- Top of page
- Disclosure of Conflict of Interests
Summary. Background: Recent studies indicate that arterial cardiovascular diseases and venous thromboembolism (VTE) share common risk factors. A family history of myocardial infarction (MI) is a strong and independent risk factor for future MI. Objectives: The purpose of the present study was to determine the impact of cardiovascular risk factors, including family history of MI, on the incidence of VTE in a prospective, population-based study. Patients and methods: Traditional cardiovascular risk factors and family history of MI were registered in 21 330 subjects, aged 25–96 years, enrolled in the Tromsø study in 1994–95. First-lifetime VTE events during follow-up were registered up to 1 September 2007. Results: There were 327 VTE events (1.40 per 1000 person-years), 138 (42%) unprovoked, during a mean of 10.9 years of follow-up. In age- and gender-adjusted analysis, age [hazard ratio (HR) per decade, 1.97; 95% confidence interval (CI), 1.82–2.12], gender (men vs. women; HR, 1.25; 95% CI, 1.01–1.55), body mass index (BMI; HR per 3 kg m−2, 1.21; 95% CI, 1.13–1.31), and family history of MI (HR, 1.31; 95% CI, 1.04–1.65) were significantly associated with VTE. Family history of MI remained a significant risk factor for total VTE (HR, 1.27; 95% CI, 1.01–1.60) and unprovoked VTE (HR, 1.46; 95% CI, 1.03–2.07) in multivariable analysis. Blood pressure, total cholesterol, HDL-cholesterol, triglycerides, and smoking were not independently associated with total VTE. Conclusions: Family history of MI is a risk factor for both MI and VTE, and provides further evidence of a link between venous and arterial thrombosis.
Venous thromboembolism (VTE), including deep vein thrombosis (DVT) and pulmonary embolism (PE), is a common disease, with serious short- and long-term complications, and a potential fatal outcome . It affects 1–3 per 1000 adults each year [1,2], and is the third most common cardiovascular disease . Although surgery, trauma, hospitalization, malignancy, immobilization, pregnancy, use of oestrogens, and inherited thrombophilia are associated with VTE [1,4], approximately 30–50% of the events occur in the absence of obvious predisposing factors .
The concept that VTE and atherosclerotic cardiovascular diseases are two distinct disease entities has recently been challenged. A higher frequency of carotid plaques in patients with idiopathic VTE has been reported . The long-term incidence of cardiovascular disease is reported to be substantially increased in patients with VTE, compared to population controls . A recent meta-analysis, based on selected prospective studies and case–control studies with verified endpoints, concluded that cardiovascular risk factors were associated with VTE . However, prospective studies on the relation between traditional cardiovascular risk factors and VTE show diverging results. In the Nurses’ Health Study, obesity, cigarette smoking, and hypertension, but not diabetes or elevated cholesterol, were independent predictors for PE . In the Physicians’ Health Study , only body mass index (BMI) and height were identified as independent risk factors for VTE, whereas smoking and abdominal obesity were predictors for VTE in another study on middle-aged men . In the LITE-study, obesity and diabetes, but not cigarette smoking, hypertension, and dyslipidemia, showed independent association to VTE . In the Copenhagen City Heart Study, total cholesterol, HDL-cholesterol (inverse), diabetes, and smoking were associated with VTE . These studies suggest that VTE and cardiovascular disorders may share some common risk factors, and that VTE may occur as the first cardiovascular event in some patients at risk for atherosclerosis.
A family history of myocardial infarction (MI) is a strong and independent risk factor for future MI  and death from cardiovascular disease . Family history of MI had synergistic interactions on risk for cardiovascular disease, together with smoking and dyslipidemia [15,16]. To the best of our knowledge, no study has investigated family history of MI as a risk factor for VTE. To address this question, we performed a prospective, population-based study and assessed the impact of traditional cardiovascular risk factors, including family history of MI, on the incidence of VTE.
- Top of page
- Disclosure of Conflict of Interests
Mean age was 45 ± 14 years, and 11 212 (52.6%) were women. There were 327 validated, first VTE events during 233 089 person-years of follow-up (mean 10.9 years). The overall crude incidence rate of VTE was 1.40 per 1000 person-years. The incidence rate increased markedly with age, and there were no major differences between men and women in subjects younger than 60 years. In subjects aged 60–69 years, the incidence was higher in men than in women (P = 0.05; Fig. 1). Subject characteristics are shown in Table 1.
Figure 1. Line graph showing the incidence of venous thromboembolism (VTE) in men and women with increasing age: the Tromsø study 1994–2007.
