Background-Particulate air pollution has been associated with excess deaths from, and increases in hospital admissions for, cardiovascular disease among older people. A study was undertaken to determine whether this may be a consequence of alterations in the blood, secondary to pulmonary inflammation caused by the action of fine particles on alveolar cells, by repeatedly measuring haematological factors in older people and relating them to measurements of exposure to airborne particles. Methods-One hundred and twelve individuals aged 60+ years in two UK cities provided repeated blood samples over 18 months, 108 providing the maximum of 12 samples. Estimates of individual exposure to particles of less than 10 µm diameter (PM 10 ), derived from a mathematical model based on activity diaries and comparative measurements of PM 10 at multiple sites and during a variety of activities, were made for each three day period prior to blood sampling. The relationships between blood values and estimates of both personal exposure and city centre measurements of PM 10 were investigated by analysis of covariance, adjusting for city, season, temperature, and repeated individual measurements. Results-Estimated personal exposure to PM 10 over the previous three days showed negative correlations with haemoglobin concentration, packed cell volume (PCV), and red blood cell count (p<0.001), and with platelets and factor VII levels (p<0.05). The changes in red cell indices persisted after adjustment for plasma albumin in a sample of 60 of the subjects. City centre PM 10 measurements over three days also showed negative correlations with haemoglobin and red cell count (p<0.001) and with PCV and fibrinogen (p<0.05), the relationship with haemoglobin persisting after adjustment for albumin. C reactive protein levels showed a positive association with city centre measurements of PM 10 (p<0.01). Based on a linear relationship, the estimated change in haemoglobin associated with an alteration in particle concentration of 100 µg/m 3 is estimated to have been 0.44 g/ dl (95% CI 0.62 to 0.26) for personal PM 10 and 0.73 g/dl (95% CI 1.11 to 0.36) for city centre PM 10 measurements. Conclusions-This investigation is the first to estimate personal exposures to PM 10 and to demonstrate associations between haematological indices and air pollution. The changes in haemoglobin adjusted for albumin suggest that inhalation of some component of PM 10 may cause sequestration of red cells in the circulation. We propose that an action of such particles either on lung endothelial cells or on erythrocytes themselves may be responsible for changing red cell adhesiveness. Peripheral sequestration of red cells oVers an explanation for the observed cardiovascular eVects of particulate air pollution.

Among babies born at term, low birthweight predicts cardiovascular risk factors and disease in adulthood. This study shows that babies born prematurely, whether or not they have intrauterine growth retardation, are predisposed to similar risks as adults.

Background. Current methods to predict survival duration of patients with pancreatic cancer are limited. The aim of this study was to determine whether certain nutritional indices and the acute‐phase protein response are prognostic factors independent of disease stage for patients with unresectable pancreatic cancer. Methods. Variables at the time of diagnosis of 102 patients with unresectable pancreatic cancer were entered into a Cox's proportional hazards model. Included in the analysis were the serum concentration of C‐reactive protein (CRP) and albumin, the extent of weight loss, age, sex, and disease stage (International Union Against Cancer criteria). Results. A multivariate analysis in which each factor was adjusted for the influence of the other factors revealed the patient age, disease stage, serum albumin, and serum CRP to be independent predictors of survival. The presence of an acute‐phase protein response was the most significant independent predictor of survival duration. The median survival of those with an acute‐phase protein response (CRP > 10 mg/L, n = 45) was 66 days compared with 222 days for those with no acute‐phase protein response (n = 57, P = 0.001, Mann‐Whitney U test). Conclusion. The acute‐phase protein response is a useful prognostic indicator for patients with unresectable pancreatic cancer. Moreover, the metabolic disturbances associated with an acute‐phase protein response of patients with pancreatic cancer may be a worthwhile therapeutic target.