Effect of dietary and antismoking advice on the incidence of myocardial infarction: a 16-year follow-up of the Oslo Diet and Antismoking Study after its close

BACKGROUND AND AIM

The Oslo Diet and Antismoking Study was a 5-year randomised controlled trial initiated in 1972-1973 and ended in 1977-1978, which showed that dietary change and smoking cessation reduced the incidence of coronary heart disease among high risk middle-aged men. In an extended follow-up we studied the incidence of myocardial infarction (MI) 16 years after the end of the trial in the intervention and control groups.

METHODS

The primary endpoint was the first occurrence of non-fatal and fatal MI including sudden death up to December 31 1993. Cases of fatal MI were identified by linkage to Statistics Norway using each subject's individual personal number. Cases of non-fatal MI were extracted from the hospital records. Cox proportional hazards regression models estimated relationships between changes in total cholesterol and triglyceride concentrations and smoking status and the primary endpoints up to 16 years following the end of the trial.

RESULTS

At 5 and 10 years following the end of the trial the incidence of MI among the 604 men in the intervention (I) and 628 in the control (C) group differed significantly (5-year event rate (I/C) =0.059/0.090; P=0.038 and 10-year event rate (I/C) =0.111/0.155; P=0.023), but the difference faded slowly and subsequently (P=0.069 at 16 years). The reduction in MI in the intervention group was primarily explained by the differences in total cholesterol and triglyceride concentrations between the groups.

CONCLUSION

This extended follow-up of the Oslo Diet and Antismoking Study found a prolonged benefit of the intervention lasting for at least a decade after the close of the trial. This finding is in accordance with statin and other studies showing that the effect of cholesterol lowering may be prolonged after the end of the intervention.

In vivo expansion coincident with excessive in vitro cell death within the memory subset of CD8+ T cells in HIV type 1 infection

In HIV-1-infected individuals the CD8+ T cell subset is considerably expanded. This has been shown to be caused predominantly by an increase in the number of CD8+CD28- T cells. To characterize further the subsets of CD8+ T cells, we have performed analyses of cell surface phenotype, T cell receptor Vbeta usage, and ability to survive in unstimulated cultures. CD8+CD28- T cells frequently expressed CD45RA. Nonetheless, coincident expression of CD95 (Fas) and high levels of integrins suggested that these cells were immunologically experienced. Contrary to what has been observed in CD8+CD28- T cells from uninfected individuals, a common hierarchy of Vbeta usage was found in CD8+CD28- T cells from HIV-1-infected individuals, which was adhered to by all the study participants, and which was shown to be similar to that observed within CD8+CD28+ T cells. Cells from the memory subsets of CD8+ T cells showed a high incidence of spontaneous death in unstimulated cultures, indicating that these cells have received signals for death by apoptosis in vivo. In contrast, most naive CD8+CD28+CD45RA+ cells survived. The CD8+CD28- memory subset is expanded in vivo despite evidence for coincident excessive cell death in vitro. Our results are consistent with the notion that frequent transitions occur from the memory CD8+CD28+CD45RA- T cell subset to the end-stage CD8+CD28- subset during HIV-1 infection.

Strict adherence to a common rank order of T-cell receptor V beta usage in human leucocyte antigen disparate individuals

In order to investigate the T-cell receptor (TCR) V beta usage in different T-cell subsets, the authors performed flow cytometric analyses using a large panel of TCR V beta-specific monoclonal antibodies on CD4+, CD8+ CD28+ and CD8+ CD28- T cells from 15 random blood donors, six umbilical cords and seven human leucocyte antigen (HLA) identical non-twin sibling pairs. The authors found that the proportion of T cells expressing each V beta gene product was similar within CD4+ and CD8+ CD28+ T cells from all samples studied. For these T-cell subsets a rank order of V beta usage could be identified which was adhered to by all donors. In contrast, within CD8+ CD28- T cells a wide variation of V beta usage was found between individuals, and no rank order correlation could be detected. Members of HLA identical sibling pairs were found to be no more similar in their usage of V beta gene products than pairs of HLA disparate random blood donors. Groups of individuals sharing HLA antigens were no different from the groups not sharing such antigens in their usage of V beta segments. The results suggest that HLA polymorphisms play no more than a minor part in determining TCR V beta usage.

Expression of costimulatory molecule CD28 on T cells in human immunodeficiency virus type 1 infection: functional and clinical correlations

To study the functional integrity of T cells from human immunodeficiency virus type 1 (HIV-1)-infected persons, CD4+ and CD8+ cells were examined for proliferation and secretion of interleukin-2 (IL-2) in response to staphylococcal superantigens and antibodies to CD3 and the alpha beta T cell receptor. A functional defect within CD8+ but not within CD4+ cells from HIV-1-infected persons was observed. Within CD8+ cells, proliferation and secretion of IL-2 was restricted to cells expressing the costimulatory molecule CD28. Such cells were proportionally reduced in HIV-1-infected persons. In patients with advanced immunodeficiency, however, evidence of functional derangement was found also within the CD28+ CD8+ cells. In a cross-sectional study of 73 HIV-infected persons and 15 controls, a significant correlation was observed between the number of CD28+ CD8+ cells and the presence of HIV-related disease. Our results suggest that regulation of expression of CD28 may play an important role in the immunopathogenesis of AIDS.

Benefit of a stroke unit: a randomized controlled trial

In a randomized controlled trial we compared the clinical outcome of acute stroke patients, 110 of whom were allocated to treatment in a stroke unit and 110 to treatment in general medical wards. No significant difference existed between these groups with regard to sex, age, marital status, medical history, or functional impairment on admission. Outcome was measured at 6 and 52 weeks after the stroke by the proportion of patients at home, the proportion of patients in an institution, the mortality, and the functional state. After 6 weeks 56.4% of the patients randomized to the stroke unit and 32.7% of the patients randomized to the general medical wards were at home (p = 0.0004), and after 52 weeks 62.7% and 44.6%, respectively, were at home (p = 0.002). After 6 weeks 36.3% of the patients from the stroke unit and 50.0% from the general medical wards were in an institution (p = 0.02); after 52 weeks 12.7% and 22.7%, respectively, were institutionalized (p = 0.016). After 6 weeks mortality was 7.3% for the stroke unit group and 17.3% for the general medical wards group (p = 0.027). After 52 weeks mortality was 24.6% for the stroke unit group and 32.7% for the general medical wards group (difference not significant). Functional state was significantly better for patients treated in the stroke unit after both 6 and 52 weeks. We conclude that care of patients with acute stroke in a stroke unit improves clinical outcome compared with treatment in general medical wards.

The histopathology of non-palpable breast lesions detected by clinical mammography

An histological study of 227 non-palpable breast lesions detected by clinical mammography revealed 64 invasive carcinomas at a mean patient age of 60.5 years. There were 10 carcinomas in situ, mean age 57.2 years. Fibrocystic disease with intraductal epithelial hyperplasia was found in 41 specimens, mean age 54.3 years. Fibrocystic disease without intraductal hyperplasia was found in 57 biopsies, mean age 50.7 years. Histological microcalcifications were found in 113 biopsies, and were considered to be a marker for epithelial proliferation of both benign and malignant kinds. Microcalcifications detectable in histological sections and by mammography differ in size by a factor of 10 or more. This difference has to be considered when comparing histological and mammographic findings.