Download figure to PowerPoint
Table 1. Baseline characteristics of subjects enrolled in the Tromsø study (1994–95)
|Age (years)||45 ± 14|
|Gender (% women)||52.6 (11 212)|
|Body mass index (kg m−2)||25.1 ± 3.8|
|Systolic blood pressure (mmHg)||133 ± 20|
|Diastolic blood pressure (mmHg)||78 ± 12|
|Triglycerides (mmol L−1)||1.52 ± 1.00|
|Total cholesterol (mmol L−1)||5.99 ± 1.29|
|HDL-cholesterol (mmol L−1)||1.50 ± 0.41|
|Smoking (%)||36.3 (7735)|
|Self-reported diabetes (%)||1.4 (305)|
|Family history of myocardial infarction (%)||24.9 (5312)|
Characteristics of the VTE events are shown in Table 2. Among the subjects with incident VTE, 64.5% had DVT and 35.5% had PE, with or without concurrent DVT (Table 2). A total of 138 events (42.2%) were unprovoked (Table 2). Cancer was the most common provoking factor, and 22% of the patients with VTE had active cancer at the time of diagnosis (Table 2).
Table 2. Characteristics of venous thromboembolism (VTE) events (n = 327): the Tromsø study (1994–2007)
| ||% (n)|
|Deep vein thrombosis||64.5 (211)|
|Pulmonary embolism||35.5 (116)|
|Clinical risk factors|
| Estrogens (HRT, oral contraceptives)||7.0 (23)|
| Heredity*||2.8 (9)|
| Pregnancy||0.6 (2)|
| Other medical conditions†||19.3 (63)|
| Surgery||18.3 (60)|
| Trauma||7.3 (24)|
| Acute medical conditions||15.6 (51)|
| Cancer||22.0 (72)|
| Immobilization (bed rest > 3 days, wheelchair)||16.8 (55)|
| Other‡||4.6 (15)|
Table 3 summarizes crude incidence rates and age- and gender-adjusted HRs for VTE by various levels of cardiovascular risk factors. Age was significantly associated with increased risk of VTE. Subjects 70 years or older had an elevenfold higher risk of VTE (HR, 11.33; 95% CI, 8.32–15.42), compared to those younger than 50 years of age. Men had a significantly higher risk of VTE than women (HR, 1.25; 95% CI, 1.01–1.55). The incidence rate of VTE increased linearly from 0.9 to 2.7 over the three categories of BMI, and obese individuals (BMI ≥ 30 kg m−2) had an almost 2-fold increased risk of VTE (HR, 1.92; 95% CI, 1.41–2.63) compared to those with BMI lower than 25 kg m−2 (Table 3). Neither systolic nor diastolic blood pressure was associated with increased risk of VTE. Test for trends in HRs were borderline significant for quartiles of triglycerides (P = 0.08) and total cholesterol (P = 0.08), whereas HDL-cholesterol showed no association with VTE (P = 0.9; Table 3). Smoking was not associated with VTE (HR, 1.04; 95% CI, 0.82–1.33). The incidence rate of VTE in subjects with self-reported diabetes was 3.71 compared to 1.37 in subjects without diabetes, and the risk of VTE by diabetes was non-significantly 1.4-fold increased (HR, 1.37; 95% CI, 0.74–2.47).
Table 3. Incidence rates (events per 1000 person-years) of venous thromboembolism (VTE) by risk factor status, and age- and gender-adjusted hazard ratios for VTE: the Tromsø study
| ||VTE events||Person-years||Incidence rate (95% confidence interval)|| Hazard ratio (95% confidence interval)|
| <50||84||154 781||0.54 (0.44–0.67)||1.00|
| 50–69||164||64 449||2.54 (2.18–2.97)||4.69 (3.61–6.10)|
| ≥70||79||13 859||5.70 (4.57–7.11)||11.33 (8.32–15.42)|
| Female||161||123 367||1.31 (1.12–1.53)||1.00|
| Male||166||109 722||1.51 (1.30–1.76)||1.25 (1.01–1.55)|
|Body mass index (kg m−2)|
| <25||113||125 030||0.90 (0.75–1.09)||1.00|
| 25–29||151||84 805||1.78 (1.52–2.09)||1.44 (1.13–1.85)|
| ≥30||63||23 254||2.71 (2.12–3.47)||1.92 (1.41–2.63)|
|Systolic blood pressure (mmHg)|
| <120||44||55 724||0.79 (0.59–1.06)||1.00|
| 120–130||54||63 666||0.85 (0.65–1.11)||0.87 (0.58–1.30)|
| 131–142||79||55 832||1.41 (1.13–1.76)||1.11 (0.76–1.62)|
| ≥143||150||57 868||2.59 (2.21–3.04)||1.02 (0.71–1.48)|
|Diastolic blood pressure (mmHg)|
| <70||54||60 671||0.89 (0.68–1.16)||1.00|
