Cardiovascular disease (CVD) is the most common cause of total mortality and premature death (before 75 years of age) in Sweden as well as in other Western countries [1] [2] . However during the period 1997 up to 2012 premature mortality in myocardial infarction has decreased with approximately 50 % in Sweden and most western countries [3] . According to the INTERHEART study, more than 90% of the risk for acute myocardial infarction could be explained by nine modifiable risk factors, most of them being lifestyle factors [4] . The same factors in addition to cardiac disease were associated to 90% of stroke [5] . Lifestyle is also related to risk for cancer [6] [7] . In a Swedish study non-smoking, moderate physical activity and minor alcohol consumption together with four healthy dietary habits could identify 60-year-old men and women with substantially lower risk for CVD and death [8] .

Some kind of global risk assessment is recommended to estimate the risk for future death in CVD. SCORE (Systematic Coronary Risk Evaluation) is one such risk assessment system, developed by the European Society of Cardiology and modified for use in Sweden [9] [10] .

Another kind of risk assessment tool, a pedagogic health profile, was developed and used in a primary prevention programme for men, 33 - 42 years of age, in the community of Habo in Sweden [11] . Each of 10 risk factors considered being most important for CVD and cancer was graded in 4 or 5 levels. This health profile was used in a health dialogue with a specially trained nurse. In that study a rather thorough examination of the non- participants was done [11] .

An improved health profile called “Health Curve” was later used in a health promotion programme for both men and women 30 and 35 years of age called “Live for Life” where community efforts were combined with a health dialogue with a nurse [12] . In the Health Curve dietary habits concerning fat and fiber intake were measured using a pedagogic food frequency questionnaire [13] .

Rather few studies concerning cohorts of men with a follow-up time of more than 20 years have been published. The study of British doctors, the Whitehall study and the Oslo study are such examples [14] - [16] . The 50-year follow-up of the male British doctors showed a progressive decrease in mortality rates among non-smokers in relation to continuous smokers, and that smoking cessation in different ages could add years to life expectancy [14] . On group level the continuous smokers lost 10 years of life. In the Whitehall study with 38-year follow-up time, there was a risk core based on smoking, diabetes, employment grade, blood pressure, serum cholesterol, and Body Mass Index (BMI) Men in the highest 5% risk score had a 15 year shorter life expectancy compared with those in the lowest 5% risk score [15] . The Oslo study was a primary prevention intervention study of men focused on lifestyle, especially smoking and dietary habits. The 28-year follow-up showed a decrease of leisure time physical activity (LTPA), and that LTPA had an independent predictive association with metabolic syndrome [16] .

In studies with long-term follow-up the study population is generally 40 years and older at baseline. There are some examples of studies where the study population is younger than 40 years at baseline [17] - [19] . The 21-year follow-up of the Cardiovascular Risk in Young Finns Study showed an increase in BMI, serum triglycerides and a decrease in blood pressure and serum cholesterol [17] . The men in Chicago Heart Association Detection Project in Industry study were 18 to 39 years at baseline [18] . In this study blood pressure above normal was significantly related to increased mortality due to CVD and all causes. The men in three large cohorts study were also 18 to 39 years at baseline [19] . The results from this study could show a graded relationship of serum cholesterol to long- term risk of CVD and all-cause mortality.

To our knowledge no studies on a cohort of men as young as in our study have been organized and carried out in primary care with a follow-up time of more than 20 years.

This 24-year follow-up survey and register study was done on all men 33 - 42 years old at baseline in the community of Habo in Sweden.

There are four aims of this study:

・ To study if a long-term follow-up of men in a rather young age group at baseline, can be organized and carried out in primary health care with a high participation rate.

・ To measure long-time changes, after approximately 24 years of follow-up, in lifestyle and biological risk markers for men taking part in the primary prevention programme in the community of Habo between 1985 and 1987.

・ To study the association between food habits and a healthy lifestyle concerning physical activity, smoking and alcohol consumption, and the association between self-rated health and the different risk groups.

・ To study morbidity from CVD, cancer and total mortality, on defined high- and low risk groups and non- participants in the whole baseline cohort of 757 men.

The overall participation rate of 83% of those who were invited to a new health dialogue and health examination was high. In the 28-year follow-up of the Oslo study the participation was 62% [16] . The men in the Oslo study were 40 - 49 years old at baseline. Only men living in Oslo and Akershus were invited to follow-up. It is comparable to our study where only men still living in Habo and neighboring communities were invited.

The most positive change in lifestyle factors was smoking cessation with a decrease from 27% to 10%. The decreasing proportion of smokers is similar to the national Swedish trend and figures from Statistics Sweden, although the prevalence of 10% daily smokers at follow-up in our study is lower. In Sweden in 1985 36% of men in age group 36 - 40 years were daily smokers, and in 2010-2011 18% of men in age group 55 - 64 years were daily smokers. In comparison with the initial examination, alcohol consumption had increased from a rather low level and LTPA had decreased. Self-rated health has in several studies been a good predictor for future morbidity and mortality [24] [25] . In our study the low-risk group and those with favorable lifestyle estimated their health more positively.

