegetables, and lower physical activity level [93–95]. These findings can be explained in part by a disinhibited eating behavior [tendency toward overeating and eating opportunistically], which is associated with less healthy food choices, thus contributing to weight gain. Because short sleep duration inevitably results in more time and opportunities for eating, short-duration sleepers who also have a high disinhibition eating behavior trait will eat more and gain more weight than short-duration sleepers who have a low disinhibition eating behavior trait [96].

After more than three decades of studies about the metabolic consequences of changes in the patterns of sleep duration, we can conclude that there is sufficient evidence to support that short sleep duration plays an important role in the AO etiology in adults. However, the association between long sleep and AO lacks more solid evidence, and explanations about its causal chain are weak. Nonetheless, disrupted eating patterns, such as having a dominance of snacks over meals, a higher intake of fat and sweets, and a lower intake of fruits and vegetables, are associated with both short and long sleep duration [95].

#### 5.3. Physical activity

decreased sleep duration from normal to short duration during the follow-up was a risk factor for increased WC [85]. Another prospective cohort investigated the association of short sleep duration (≤5 h) among women in the first year postpartum with adiposity status at 3 years postpartum. The authors demonstrated that postpartum sleep of ≤5 h/day was associated with

In order to assess the magnitude and consistency of the relation of insufficient sleep and WC, a meta-analysis of cross-sectional studies was conducted in 2014. The analyses included 21 studies (total of 56,259 participants); the results confirmed a significant negative relation between sleep duration and WC, and longer sleep duration was not investigated [86]. However, What is short sleep and long sleep? Grandner et al. summarized the laboratory findings and epidemiological studies that investigated the consequences of acute reductions in sleep time, and they proposed to describe short sleep duration as <6 h/day [87]. However, there is no robust definition of long sleep duration, although several studies have defined long

Finally, clinical trials that measured AO as an outcome are sparse, probably because it is difficult to maintain individuals on sleep deprivation for a long period of time. In a randomized study of adults, sleep curtailment was shown to undermine dietary efforts to reduce adiposity, after 14 days of moderate caloric restriction with 8.5- or 5.5-h nighttime sleep opportunity. It reduced the fraction of weight loss by 55% and increased the loss of fat-free body mass by 60% [88]. The hormonal regulation that occurs during sleep and its multiple peripheral effects depend on sleep duration and quality, indicating that sleep deprivation has deleterious health effects. In a randomized study, acute sleep deprivation in adults has been associated with changes in thermoregulation and, consequently, reductions in total energy expenditure in humans. Sleep-deprived individuals may also increase their total caloric intake due to the impact of sleep deprivation on peripheral regulators of satiety. Studies have associated sleep deprivation with lower leptin and higher ghrelin levels, with consequent increases

Furthermore, reduced sleep duration is known to have consequences on individuals' social routines and lifestyles, particularly regarding dietary habits and level of physical activity. In this sense, short sleep duration is associated with irregular eating habits, such as a dominance of snacks over meals, increased intake of fat and sweets, increased energy intake and, on the other hand, reduced intake of fruits and vegetables, and lower physical activity level [93–95]. These findings can be explained in part by a disinhibited eating behavior [tendency toward overeating and eating opportunistically], which is associated with less healthy food choices, thus contributing to weight gain. Because short sleep duration inevitably results in more time and opportunities for eating, short-duration sleepers who also have a high disinhibition eating behavior trait will eat more and gain more weight than short-duration sleepers who have a

After more than three decades of studies about the metabolic consequences of changes in the patterns of sleep duration, we can conclude that there is sufficient evidence to support that short sleep duration plays an important role in the AO etiology in adults. However, the association between long sleep and AO lacks more solid evidence, and explanations about its

a higher WC at 3 years postpartum, compared to sleeping >5 h [86].

sleepers as those who have a habitual sleep duration of more than 8 h.

in appetite and weight gain [89–92].

80 Adiposity - Epidemiology and Treatment Modalities

low disinhibition eating behavior trait [96].

