There are a range of methods that are more accurate than BMI for estimating body fat [26]. The most precise methods are imaging approaches including CT and MRI. These methods are used in research studies to not only measure body fat but to measure visceral fat content which is more closely related to metabolic disorders than total body fat. Dual energy X-ray absorptiometry (DXA) provides accurate estimates of body fat and can provide some information on the relative amount of abdominal fat as compared to lower body subcutaneous fat. However, DXA has limited clinical utility because of cost and patient exposure to radiation. Underwater weighing used to be used to estimate percent body fat based on the principle that lean tissue is more dense than fat. This method used body weight and volume to calculate density and from that, estimate body fat content. A more recent technique that uses this same principle is air displacement plethysmography (Bod Pod; Cosmed and others). The advantages of this method are that it is almost as accurate as DXA, is quick and relatively easy to perform, is less expensive than DXA and does not expose the patient to radiation. It however does not provide information on regional fat distribution. Bioelectrical impedance analysis (BIA) provides an estimate of body fat based on the differential conductivity of lean tissue as compared to fat tissue. A range of devices are available that measure conductivity between 2 fingers, 2 hands, a hand, and a foot or 2 feet (Tanita, Omron, and others) and use this measurement to estimate body fat. While this method is portable, easy to perform, inexpensive and has minimal risk (not recommended for individuals with pacemakers), it is not as accurate or reproducible as the other methods listed above. It is affected by a patient’s state of hydration and is less accurate in very obese individuals [27].

While obtaining estimates of body fat from one of these methods is more accurate than estimates of fatness determined by BMI, there are currently no broadly accepted guidelines for what a healthy level of body fat is for adult men and women across the lifespan. Estimates of body fat may be helpful in motivating patients to start a weight loss program and in providing ongoing positive feedback and motivation during weight loss.

Recently the relationship between modest increases in BMI and mortality has been reexamined. A meta-analysis of 97 studies of more than 2.88 million individuals and more than 270,000 deaths found that the lowest relative risk of mortality was seen in individuals with a BMI between 25 and 30 kg/m², not those with a BMI < 25 kg/m². This study found that while the risk of mortality rose in obese individuals considered together, there was no evidence of increased mortality in those with a BMI between 30 and 35 kg/m² [28]. While this conclusion has been challenged [29] it may be that adverse effects of obesity are mostly seen in younger and middle aged individuals where excess adiposity predisposes to the development of diabetes and cardiovascular diseases but that once these disorders develop excess weight may be less harmful or may even be advantageous. The surprising finding that obese individuals with a range of health problems may actually do better than their lean counterparts has been termed the “obesity paradox” [30].

In addition, studies have shown that 25–30 % of obese individuals do not have evidence of metabolic disease [31, 32]. It is not clear that these so-called “healthy obese” individuals are at increased risk for morbidity or mortality when compared to normal weight individuals who have markers of insulin resistance. One recent meta-analysis that included eight studies of more than 61,000 individuals found that metabolically healthy obese individuals had increased risk of all-cause mortality and/or cardiovascular events as compared to metabolically healthy normal weight individuals. However, all metabolically unhealthy groups, normal weight, overweight and obese had increased risk compared to the metabolically healthy obese subjects [33]. This study found no difference in risk between normal weight, overweight and obese subjects who were metabolically unhealthy. While this issue is controversial and far from settled, it does seem clear that BMI alone is not sufficient for risk assessment in overweight and obese individuals and that other factors such as blood pressure, insulin resistance, hyperlipidemia, and systemic inflammation likely play important roles in the development of metabolic disease and should be considered when assessing the overweight or obese patient.

In response to the perceived limitations of a “BMI centric” approach to obesity risk assessment, a number of alternative strategies to risk assessment have been proposed. The oldest is the concept of the “metabolic syndrome.” The cluster including insulin resistance, glucose intolerance, hypertension, hyperlipidemia, activation of inflammatory pathways, endothelial dysfunction, and non-alcoholic steatohepatitis has been called syndrome X, the insulin resistance syndrome and other names, but most now refer to this condition as the metabolic syndrome. The metabolic syndrome came into broader awareness when formal diagnostic criteria were proposed first by the World Health Organization and then the National Cholesterol Education Program in their Adult Treatment Panel III guidelines (NCEP-ATPIII) [34]. Table 12.3 lists the diagnostic criteria for the metabolic syndrome advocated in these older guidelines. The American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD) took the provocative position that there is currently inadequate information available to accurately define the metabolic syndrome and that this designation should not be used in routine clinical practice [35]. This view grew out of a belief that the root cause of this clustering is not known and could include obesity, insulin resistance, or inflammation. The authors of this paper also felt that while the clustering of these conditions increases the risk of cardiovascular disease, it was not clear that the syndrome had any greater risk than the sum of the component parts. However, most experts agree that considering a range of variables in the risk stratification of overweight and obese patients is important. Disagreement comes in what variables to include and how to weigh these variables [36, 37].

