We sought to establish the extent and basis for adaptive changes in beta cell numbers in human pregnancy. Pancreas was obtained at autopsy from women who had died while pregnant (n = 18), post-partum (n = 6) or were not pregnant at or shortly before death (controls; n = 20). Pancreases were evaluated for fractional pancreatic beta cell area, islet size and islet fraction of beta cells, beta cell replication (Ki67) and apoptosis (TUNEL), and indirect markers of beta cell neogenesis (insulin-positive cells in ducts and scattered beta cells in pancreas). The pancreatic fractional beta cell area was increased by ∼1.4-fold in human pregnancy, with no change in mean beta cell size. In pregnancy there were more small islets rather than an increase in islet size or beta cells per islet. No increase in beta cell replication or change in beta cell apoptosis was detected, but duct cells positive for insulin and scattered beta cells were increased with pregnancy. The adaptive increase in beta cell numbers in human pregnancy is not as great as in most reports in rodents. This increase in humans is achieved by increased numbers of beta cells in apparently new small islets, rather than duplication of beta cells in existing islets, which is characteristic of pregnancy in rodents.

We sought to establish the relationship between fasting plasma glucose concentrations and pancreatic fractional beta cell area in adult cynomolgus monkeys (Macaca fascicularis). Fasting plasma glucose and pancreatic fractional beta cell area were measured in 18 control and 17 streptozotocin-treated adult primates (17.0 ± 1.2 vs 15.4 ± 1.2 years old). Fasting plasma glucose was increased (12.0 ± 2.0 vs 3.4 ± 0.1 mmol/l, p < 0.01) and fractional beta cell area was decreased (0.62 ± 0.13% vs 2.49 ± 0.35%, p < 0.01) in streptozotocin-treated monkeys. The relationship between fasting plasma glucose and pancreatic fractional beta cell area was described by a wide range of beta cell areas in controls. In streptozotocin-treated monkeys there was an inflection of fasting blood glucose at ∼50% of the mean beta cell area in controls with a steep increase in blood glucose for each further decrement in beta cell area. In adult non-human primates a decrement in fractional beta cell area of ∼50% or more leads to loss of glycaemic control.