Complex interactions occur among adipose tissue, the central nervous system, bone and pancreas to integrate bone remodelling, glucose, lipid and energy metabolism. Data obtained largely from the judicious use of gain-of-function and loss-of-function genetic mouse models show that leptin, an adipocyte-secreted product, indirectly inhibits bone accrual through a central pathway comprising the hypothalamus and central nervous system. Increased sympathetic output acting via β2-adrenergic receptors present in osteoblasts decreases bone formation and causes increased bone resorption. Insulin is a key molecular link between bone remodelling and energy metabolism. Insulin signalling in the osteoblasts increases bone formation and resorption as well as the release of undercarboxylated osteocalcin. An increase in the release of bone-derived undercarboxylated osteocalcin into the systemic circulation enables it to act as a circulating hormone to stimulate insulin production and secretion by pancreatic β-cells and adiponectin by adipocytes. Insulin sensitivity increases, lipolysis and fat accumulation decreases while energy expenditure increases. Whether this model of integrative physiology involving the skeleton, pancreas and adipose tissue, so elegantly demonstrated in rodents, is applicable to humans is controversial. The mouse Esp gene, encoding an intracellular tyrosine phosphatase that negatively regulates insulin signalling in osteoblasts, is a pseudogene in humans, and a homolog for the Esp gene has so far not been identified in humans. A close homologue of Esp, PTP1B, is expressed in human osteoblasts and could take the role of Esp in humans. Data available from the limited number of clinical studies do not provide a sufficient body of evidence to determine whether osteocalcin or undercarboxylated osteocalcin affects glucose metabolism in humans.