A second approach to risk stratification was proposed in 2009 by Sharma and Kushner [38]. This system known as the Edmonton Obesity Staging system is depicted in Table 12.4. It focuses on the patient’s risk for and the presence of both cardiovascular and mechanical complications of obesity to “stage” obese patients with the goal of targeting treatment efforts to those who are the most likely to benefit. A strength of this system is that it takes into account more than just cardiometabolic risk in assessing the burden of disease associated with obesity. One limitation of this system is that it relies on clinical judgment to determine the specific level of disability. A second concern is that it does not establish quantitative cut points for many of the characteristics that are in the evaluation scheme.

A third approach has recently been proposed by Garvey and coworkers [39]. This approach that they call the “Cardiometabolic Disease Staging System” depicted in Table 12.5, divides patients into five risk categories using specific measurable parameters readily available to care providers including waist circumference, blood pressure, fasting blood levels of glucose, 2 h glucose levels during an oral glucose tolerance test (OGTT), fasting triglycerides, and HDL-C. The advantages of this system are that the parameters are quantitative and the cut points are based on several large epidemiological studies of the Coronary Artery Risk Development in Young Adults (CARDIA) cohort and the National Health and Nutrition Examination Survey (NHANES) cohort.

One should not conclude that there is doubt about the adverse health effects of obesity, only that emerging data suggests that the relationship between weight and health is complex. These different staging approaches have not been embraced by the most recent guideline documents but both the 2013 AHA/ACC/TOS Obesity Guideline as well as the 2013 ACC/AHA Guideline on the Assessment of Cardiovascular Risk [37] advocate consideration of other factors besides BMI in the risk stratification of obese patients.

The pattern of weight change over time in an individual patient often gives important clues as to likely causes of weight gain, past successes and challenges in weight loss and the reasons why the person is seeking assistance with their weight at this time. Asking questions about the history of weight gain including maximum lifetime weight, factors that were associated with periods of weight gain, successes and limitations of previous weight loss attempts. Kushner has suggested that one way to get at this information efficiently is to have the patient draw a graph of their weight over time [40]. In this manner, triggers for weight gain such as pregnancy, smoking cessation, the introduction of a new medication, depression or a musculoskeletal injury can be identified and the clinician can help the patient see the connection between these events and weight gain [40]. A history of obesity during adolescence with progressive weight gain during adulthood strongly argues against a medical condition such as Cushing’s syndrome or hypothyroidism as the cause of obesity.

Assessing previous weight loss efforts is also important. Many patients comment with frustration that “diets never work for me.” Often though, when discussed in greater detail, previous efforts are revealed to have produced the expected degree of weight loss (3–8 %) that was not maintained because of difficulties in sustaining the chosen weight loss strategy. Acknowledging and exploring these previous weight loss attempts can provide a useful platform for discussing the amount of weight that is commonly lost with a diet and exercise program and to explore strategies that were or were not successful previously as a prelude to a discussion of potential future approaches to treatment. Asking specifically about what prompted previous weight loss attempts, how much weight was lost, what was successful about those previous attempts and what were the circumstances of the termination of those efforts can help you understand how to help the person plan future weight loss attempts. This kind of discussion allows the clinician to provide empathy and support around what are extremely common, almost expected periods of relapse. In addition, the patient’s own experiences can be leveraged to emphasize the critical need for long-term behavior change strategies if maintenance of weight loss is the goal. It is important to emphasize to the patient that they can learn from previous weight loss attempts and that if they do, future attempts need not be a replay of prior attempts. Elements of treatment such as cost, time commitment, social support, types of foods consumed, self-monitoring, exercise, and the impact of special occasions, chronic illnesses, vacations and work can be explored. Things that did work as well as barriers to success can be identified and incorporated into a new plan.

