1. Six elderly (66-71 years) and six young (20-23 years) subjects (half of each group women) were cooled for 2 h in moving air at 18°C to investigate possible causes of increased mortality from arterial thrombosis among elderly people in cold weather. Compared with thermoneutral control experiments, skin temperature (trunk) fell from 35.5 to 29.5°C, with little change in core temperature. 2. Erythrocyte count rose in the cold from 4.29 to 4.69 × 10 12 /l, without a change in mean corpuscular volume, indicating a 14% or 438 ml decline in plasma volume; increased excretion of water, Na + and K + accounted for loss of only 179 ml of extracellular water. 3. Plasma cholesterol and fibrinogen concentrations rose in the elderly subjects from 4.9 mmol/l and 2.97 g/l (control) to 5.45 mmol/l and 3.39 g/l in the cold, and in the young subjects from 3.33 mmol/l and 1.84 g/l (control) to 3.77 mmol/l and 2.07 g/l in the cold. Increases were significant for the elderly subjects, the young subjects and the group as a whole, except for cholesterol in the young subjects, and all were close to those expected from the fall in plasma volume. 4. Plasma levels of Protein C and factor X did not increase significantly in the cold in the elderly subjects, young subjects, or the group as a whole. 5. The results suggest that loss of plasma fluid in the cold concentrates major risk factors for arterial thrombosis, while small molecules, including protective Protein C, redistribute to interstitial fluid.

1. Changes in serum total and lipoprotein fraction triglyceride and cholesterol levels were studied in 24 adults on home haemodialysis. Half the patients were randomly allocated to a low cholesterol (mean 200 mg/day), fat-modified diet (mean polyunsaturated/saturated fat ratio of 1.0 with a mean of 43% of the total energy content derived from fat). 2. Before dietary manipulation, triglyceride levels in all lipoprotein fractions were significantly higher ( P < 0.02) than in a control group of age and sex matched normal subjects. Total cholesterol, very-low-density-lipoprotein (VLDL) and low-density-lipoprotein (LDL) cholesterol were also significantly raised ( P < 0.02), but high-density-lipoprotein (HDL) cholesterol was normal. In the patients on a fat-modified diet triglyceride levels did not alter in any of the lipoprotein fractions. Total cholesterol and LDL cholesterol levels fell significantly into the normal range ( P < 0.002 and < 0.001 respectively) but VLDL and HDL cholesterol levels did not change. 3. Hypertriglyceridaemia is the most common lipid abnormality in patients with renal failure and a long-term fat-modified diet is, therefore, of limited therapeutic importance in these patients unless there is a low HDL/LDL cholesterol ratio.