| 70–76||52||56 936||0.91 (0.69–1.19)||0.77 (0.53–1.14)|
| 77–84||82||57 140||1.44 (1.16–1.79)||0.93 (0.66–1.33)|
| ≥85||139||58 342||2.38 (2.02–2.81)||1.08 (0.78–1.51)|
|Triglycerides (mmol L−1)|
| <0.87||55||58 705||0.94 (0.72–1.22)||1.00|
| 0.87–1.24||69||57 255||1.21 (0.96–1.53)||1.01 (0.71–1.45)|
| 1.25–1.87||93||58 767||1.58 (1.29–1.94)||1.15 (0.82–1.61)|
| ≥1.87||110||58 362||1.88 (1.56–2.27)||1.28 (0.92–1.78)|
|Total cholesterol (mmol L−1)|
| <5.06||38||56 939||0.67 (0.49–0.92)||1.00|
| 5.06–5.85||57||58 132||0.98 (0.76–1.27)||0.99 (0.66–1.50)|
| 5.86–6.81||89||58 929||1.51 (1.23–1.86)||1.10 (0.75 1.63)|
| ≥6.82||143||59 089||2.42 (2.05–2.85)||1.29 (0.89–1.88)|
|HDL-cholesterol (mmol L−1)|
| <1.21||86||56 387||1.53 (1.23–1.88)||1.00|
| 1.21–1.44||71||58 391||1.22 (0.96–1.53)||0.80 (0.59–1.10)|
| 1.45–1.74||91||59 852||1.52 (1.24–1.87)||0.98 (0.73–1.33)|
| ≥1.75||79||58 459||1.35 (1.08–1.69)||0.80 (0.58–1.10)|
| No||227||147 876||1.53 (1.34–1.74)||1.00|
| Yes||100||85 213||1.17 (0.96–1.42)||1.04 (0.82–1.33)|
| No||316||230 124||1.37 (1.23–1.53)||1.00|
| Yes||11||2965||3.71 (2.05–6.70)||1.35 (0.74–2.47)|
|Family history of myocardial infarction|
| No||211||174 845||1.21 (1.06–1.38)||1.00|
| Yes||116||58 244||1.99 (1.66–2.39)||1.31 (1.04–1.65)|
The incidence rate of VTE in subjects with family history of MI was 1.91 (95% CI, 1.66–2.39) per 1000 person-years compared to 1.21 (95% CI, 1.06–1.38) in those without family history of MI. Subjects with family history of MI had a significant 1.3-fold increased risk of VTE (HR, 1.31; 95% CI, 1.04–1.65; Table 3).
Table 4 shows multivariable HRs for total and unprovoked VTE. Family history of MI remained significant as a risk factor for both total VTE (HR, 1.27; 95% CI, 1.01–1.60) and unprovoked VTE (HR, 1.46; 95% CI, 1.03–2.07) after adjustments for all the other traditional cardiovascular risk factors (Table 4). The risk of unprovoked VTE increased with increasing number of affected relatives, and subjects with family history of MI in ≥2 first degree relatives had a 1.75-fold increased risk of unprovoked VTE compared to subjects without family history of MI (HR, 1.75; 95% CI, 0.90–3.41). Age (HR per decade, 1.90; 95% CI, 1.74–2.07), male gender (HR, 1.36; 95% CI, 1.08–1.73), and BMI (HR per 3 kg m−2, 1.24; 95% CI, 1.15–1.35) were also significantly associated with VTE in the multivariable model, whereas blood pressure, blood lipids, and smoking showed no association with VTE (Table 4).
Table 4. Multivariable hazard ratios (HRs) with 95% confidence interval (CI) for venous thromboembolism (VTE): the Tromsø study (1994–2007)
| ||Total VTE||Unprovoked VTE|
|HR (95% CI)||HR (95% CI)|
|Family history of myocardial infarction (yes/no)||1.27 (1.01–1.60)||1.46 (1.03–2.07)|
|Age (decade)||1.90 (1.74–2.07)||1.80 (1.59–2.06)|
|Gender (male)||1.36 (1.08–1.73)||1.44 (1.00–2.07)|
|Body mass index (3 kg m−2)||1.24 (1.15–1.35)||1.30 (1.16–1.47)|
|Smoking (yes/no)||1.14 (0.89–1.46)||0.95 (0.64–1.40)|
|Self-reported diabetes (yes/no)||1.24 (0.67–2.29)||1.08 (0.39–2.96)|
|HDL-cholesterol (0.5 mmol L−1)||1.02 (0.87–1.20)||1.26 (1.00–1.60)|
|Total cholesterol (1 mmol L−1)||1.07 (0.97–1.18)||1.01 (0.87–1.18)|
|Triglycerides (1 mmol L−1)||0.90 (0.78–1.03)||0.95 (0.77–1.17)|
|Diastolic blood pressure (10 mmHg)||0.98 (0.85–1.12)||0.89 (0.72–1.10)|
Surprisingly, HDL-cholesterol was associated with an increased risk of unprovoked VTE (HR, 1.26; 95% CI, 1.00–1.60; Table 4). Age- and gender-stratified analysis revealed that the increased risk of unprovoked VTE by HDL-cholesterol was highest in women below 50 years of age (data not shown). The increased risk of VTE by HDL-cholesterol in young women was not explained by estrogenic supplementation or menopausal status, as adjustment for these variables did not attenuate the HR.
