Obesity can be defined in absolute or relative terms. In practical settings, obesity is typically evaluated in absolute terms by measuring BMI (body mass index), but also in terms of its distribution through waist circumference or waist-hip circumference ratio measurements. In addition, the presence of obesity needs to be regarded in the context of other risk factors and comorbidities (other medical conditions that could influence risk of complications).
BMI, or body mass index, is a simple and widely used method for estimating body fat. In epidemiology BMI alone is used as an indicator of prevalence and incidence.
BMI was developed by the Belgian statistician and anthropometrist Adolphe Quetelet. It is calculated by dividing the subject's weight by the square of his/her height, typically expressed either in metric or US "Customary" units:
Metric: BMI = kg / m2
Where kg is the subject's weight in kilograms and m is the subject's height in metres.
US/Customary: BMI = lb * 703 / in2
Where lb is the subject's weight in pounds and in is the subject's height in inches.
The current definitions commonly in use establish the following values, agreed in 1997 and published in 2000:
* A BMI less than 18.5 is underweight
* A BMI of 18.5–24.9 is normal weight
* A BMI of 25.0–29.9 is overweight
* A BMI of 30.0–39.9 is obese
* A BMI of 40.0 or higher is severely (or morbidly) obese
* A BMI of 35.0 or higher in the presence of at least one other significant comorbidity is also classified by some bodies as morbid obesity.
In a clinical setting, physicians take into account race, ethnicity, lean mass (muscularity), age, sex, and other factors which can affect the interpretation of BMI. BMI overestimates body fat in persons who are very muscular, and it can underestimate body fat in persons who have lost body mass (e.g. many elderly). Mild obesity as defined by BMI alone is not a cardiac risk factor, and hence BMI cannot be used as a sole clinical and epidemiological predictor of cardiovascular health.
BMI does not take into account differing ratios of adipose to lean tissue; nor does it distinguish between differing forms of adiposity, some of which may correlate more closely with cardiovascular risk. Increasing understanding of the biology of different forms of adipose tissue has shown that visceral fat or central obesity (male-type or apple-type obesity) has a much stronger correlation, particularly with cardiovascular disease, than the BMI alone.
The absolute waist circumference (>102 cm in men and >88 cm in women) or waist-hip ratio (>0.9 for men and >0.85 for women) are both used as measures of central obesity.
Body fat measurement
An alternative way to determine obesity is to assess percent body fat. Doctors and scientists generally agree that men with more than 25% body fat and women with more than 30% body fat are obese. However, it is difficult to measure body fat precisely. The most accepted method has been to weigh a person underwater, but underwater weighing is a procedure limited to laboratories with special equipment. Two simpler methods for measuring body fat are the skinfold test, in which a pinch of skin is precisely measured to determine the thickness of the subcutaneous fat layer; or bioelectrical impedance analysis, usually only carried out at specialist clinics. Their routine use is discouraged.
Other measurements of body fat include computed tomography (CT/CAT scan), magnetic resonance imaging (MRI/NMR), and dual energy X-ray absorptiometry (DXA).
Risk factors and comorbidities
The presence of risk factors and diseases associated with obesity are also used to establish a clinical diagnosis. Coronary heart disease, type 2 diabetes, and sleep apnea are possible life-threatening risk factors that would indicate clinical treatment of obesity. Smoking, hypertension, age and family history are other risk factors that may indicate treatment.
Effects on health
A large number of medical conditions have been associated with obesity. Health consequences are categorised as being the result of either increased fat mass (osteoarthritis, obstructive sleep apnea, social stigma) or increased number of fat cells (diabetes, cancer, cardiovascular disease, non-alcoholic fatty liver disease). Mortality is increased in obesity, with a BMI of over 32 being associated with a doubled risk of death. There are alterations in the body's response to insulin (insulin resistance), a proinflammatory state and an increased tendency to thrombosis (prothrombotic state).
Disease associations may be dependent or independent of the distribution of adipose tissue. Central obesity (male-type or waist-predominant obesity, characterised by a high waist-hip ratio), is an important risk factor for the metabolic syndrome, the clustering of a number of diseases and risk factors that heavily predispose for cardiovascular disease. These are diabetes mellitus type 2, high blood pressure, high blood cholesterol, and triglyceride levels (combined hyperlipidemia).
