The aim of the study was to evaluate a range of potential risk factors in the etiology of varicose veins by a longitudinal follow-up study in a general middle-aged population. Studied factors were parity, use of oral contraceptives and hormone replacement therapy, lifestyle factors and self-reported family history of varicose veins. The purpose was also to evaluate misclassification in self-reported family history of varicose veins.

The five-year follow-up data was obtained through postal questionnaires from three cohorts of residents of the city of Tampere in Finland. The number of original target population of 40-, 50- and 60-year-olds was 6,874. The response rate was 81% (5,568) at entry. An almost identical questionnaire was sent five years later to those who had responded to the first questionnaire and the response rate was 88% (4,903). The questionnaire included questions related to medical conditions, lifestyle, varicose veins and family history of varicose veins. A sub-sample of subjects underwent a physical examination but family members were not clinically examined. Incidence of varicose veins was studied as an indicator of risk in those free of varicose veins at entry.

It was found that new varicose veins appeared more often in women with three or more births than in nulliparous women, incidence odds ratio, IOR 2.0 (95% confidence interval, CI: 1.0 3.9). Subjects who consumed alcohol weekly had a higher incidence of varicose veins compared to non-users, IOR 1.5 (95% CI: 1.05 2.3). The result was statistically significant only in women. Compared to non-smokers, subjects who had ever smoked for over a year, and of these heavier smokers (≥15 cigarettes a day), had higher risk of varicose veins, IORs 1.3 (95% CI: 0.9 1.8) and 1.8 (95% CI: 1.1 2.8) respectively. Daily use of meat seemed to decrease the risk of varicose veins compared to infrequent use (0 2 meals a week) of meat, particularly in women. Dietary fiber intake, use of contraceptive pills or hormone replacement therapy did not have an effect on the incidence of varicose veins.

In those with a positive family history of varicose veins the risk of varicose veins was increased compared to those with negative family history, IOR 1.6 (95% CI: 1.1 2.3). When genders were studied separately, the higher risk was statistically significant only in women, IOR 1.8 (95% CI: 1.1 2.8), but not in men, IOR 1.4 (95% CI: 0.7 2.6). The estimates of the risk were much lower than our prevalence results (OR 6.6 in men and OR 4.9 in women, respectively) and also those usually reported in the literature based on cross-sectional studies with prevalence estimates.

Subject s own varicose vein status or change in it was associated with self-reported family history of varicose veins. For example, if a person who at entry reported negative family history developed varicose veins during follow-up, the risk of reporting a family member with varicose veins after the follow-up was sixfold, OR 6.0 (95% CI: 2.0 47.8). If another person reported varicose veins at entry and at the end of the follow-up, the risk of reporting a new family member with varicose veins was less than 1 (OR 0.8 (95% CI: 0.6 0.9)).

In conclusion, in women, multiparity, alcohol consumption and positive family history of varicose veins increase the risk of varicose veins, whereas meals containing meat may reduce the risk. Smoking is likely a risk factor of varicose veins in both genders. The strong effect of family history on the risk of varicose veins in the literature is affected by differential misclassification bias.