Strong evidence shows that physical inactivity increases the risk of many adverse health conditions, including coronary heart disease and type 2 diabetes mellitus, and it also shortens life expectancy [97]. The therapeutic role of physical activity in cardiovascular and metabolic disease involves multiple mechanisms, and one of these pathways appears to be via its effect on abdominal fat [98]. Physical activity has a role in the prevention, treatment, and control of AO. However, this chapter has focused on the epidemiology of AO, and thus we have decided to describe the predictor role of physical activity on the occurrence of AO.

Several prospective studies have evaluated the association between physical activity and AO. In the European Prospective Investigation into Cancer and Nutrition [EPIC] prospective cohort study, 84,511 men and 203,987 women were followed for 5.1 years. Physical activity predicted a lower WC in men and in women, regardless of baseline body weight, baseline WC, and other confounding factors [99]. A nationally representative population-based cohort study, The Australian Diabetes, Obesity and Lifestyle Study [AusDiab] found similar results. It followed 2,191 men and 2,650 women for 5 years. Men and women with AO showed the odds of reducing physical levels from baseline to follow-up for 1.40 (95% CI 1.10;1.79) and 1.44 (95% CI 1.16;1.80), respectively, compared to those with a normal WC [100].

In the same direction, long-term longitudinal studies (follow-up ≥ 20 years) have shown that youth physical activity had an effect on AO through the maintenance of physical activity in adulthood [101, 102]. A longitudinal study data by Finns on the cardiovascular risk in young people followed 1319 men and women for 21 years. Structural equation analysis indicated that the prevalence of AO in adulthood was directly affected by adult physical activity and indirectly via youth physical activity. Obesity was significantly reduced from youth to adulthood. The model accounted for 19% of AO in men and 13% in women [103]. Another long-term longitudinal study has showed sex differences in the association between physical activity and AO. The Coronary Artery Risk Development in Young Adults study (CARDIA), a 20-year follow-up prospective longitudinal study with 3,554 men and women, showed that maintaining high levels of activity was associated with smaller gains in WC compared with low activity levels. Men maintaining high activity gained 3.1 fewer centimeters in WC, and women maintaining higher activity gained 3.8 fewer centimeters [104].

Studies only among women showed similar results. In a prospective cohort with 233 middleaged women, after a 20-month follow-up, changes in physical activity remained as an independent predictor of abdominal fat change after adjusting for changes in total body fat and total energy intake. Compared with women who maintained or decreased physical activity, women who increased physical activity had approximately half the risk [RR (relative risk) = 0.52 95% CI: 0.27;0.98] of gaining abdominal fat [105]. In Southern Brazil, a case-control study investigated the association between the practice of physical activities in adolescence and AO in adulthood among women who were shift workers. Women who participated in five or more physical activities in adolescence were 50% less likely to have AO than women who participated in one activity or no physical activities (OR = 0.50 95% CI: 0.27;0.93) [106].

Finally, evidence from clinical trials can help us understand how physical activity reduces body fat, preferentially from the abdominal area. In the past decade, several systematic reviews have summarized results from randomized controlled and nonrandomized controlled trials [107–109]. They support moderate physical activity as an effective intervention for reducing abdominal fat in overweight or obese subjects. Also increasing physical activity expressed as energy expended per week was positively related to reductions in total adiposity in a dose-response manner. The gender influence presents controversial results due to low numbers for comparison among those reviews. However, the most recent review showed a stronger relationship between physical activity and visceral fat in women than in men. However, it is difficult to compare differences in the amount of visceral fat reduction by aerobic exercise between men and women, as women generally store a greater total fat mass relative to body weight than men. Also, energy cost of exercise interventions and energy expenditure associated with sedentary behavior may differ considerably between genders [109].