Apart from the metabolic syndrome, obesity is also correlated with a variety of other complications. For some of these complaints, it has not been clearly established to what extent they are caused directly by obesity itself, or have some other cause (such as limited exercise) that causes obesity as well.
* Cardiovascular: congestive heart failure, enlarged heart and its associated arrhythmias and dizziness, cor pulmonale, varicose veins, and pulmonary embolism
* Endocrine: polycystic ovarian syndrome (PCOS), menstrual disorders, and infertility
* Gastrointestinal: gastroesophageal reflux disease (GERD), fatty liver disease, cholelithiasis (gallstones), hernia, and colorectal cancer
* Renal and genitourinary: erectile dysfunction, urinary incontinence, chronic renal failure, hypogonadism (male), breast cancer (female), uterine cancer (female), stillbirth
* Integument (skin and appendages): stretch marks, acanthosis nigricans, lymphedema, cellulitis, carbuncles, intertrigo
* Musculoskeletal: hyperuricemia (which predisposes to gout), immobility, osteoarthritis, low back pain
* Neurologic: stroke, meralgia paresthetica, headache, carpal tunnel syndrome, dementia
* Respiratory: dyspnea, obstructive sleep apnea, hypoventilation syndrome, Pickwickian syndrome, asthma
* Psychological: Depression, low self esteem, body dysmorphic disorder, social stigmatization
While being severely obese has many health ramifications, those who are somewhat overweight face little increased mortality or morbidity. Osteoporosis is known to occur less in slightly overweight people.
Causes and mechanisms
Most researchers have concluded that the combination of an excessive nutrient intake and a sedentary lifestyle are the main cause for the rapid acceleration of obesity in Western society in the last quarter of the 20th century. 
Despite the widespread availability of nutritional information in schools, doctors' offices, on the internet and on groceries, it is evident that overeating remains a substantial problem. For instance, reliance on energy-dense fast-food meals tripled between 1977 and 1995, and calorie intake quadrupled over the same period.
However, dietary intake in itself is insufficient to explain the phenomenal rise in levels of obesity in much of the industrialized world during recent years. An increasingly sedentary lifestyle also has a significant role to play. More and more research into child obesity, for example, links such things as the school run, with the current high levels of this disease. 
Less well established life style issues which may influence obesity include a stressful mentality and insufficient sleep.
As with many medical conditions, the calorific imbalance that results in obesity often develops from a combination of genetic and environmental factors. Polymorphisms in various genes controlling appetite, metabolism, and adipokine release predispose to obesity, but the condition requires availability of sufficient calories, and possibly other factors, to develop fully. Various genetic abnormalities that predispose to obesity have been identified (such as Prader-Willi syndrome, Bardet-Biedl syndrome, MOMO syndrome, leptin receptor mutations and melanocortin receptor mutations), but known single-locus mutations have been found in only about 5% of obese individuals. While it is thought that a large proportion of the causative genes are still to be identified, much obesity is likely the result of interactions between multiple genes, and non-genetic factors are likely also important.
A 2007 study identified fairly common mutations in the FTO gene; heterozygotes had a 30% increased risk of obesity, while homozygotes faced a 70% increased risk.
On a population level, the thrifty gene hypothesis postulates that certain ethnic groups may be more prone to obesity than others, and the ability to take advantage of rare periods of abundance and use such abundance by storing energy efficiently may have been an evolutionary advantage in times when food was scarce. Individuals with greater adipose reserves were more likely to survive famine. This tendency to store fat is likely maladaptive in a society with stable food supplies.
Certain physical and mental illnesses and particular pharmaceutical substances may predispose to obesity. Apart from the fact that correcting these situations may improve the obesity, the presence of increased body weight may complicate the management of others.
Medical illnesses that increase obesity risk include several rare congenital syndromes (listed above), hypothyroidism, Cushing's syndrome, growth hormone deficiency. Smoking cessation is a known cause for moderate weight gain, as nicotine suppresses appetite. Certain medications (e.g. steroids, atypical antipsychotics, some fertility medication) may cause weight gain.
Mental illnesses may also increase obesity risk, specifically some eating disorders (bulimia nervosa and binge eating disorder).