The biological risk factors, except for diastolic blood pressure had deteriorated significantly for the entire co-

^**p < 0.01, ^***p < 0.001 follow-up values compared with baseline values. Registrations for waist, hip, WHR, PEF were not complete during the first examination. Some questionnaire responders did not report their anthropometric measurements. During follow-up only those who took part in the health examination had measurements for blood pressure and serum lipids.

^*p < 0.05, ^**p < 0.01, ^***p < 0.001 values for high-risk group compared with low-risk group. Registrations for waist, hip, WHR, PEF were not complete during the first examination. Some questionnaire responders did not report their anthropometric measurements. During follow-up only those who took part in the health examination had measurements for blood pressure and serum lipids.

hort and especially for the high-risk group. The fact that serum cholesterol was not higher in the high-risk group could be explained by differences in lipid lowering medication. Medication for high serum cholesterol was used by 19% in the high-risk group and by 9% in the low-risk group. The change and deterioration of most biological risk markers is similar to men in the Whitehall study [26] . A study in Gothenburg of risk factors in a group of men 50 years old and another group of men 60 years old showed impaired anthropometric data, blood pressure, but not serum lipids among the older men [27] .

The distribution in high-risk and low-risk groups based on the health profile seemed, at group level, to be able to predict a significant higher incidence of CVD and cancer, and a non-significant higher incidence of total mortality, in the high-risk group in comparison with the low-risk group. The baseline non-participant group had a higher incidence of CVD, and a higher mortality than the low-risk group. Other studies have also shown that non-par- ticipants have a higher morbidity and mortality than participants [28] - [30] . In our earlier study of the baseline non- participant group this group could be divided into three different groups [11] . Those men who did not participate because they had already taken part in a similar health examination, those who already had frequent contacts with health care because if different diseases, and those who were not interested [11] . It is reasonable to think that the higher morbidity from CVD and mortality in the non-participant group can be linked primarily to these last two groups.

In most studies with long term follow-up the study population is 40 years and older at baseline. There are some examples where the study population is younger and single risk factors can predict CVD. In the Chicago Heart Association Detection Project in Industry blood pressure was related to mortality in CVD after 25 years [18] . In a study of three cohorts serum cholesterol had a long time relationship to CVD and all-cause mortality [19] .

The SCORE-model based on age, gender, serum cholesterol, systolic blood pressure and smoking is designed to predict risk for mortality in CVD in 10 years for people 40 to 65 years [9] . This model was compared with the Framingham risk score and WHO/ISH model in Asian population 40 - 65 years, and was considered suitable to identify high CVD risk [31] . The SCORE-model is also compared to a Swedish consultation-based method without laboratory examinations on a population 40 - 59 years with equal result [32] . An aim for a future study can be to compare our health profile with the SCORE-model.

To our knowledge there are few examples of health prediction instruments for mortality of both cancer and CVD. The computer based Real Age is one example capable to predict mortality both for cancer and CVD [33] . In Real Age people could register health habits, family history, serum cholesterol, blood pressure and several psychosocial data. The data used in the Real Age register are similar to those in our health profile.

The strength in our study is the long follow-up time as well as the high participation rate among those who were invited. The limitations are lack of values for blood tests and blood pressures for those who were not invited, who also self-reported their own measurements of weight, height, and waist and hip circumferences. These data were thus not standardized according to the measurements that the nurses performed of the participants at the health care center. In spite of this we consider the self-reported data useful, and can be included in the data analyses. The Real Age register is also based on self-reported data [33] . There is always a risk for participation bias in follow-up studies. The high participation rate in our study and that those who were not invited were sent a postal questionnaire and asked to measure their weight, length, waist and hip circumferences are assumed to reduce the risk for participation bias. Another method to improve participation rate is to offer home visits to non-participants, as was done in the 32-year follow-up of the population study of women in Gothenburg [34] .

In conclusion this study shows that it was possible to organize and carry out a long-term follow-up study, of this at baseline relative young group of men, in primary health care with a high participation rate. The categorization of the participants in high-risk and low-risk groups from values in the health profile could predict the incidence of CVD and cancer after 24 years of follow-up. Concerning lifestyle factors smoking and LTPA had decreased. Alcohol consumption had increased but from a relatively low level. Biological risk markers, except diastolic blood pressure, deteriorated significantly with age.

This study was supported by grants from the Medical Research Council of Southeast Sweden (FORSS) and Futurum County Council of Jönköping Sweden.

The study was approved by the Regional Ethical Review Board in Linköping (Dnr M167-09).

The authors report no conflict of interest. The authors alone are responsible for the content and writing of the paper.