Furthermore, the interplay between physical activity and AO may differ by gender. Although the practice of physical activity may be more beneficial for women than men, it has been shown that inactive women with a large WC had a greater risk of cardiovascular heart diseases (CHD) than inactive men with a large WC [98, 110]. In this sense, Kin and Han [98] examined the interaction between physical activity and AO in relation to the Framingham Risk Score in 2,112 adults, using data from the 2007 Korean National Health and Nutrition Examination Survey (NHANES IV). The findings suggested that the risk of CHD associated with physical inactivity and AO was much stronger in women than in men. Inactive men with a large WC had an OR for CHD risk of 2.91 (95% CI 1.63;5.22), compared with physically active men with a normal WC, while inactive women with a large WC had an OR of 6.37 (95% CI 3.44;11.80), compared with active women with a normal WC [98]. Thus, at the same time that women could have greater benefits from the practice of physical activity compared with men, they also have more deleterious effects from physical inactivity.

#### 5.4. Smoking

Smoking is one of the major causes of preventable death in developed countries and presents a complex interaction with abdominal and general obesity. People with both conditions are at high risk of cardiovascular disease and have a substantially reduced life expectancy. Unfortunately, for many smokers, the fear of weight gain is a considerable barrier to cessation. Thus, understanding the role of smoking in the occurrence of AO is a very important issue.

Smokers have a lower body weight and BMI than nonsmokers. Nevertheless, the association between smoking and WC is inverse. Large cross-sectional studies have supported the hypothesis that current smokers tend to have a larger WC than nonsmokers, suggesting that smoking may increase central fat accumulation [111–113]; prospective studies showed that the number of cigarettes plays an important role in this association [111, 114, 115]. A cross-sectional study with a nationally representative sample of 9047 Scotland adults found that in women who were current smokers, the mean WC was higher compared with nonsmokers [111]. In the same way, a prospective study of five consecutive birth cohorts of Finnish twins (n = 4,296) analyzed the effect of adolescent smoking on AO in early adulthood. Smoking at least 10 cigarettes a day when aged 16–18 years increased the risk of adult AO in 34%, regardless of current smoking status [114]. A cohort of 79,236 white, non-Hispanic healthy adults estimated the effects of behaviors on a 10-year change in odds ratios for WC gain. Women who continued smoking up to a pack a day experienced a small protection against WC gain, but this effect could not be confirmed for women smoking more than a pack a day.

or more physical activities in adolescence were 50% less likely to have AO than women who

Finally, evidence from clinical trials can help us understand how physical activity reduces body fat, preferentially from the abdominal area. In the past decade, several systematic reviews have summarized results from randomized controlled and nonrandomized controlled trials [107–109]. They support moderate physical activity as an effective intervention for reducing abdominal fat in overweight or obese subjects. Also increasing physical activity expressed as energy expended per week was positively related to reductions in total adiposity in a dose-response manner. The gender influence presents controversial results due to low numbers for comparison among those reviews. However, the most recent review showed a stronger relationship between physical activity and visceral fat in women than in men. However, it is difficult to compare differences in the amount of visceral fat reduction by aerobic exercise between men and women, as women generally store a greater total fat mass relative to body weight than men. Also, energy cost of exercise interventions and energy expenditure associated with sedentary behavior may differ considerably between

Furthermore, the interplay between physical activity and AO may differ by gender. Although the practice of physical activity may be more beneficial for women than men, it has been shown that inactive women with a large WC had a greater risk of cardiovascular heart diseases (CHD) than inactive men with a large WC [98, 110]. In this sense, Kin and Han [98] examined the interaction between physical activity and AO in relation to the Framingham Risk Score in 2,112 adults, using data from the 2007 Korean National Health and Nutrition Examination Survey (NHANES IV). The findings suggested that the risk of CHD associated with physical inactivity and AO was much stronger in women than in men. Inactive men with a large WC had an OR for CHD risk of 2.91 (95% CI 1.63;5.22), compared with physically active men with a normal WC, while inactive women with a large WC had an OR of 6.37 (95% CI 3.44;11.80), compared with active women with a normal WC [98]. Thus, at the same time that women could have greater benefits from the practice of physical activity compared with men, they also

Smoking is one of the major causes of preventable death in developed countries and presents a complex interaction with abdominal and general obesity. People with both conditions are at high risk of cardiovascular disease and have a substantially reduced life expectancy. Unfortunately, for many smokers, the fear of weight gain is a considerable barrier to cessation. Thus, understanding the role of smoking in the occurrence of AO is a very important