Weight change is produced by a long-term imbalance between energy intake (EI) and energy expenditure (EE). Weight gain only occurs when EI > EE, and weight loss will only occur when EE > EI. The problem is that it is extremely difficult to accurately measure either EI or EE in a clinical environment. An extensive body of research demonstrates that virtually everyone underestimates EI when asked to self-report food intake. The best measure of EE is a method known as doubly labeled water. This method can accurately determine EE over a period of weeks in free living individuals. If weight is stable then EE = EI. In a number of studies self-reported food intake underestimated measured EE by an average of almost 30 % [41, 42]. A number of factors including BMI, previous weight loss history, and fear of negative evaluation have been shown to be associated with underreporting of EI [43].

The reality that people tend to underreport food intake however does not undermine the importance of gaining as much information as is reasonably possible on this important parameter. Information on food intake can be easily obtained in an office visit using a 24 h., 3 day or 7 day dietary recall or a food frequency questionnaire. Information about meal patterns, fast food consumption, calories consumed in beverages and “trigger foods” that tend to be overeaten can be identified. Diet record forms can be printed and available in the office so that patients can collect more extensive information between visits. Tools that help patients estimate portion sizes can help improve the quality of information obtained from diet records as well as building a foundation on which dietary interventions can be built. In fact self-monitoring of the diet appears to be one of the most important features of both successful short and long-term weight loss [44]. Keeping detailed food records can provide useful information not only about the foods that were consumed but about situations and precipitating factors associated with overeating. The patient can be encouraged to look for and record details of the “chain of events” that led to a loss of control over food choices. Were meals skipped? Was stress involved? What were the circumstances around which the particular foods overeaten were available? Was food eaten while the person was engaged in other activities such as television watching? In this manner the patient can begin to identify points along this sequence of events that could be modified through alternative approaches to similar situations that will likely recur in the future. While the information may not be completely accurate, asking for a self-report of food intake such as a 24-h dietary recall on each office visit emphasizes to the patient that the clinician feels that this information is critical in assessing weight health.

For those patients who use the internet and computer programs regularly, a number of diet monitoring tools are available for either PDA or PC based use. The US Department of Agriculture has a website that allows individuals to track their diet (https://​www.​supertracker.​usda.​gov/​default.​aspx) and another site where information on recommended intakes of a wide variety of nutrients can be found (http://​fnic.​nal.​usda.​gov/​dietary-guidance/​dietary-reference-intakes). While these sites are free and contain a good deal of useful information, many patients find them difficult to navigate and find the lists of foods incomplete for diet logging. Some other sites that are well reviewed by consumers for dietary self-monitoring include MyFitnessPal (www.​myfitnesspal.​com), Sparkpeople (www.​sparkpeople.​com, also provides social support and weight loss advice) and CalorieKing (www.​calorieking.​com, has an extensive database of foods that can be used to estimate energy intake). These are just a few of the many excellent sites available at this time for dietary self-monitoring.

Energy expenditure is made up of three components: basal metabolic rate (BMR), which can be estimated as resting energy expenditure (REE) which has also been called resting metabolic rate (RMR), thermic effect of food, which makes up only a small fraction of total daily energy expenditure, and energy expended in physical activity (EEPA), which is by far the most variable between individuals. Although patients often complain that they have a “low metabolic rate,” careful studies have conclusively shown that REE is linearly related to lean body mass [45]. This means that heavier people have higher REE than thin individuals, and as a result need to eat more on average each day to maintain their higher weight. It is likely that the rise in prevalence of obesity is the result not only of increased EI associated with the modern food environment, but also due to a reduction in the habitual levels of EEPA associated with a modern environment filled with technologies designed to reduce the need for physical labor [46, 47]. There is increasing evidence that the low levels of physical activity that characterize a sedentary lifestyle are associated with not only obesity, type 2 diabetes, cardiovascular disease, but also some types of cancer and increased overall mortality [48–52]. Conversely, increased levels of physical activity and high levels of cardiorespiratory fitness are associated with reduced levels of morbidity and cardiovascular mortality [53–55]. A physically active lifestyle is one of the top ten health indicators for Americans in the Healthy People 2020 objectives [56].

Clinicians can and should solicit information about usual levels of physical activity as part the initial evaluation and at follow up visits. Questions such as “how often do you engage in planned physical activity?” or “do you ever walk for exercise?” can be helpful. Asking about participation in sports or active pursuits in the past can also provide a useful background on which plans for increases in physical activity to manage weight can be based. Questions about the amount of time spent in sedentary activities such as television watching, using the computer, or reading also provide useful information about habitual activity levels. In addition, time spent in these sedentary activities may be available for active pursuits should the person choose to increase their physical activity level. A number of physical activity questionnaires are available to obtain more in depth information on energy expended in activities of daily living as well as planned bouts of exercise (http://​www.​health.​gov/​PAGuidelines/​). As is the case with assessing EI, there are substantial limitations to the assessment of EE by self-report. People tend to underreport food intake and over report levels of physical activity. Adults overestimate EEPA by as much as 50 % [57, 58]. A recent scientific statement from the American Heart Association provides a comprehensive guide to the tools available to assess physical activity [59].

