The design and methods of the Oslo Study of 1972/3 have been described earlier [20
]. In brief, all men born in 1923–32 and residing in the city of Oslo were invited to a screening examination for cardiovascular diseases and risk factors. A total of 16 209 men aged 40–49 years at baseline attended, which represented 63% of this age group. Height, weight, and blood pressure were measured and a blood sample taken in the non-fasting state was used for measurements of total serum cholesterol, triglycerides and glucose. Time since the last meal was recorded. Participants filled in a questionnaire regarding prevalent diseases and symptoms of cardiovascular origin and diabetes, smoking habits, degree of physical activity at work and at leisure and a few questions about mental stress. LTPA was defined in four groups as follows: sedentary/light: usually reading, watching television or other sedentary occupations at leisure; moderate: walking, bicycling or other forms of physical activity including walking or bicycling to and from the place of work and a Sunday walk totalling at least four hours a week; moderately vigorous: exercise, sports, heavy gardening and similar activities totalling at least 4 hours a week; vigorous: hard training or competition sports regularly several times a week. The respondent was asked to use the average amount of activity, if activity varied much for example between summer and winter. This question has been found to show solid correlation to physical fitness and have good predictive validity in prospective studies [21
In 2000–2001 the conductance of the Oslo Health Study 2000–2001 (abbreviated as HUBRO), a population-based survey of selected birth cohorts living in Oslo in 2000–2001 [24
] took place. Additional to that study all men originally invited to the Oslo Study in 1972/3 and resident in Oslo or the neighbouring county of Akershus were invited to a repeat screening examination (designated Oslo II)[25
]. Height, weight, and blood pressure were measured again and non-fasting blood samples were taken for measurements of serum total cholesterol, triglycerides and glucose. This time high density cholesterol (HDL) cholesterol was also included. Time since the last meal was recorded. Attendees filled in two questionnaires covering smoking habits and the same questions concerning LTPA as they did in 1972/3 as well as a number of other issues including educational attendance in number of years.
Some of the participants who met the criteria for inclusion in Oslo II were invited to the screening through HUBRO while others were invited through another study promoting physical activity in the community [26
]. Both studies used the same screening procedures and questionnaires as Oslo II. In total 1 095 men from these studies participated and their data were later included in the Oslo II study database. Moreover 813 eligible subjects for Oslo II were participating in three ongoing clinical trials. These men were invited to fill in the same questionnaires that were used in Oslo II at the close of the trials and they provided fasting blood sample values which accordingly were not adjusted to eight hours since last meal. The Norwegian Data Inspectorate allowed the data and the measurements in these five studies to be added to the Oslo II database.
Men who were dead or had emigrated (n = 1 655), men living outside the catchment area of Oslo and Akershus (n = 1 278) and men with unknown addresses (n = 2 944) were excluded from the Oslo II screening leaving 10 328 candidates for the study. Of these 6 410 of the men who attended the baseline study, also attended the follow up, resulting in an attendance proportion of 62.0%. Finally 6 382 men without reported diabetes in 1972/3 (n = 28 had diabetes or non fasting glucose ≥ 11.1 mmol/l) and with systolic blood pressure measurements both in 1972/3 and in 2000 constituted the population for analysis.
We based the definition of the presence or absence of the metabolic syndrome in year 2000 based on a modification of the National Cholesterol Education Program Adult Treatment Program III criteria (NCEP III) [27
]. Because of the unavailability of waist circumference, BMI was used to replace it with a cut-off of a BMI ≥30.0 kg/m2
corresponding to about a waist circumference of 102 cm. Furthermore fasting and non-fasting glucose levels were adjusted to correspond to eight hours since last meal as the criteria require fasting levels. Triglyceride levels were also adjusted to eight hours since the last meal. In summary the five criteria of the metabolic syndrome were defined as follows: triglycerides ≥1.7 mmol/l adjusted for the last meal, glucose ≥6.1 mmol/l adjusted for the last meal, BMI ≥30.0 kg/m2
, blood pressure ≥135/85 mmHg (use of antihypertensive medication was not included in the hypertension definition), and HDL cholesterol <1.03 mmol/l. The metabolic syndrome was defined as the presence of at least three out of five criteria. We could not define MS in 1972/3 according to NCEP III criteria because we lacked measurements of HDL-C.
Smoking was categorized as never (reference), previous and current. LTPA was grouped as the same four groups at both screenings. In year 2000 the definition of diabetes included self reported diabetes, men who took oral antidiabetic agents, used insulin or had a non fasting glucose ≥ 11.1 mmol/l.
Ethics and approvals
All the participants of the Oslo Study have given their written signed consent. The Norwegian Data Inspectorate has approved the Oslo Study, it has been cleared by the Regional Committee for Medical Research Ethics and it has been conducted in full accordance with the World Medical Association Declaration of Helsinki.
Because blood samples were non-fasting (except for subjects that participated in one of the three randomised trials mentioned above), second order regression equations were fitted between level of triglycerides and of glucose and time since last meal. The values were adjusted to eight hours after the last meal. Odds ratios (OR) with 95% confidence limits were calculated by logistic regression analyses. The metabolic syndrome in 2000 and diabetes in 2000 were dependent variables. Age, length of education in years, smoking, LTPA, levels of glucose and triglycerides, body mass index and systolic blood pressure were independent variables. Likelihood ratio tests were used to test the significance of the factors. Test of trend for LTPAwas done assuming interval scale of the variable. In 2000 the question regarding LTPA was answered by only 66% of attendees since HUBRO did not include this question in its survey. Thus the comparison of LTPA in 2000 to 1972/3 had correspondingly fewer data points.
Spearman's correlation coefficient was used to estimate the relationship of LTPA in 2000 to LTPA in 1972/3 while Pearsons' correlation coefficient was used for the relationship of BMI in 2000 to BMI in 1972/3. A Chi square test was used to test for changes in smoking habits. The SPSS 13.0 software program was used for all analyses.