The objective of this study was to investigate how different obesity measures link to circulating metabolites, and whether the connections are due to genetic or environmental factors. A cross-sectional analysis was performed on follow-up survey data at the Chinese National Twin Registry (CNTR), which was conducted in four areas of China (Shandong, Jiangsu, Zhejiang and Sichuan) in 2013. The survey collected detailed questionnaire information and conducted physical examinations, fasting blood sampling and untargeted metabolomic measurements among 439 adult twins. Linear regression models and bioinformatics analysis were used to examine the relation of obesity measures, including body mass index (BMI), waist circumference (WC) and waist-to-hip ratio (WHR) with serum metabolite levels and related pathways. A co-twin control study was additionally conducted among 15 obesity-discordant monozygotic (MZ) pairs (intrapair BMI difference >3 kg/m2) to examine any differences in metabolites controlling for genetic factors. Eleven metabolites were associated with BMI, WC and WHR after controlling for genetic and shared environmental factors. Pathway analysis identified pathways such as phenylalanine metabolism, purine metabolism, valine, leucine and isoleucine biosynthesis that were associated with obesity. A wide range of unfavorable alterations in the serum metabolome was associated with obesity. Obesity-discordant twin analysis suggests that these associations are independent of genetic liability.

Residual feed intake (RFI), defined as the difference between an animal’s actual feed intake and expected feed intake over a specific period, is an inheritable character of feed conversion efficiency in dairy cows. Research has shown that a lower RFI could improve the profitability of milk production. This study explored variation in RFI by comparing the differences in body size, milk performance, feeding behavior, and serum metabolites in 29 Holstein cows in mid lactation. The cows were selected from a total of 84 animals based on their RFI following feedlot tests. Selected cows were ranked into high RFI (RFI >1 SD above the mean, n=14) and low RFI (RFI<1 SD below the mean, n=15). The low RFI cows (more efficient) consumed 1.59 kg/day less dry matter than the high RFI group (P<0.01), while they produced nearly equal 4% fat-corrected milk. The milk : feed ratio was higher for the low RFI group than for the high RFI group (P<0.05). The levels of milk protein (P<0.01), total solids (P<0.05), and nonfat solids (P<0.05) were also higher for the low RFI group, whereas milk urea nitrogen was lower (P<0.01). The daily feeding duration was shorter for the low RFI group than for the high RFI group (P<0.01). No significant differences were found in levels of glucose, β-hydroxybutyrate, prolactin, insulin, IGF-1, growth hormone or ghrelin, but the level of neuropeptide Y was higher (P<0.01) and levels of leptin and non-esterified fatty acid (P<0.05) were lower for the low RFI group than for the high RFI group. There were substantial differences between cows with different RFI, which might affect the efficiency of milk protein metabolism and fat mobilization.

Energetic efficiency was evaluated in composite bred heifers born from dams receiving 1·8 or 1·2 kg/d winter supplementation for approximately 80 d before parturition. Heifers were then developed post-weaning and randomly assigned to heifer development treatments of either control (100 %; ad libitum; n 8/year) or restricted (80 %; fed 80 % of supplementation fed to controls adjusted to a common body weight: n 8/year) in a 2-year study. A glucose tolerance test (GTT) and acetate irreversible loss test (AILT) were administered to heifers at the termination of a 140 d development period when the heifers were approximately 403 d of age and consumed a silage-based diet, and again at 940 d of age when pregnant with their second calf and grazing dormant forage. No differences were measured (P>0·08) for dam winter nutrition or heifer development treatment for baseline serum metabolites or measures in either the GTT or the AILT. However, changes in baseline serum concentrations (P>0·05) were different between metabolic challenges, which occurred at different stages of development. No difference in acetate disappearance (P = 0·18) and half-life (P = 0·66) was measured between the two metabolic challenges. A trend for glucose half-life to be shorter in heifers born from dams receiving in utero winter treatments that supplied 1·2 kg/d of winter supplementation was observed (P = 0·083). Heifers developed with lower total DM intake during a 140 d development period had similar glucose and acetate incorporation rates as ad libitum-fed heifers when evaluated at two different production stages.