Smokers have a lower body weight and BMI than nonsmokers. Nevertheless, the association between smoking and WC is inverse. Large cross-sectional studies have supported the hypothesis that current smokers tend to have a larger WC than nonsmokers, suggesting that smoking may increase central fat accumulation [111–113]; prospective studies showed that the number of cigarettes plays an important role in this association [111, 114, 115]. A cross-sectional study

have more deleterious effects from physical inactivity.

participated in one activity or no physical activities (OR = 0.50 95% CI: 0.27;0.93) [106].

genders [109].

82 Adiposity - Epidemiology and Treatment Modalities

5.4. Smoking

issue.

In addition, studies of smokers have shown a dose-response association between total smoking amount during a lifetime and AO. A cross-sectional study with 6,123 Caucasians from Switzerland found that compared with light-smoking men, the odds ratio for AO was 1.28 (95% CI 0.78–2;10) for moderate smokers and 1.94 (95% CI 1.15;3.27) for heavy smokers and 1.07 (95% CI 0.72–1.58) and 2.15 (95% CI 1.26;3.64) in women who were moderate and heavy smokers, respectively [115]. A prospective study in South Korea, with 283 subjects, found results that point to the same direction. WC and VAF area showed a J- or U-shaped association with total smoking during lifetime. And, after restricting the analyses to past/ current smokers, they found positive dose-response associations between smoking pack-years and AO [116].

Recently, epigenetic studies have shown that genetic factors may have an important role in this association. A genetic variant in chromosome 15 CHRNA5-CHRNA3-CHRNB4 gene region (rs16969968) was associated with smoking quantity. It was also associated with a lower BMI and a greater WC. In this sense, a Mendelian randomization of the meta-analyses of 29 studies comprising almost 150,000 participants using this genetic variant as a proxy for smoking heaviness was conducted. The results suggested that for every copy of the minor allele associated with cigarette consumption (i.e., increasing cigarette per day consumption by approximately one cigarette), WC will be increased by 0.14%, if BMI were to remain constant. These findings strengthen the hypothesis that a preferential redistribution toward central adiposity is associated with higher cigarette consumption.

The physiological mechanism linking smoking and AO was not fully elucidated, but cortisol and sex hormones could play a possible role in greater WC among smokers. Smokers have higher fasting plasma cortisol concentrations than nonsmokers, and VAF is influenced by the cortisol concentration. In addition, an imbalance between male and female sex hormones [estrogens/androgen in women] and a decrease in testosterone, in men, are observed in smokers [117]. At the same time, unhealthy behavior was also observed in subjects who smoked. Some researchers have pointed to the existence of dangerous behaviors that included smoking, lower physical activity, and unhealthy eating habits in adults, and these behaviors were usual among low SES subjects [106, 118].

Unfortunately, smoking cessation also has deleterious effects on the body composition. Several studies have reported an increase in weight after smoking cessation [119, 120]. A prospective study Inter99 (1-year follow-up) investigated changes in WC after smoking cessation. Smoking cessation resulted in substantial increase in weight and central fat. Women quitters gained more weight and had a higher increase in WC than men [121]. Kahn's cohort study showed that quitting smoking was associated with approximately doubling the likelihood of WC gain compared with those who were current smokers or those who never smoked [122]. However, the increase in WC may be explained by the increase in weight; consequently, future studies are needed to elucidate this issue.

The role of smoking in increased overall obesity is well documented in the literature, while its relationship with AO has received less attention. However, current evidence suggests a causal relationship between smoking and fat distribution. Smoking increases the risk of AO and, among smokers, the number of cigarettes smoked per day was also positively associated with AO. We still need further evidence in order to understand the real impact of smoking cessation in the body composition. Thus, smoking should be avoided in order to prevent AO; current smokers should be informed that they are more prone to central fat accumulation and to the inherent additional health risks; the increase in unhealthy central fat after smoking cessation needs attention, mainly among women, who pay more attention to body shape and could avoid smoking cessation.
