Obesity (BMI C30 kg/m 2 ) increases the risk of developing lifestyle-related diseases. A subgroup of obese individuals has been described as ''metabolically healthy, but obese'' (MHO). In contrast to at-risk obese (ARO), the MHO phenotype is defined by a favourable lipid profile and a normal or only slightly affected insulin sensitivity, despite the same amount of body fat. The objective was to characterize the metabolic phenotype of MHO subjects. We screened a variety of genes involved in lipid metabolism and inflammation in peripheral blood mononuclear cells (PBMC). Obese subjects (men and women; 18-70 years) with BMI C30 kg/m 2 were characterized as MHO (n = 9) or as ARO (n = 10). In addition, eleven healthy, normal weight subjects characterized as healthy by the same criteria as described for the MHO subjects were included. We found that with similar weight, total fat mass and fat mass distribution, the ARO subjects have increased plasma levels of gamma-glutamyl transpeptidase and free fatty acids. This group also has altered expression levels of a number of genes linked to lipid metabolism in PBMC with reduced gene expression levels of uncoupling protein 2, hormone-sensitive lipase and peroxisome proliferatoractivated receptor d compared with MHO subjects. The present metabolic differences between subgroups of obese subjects may contribute to explain some of the underlying mechanisms causing the increased risk of disease among ARO subjects compared with MHO subjects.

Chronic systemic inflammatory response is proposed as an underlying mechanism for development of cancer cachexia. We conducted a prospective study to examine changes in inflammatory biomarkers during the disease course and the relationship between inflammatory biomarkers and cachexia in patients with inoperable pancreatic cancer. Twenty patients, median (range) age 67.5 (35-79) years, 5 females, were followed for median 5.5 (1-12) months. Cachexia was diagnosed according to the 2011 consensus-based classification system (weight loss >5 % past six months, BMI < 20 kg/m(2) and weight loss >2 %, or sarcopenia) and the modified Glasgow Prognostic score (mGPS) that combines CRP and albumin levels. Inflammatory biomarkers were measured by enzyme immunoassays. The patients had increased levels of most inflammatory biomarkers, albeit not all statistically significant, both at study entry and close to death, indicating ongoing inflammation. According to the consensus-based classification system, eleven (55 %) patients were classified as cachectic upon inclusion. They did not differ from non-cachectic patients with regard to inflammatory biomarkers or energy intake. According to the mGPS, seven (35 %) were defined as cachectic and had a higher IL-6 (p < 0.001) than the non-cachectic patients. They also had a slightly, but insignificantly longer survival than non-cachectic patients (p = 0.08). The mGPS should be considered as an additional framework for identification of cancer cachexia.

There is no universally accepted definition of cancer cachexia. Two classifications have been proposed; the 3-factor classification requiring ≥ 2 of 3 factors; weight loss ≥ 10%, food intake ≤ 1500 kcal/day, and C-reactive protein ≥ 10 mg/l, and the consensus classification requiring weight loss >5% the past 6 mo, or body mass index <20 kg/m(2) or sarcopenia, both with ongoing weight loss >2%. Precachexia is the initial stage of the cachexia trajectory, identified by weight loss ≤ 5%, anorexia and metabolic change. We examined the consistency between the 2 classifications, and their association with survival in a palliative cohort of 45 (25 men, median age of 72 yr, range 35-89) unresected pancreatic cancer patients. Computed tomography images were used to determine sarcopenia. Height/weight/C-reactive protein and survival were extracted from medical records. Food intake was self-reported. The agreement for cachexia and noncachexia was 78% across classifications. Survival was poorer in cachexia compared to noncachexia (3-factor classification, P = 0.0052; consensus classification, P = 0.056; when precachexia was included in the consensus classification, P = 0.027). Both classifications showed a trend toward lower median survival (P < 0.05) with the presence of cachexia. In conclusion, the two classifications showed good overall agreement in defining cachectic pancreatic cancer patients, and cachexia was associated with poorer survival according to both.