More objective information about habitual levels of physical activity can be obtained through the use of physical activity monitoring systems. The simplest of these is the pedometer or step counter. These devices are worn at the waist and count the number of steps accumulated over a day or week [60, 61]. A pedometer can be purchased for $10–$30 and can be used to characterize an individual as sedentary (2–5,000 steps/day), normal activity (5–8,000 steps/day), meeting guidelines for PA at a level to prevent weight gain (8–11,000 steps/day), highly active or active at a level commensurate with that needed to produce and maintain weight loss (11–15,000 steps/day). Pedometers have limitations. Some cheaper models may be inaccurate, and accuracy may be reduced in obese individuals due to difficulties in keeping the device in a proper vertical alignment when worn on the belt and reduced sensitivity with slow walking speeds. Like dietary self-monitoring, physical activity self-monitoring using either a pedometer or minutes of moderate physical activity per week is valuable not only in assessing the causes of weight gain, but for laying a foundation for subsequent interventions [62, 63]. Over the last few years a large number of new physical activity monitoring systems have emerged for the consumer market. These devices cost about $100–$200 and provide data that some patients find more helpful than that provided by a typical pedometer. While the field is moving rapidly some of the leaders in this market include several Fitbit devices, the Nike Fuel, the Jawbone, and several devices from BodyMedia. These devices and others under development combine measures of movement in space with other physiological measures such as heart rate, skin temperature and galvanic skin response to estimate EEPA in free living individuals. Many of these devices interface with computer software packages that allow the tracking of specific activities at specific times of day, logging of activities over time and even the potential to provide data to personal trainers or healthcare providers. A number of other devices including the Actical (Philips), ActiGraph (Actigraph Corp.), ActivPAL (PAL Technologies Ltd) and the RT6 (Stayhealthy) have been used in research settings and have been well validated [64]. However, these systems tend to be more costly and complex requiring specialized software for analysis making them much less user friendly than the devices designed for the consumer market.

Another tool that can be used clinically to measuring energy expenditure is indirect calorimetry. The indirect calorimeter measures air flow and the difference in the concentration of oxygen between inspired and expired air to determine oxygen consumption, which is then used to calculate energy expenditure in kcal/h. When measured in the resting state, indirect calorimetry gives an estimate of REE/RMR that can be used to estimate daily energy requirements. For most people, total daily energy expenditure (which equals daily energy intake for weight maintenance) is roughly 1.3–1.5 times RMR. A number of indirect calorimetry systems are commercially available to consumers and healthcare providers for the measurement of RMR. These MedGem/BodyGem products [65] (Microlife Medical Home Solutions), the Reevue indirect calorimeter (Korr Medical Technologies Inc.) and several instruments manufactured by the Cosmed Pulmonary Function Equipment company to name just a few [66]. It is not clear how accurate these devices are in real clinical environments. The primary role for these devices at this time is to provide patients with some objective information about their energy intake needs. Many patients believe that they have a “low metabolic rate” and devices like these can provide direct evidence of what their metabolic rate is.

Weight gain associated with the introduction of medications to treat comorbid illnesses is a common problem. The most commonly implicated medications include anti-diabetic medications [67] (sulfonylureas, thiazolidinediones, insulin) as well as a wide range of psychotropic medications. The antipsychotic drugs clozapine, olanzepine, risperidone, and quetiapine have all been associated with weight gain as well as abnormalities in glucose homeostasis [68]. A number of antidepressant medications including amitriptyline, mirtazapine, and some serotonin reuptake inhibitors may promote weight gain in some patients. Other drugs that are used as mood stabilizers including lithium, valproic acid, and carbamazepine and the anti-epileptic drugs valproate, carbamazepine, and gabapentin can also promote weight gain. Historically psychiatrists and neurologists have paid little attention to the weight related side effects of some of the medications that they commonly prescribe. This situation is fortunately changing, but it is still common for a patient to be placed on a psychotropic medication or an anti-epileptic medication, experience substantial weight gain without the knowledge of the provider that initially prescribed the medication.