- Top of page
- Disclosure of Conflict of Interests
Recent studies have suggested that atherosclerotic cardiovascular diseases and VTE share common risk factors [6,8–12,17]. In this study, age, male gender, BMI, and family history of MI were identified as independent risk factors for VTE. Other traditional cardiovascular risk factors, including blood pressure, total cholesterol, and triglycerides were not independent predictors for VTE. Surprisingly, HDL-cholesterol was significantly associated with increased risk of unprovoked VTE in multivariable analysis.
Familial aggregation of coronary heart disease is a strong and independent risk factor for arterial cardiovascular events [13,14]. To the best of our knowledge, we have for the first time identified an association between family history of MI and VTE. The risk of unprovoked VTE in our study was of similar magnitude to arterial cardiovascular events in familial aggregates of cardiovascular diseases [13,14].
A family history of cardiovascular disease has previously been identified as a predictor for subclinical carotid atherosclerosis , and as an independent predictor for coronary calcification in young subjects . A higher frequency of carotid plaques  and coronary calcification  has been reported in patients with VTE. However, the impact of subclinical atherosclerosis as a risk factor for VTE is controversial. A recent prospective study showed a significant association between carotid intima media thickness (IMT) and first episode of VTE, but the association disappeared after adjusting for cardiovascular risk factors , whereas no association was observed in subjects above 65 years of age . Atherosclerosis is associated with endothelial dysfunction, coagulation activation, and platelet activation , with a subsequent increase in risk of arterial thrombosis [23,24]. Thus, it may be speculated, that underlying endothelial dysfunction, coagulation, and platelet activation are a plausible link between arterial and venous thrombosis by appearing in the slow-flowing venous system. The latter assumption is supported by a prospective study, showing that a first arterial cardiovascular event is associated with subsequent development of VTE .
Prospective studies in elderly  and middle-aged  subjects have failed to confirm subclinical atherosclerosis, as an independent risk factor for VTE. Cardiovascular risk factors, such as smoking and dyslipidemia, had synergistic interactions with family history on the risk for cardiovascular disease [15,16]. In our study, traditional cardiovascular risk factors, including blood pressure, blood lipids, and smoking, were not independent risk factors for VTE, and did not influence family history of MI as a risk factor for VTE in multivariable analysis. These findings indicate that other factors than the atherosclerotic process itself may also contribute to the link between arterial and venous thrombosis. In addition, the sequence of thrombotic events is not predetermined. Patients who have suffered a VTE event are reported to have increased risk of an arterial thromboembolic event during long-term follow-up . Lack of associations with other traditional cardiovascular risk factors, and undetermined sequence of thrombotic events, suggest that increased risk of VTE in subjects with family history of MI is probably a result of genetic and environmental determinants that exert their impact by as yet unknown mechanism(s).
Another possible link between arterial thromboembolism and VTE could be the metabolic syndrome, a cluster of cardiovascular risk factors associated with risk of cardiovascular disease and overall mortality [25,26]. Recent case–control studies have reported a 2-fold higher prevalence of the metabolic syndrome in subjects with idiopathic VTE  and recurrent VTE  compared to controls. However, this association has not yet been confirmed in prospective studies of general populations.
Some studies [12,29,30] have shown decreased HDL-cholesterol in patients with VTE, suggesting a protective role of HDL-cholesterol on the risk of VTE, a concept supported by the antithrombotic properties of HDL . Other studies [11,32,33], however, were unable to establish any association between HDL-cholesterol and risk of VTE. In contrast to previous studies, HDL-cholesterol was associated with increased risk of unprovoked VTE, and apparently the risk was highest in young women. The finding was not explained by estrogenic supplementation or menopausal status, as adjustments for these variables did not attenuate the HR. The association of HDL-cholesterol and increased risk of unprovoked VTE could possibly be explained by chance, or by unrecognized confounders.
Most previous prospective studies may have limited implications to the general population, as they were carried out in specific population cohorts, such as male physicians, female nurses, middle-aged or elderly [3,9–11], or because no risk assessment for cardiovascular risk factors was performed . Our prospective, population-based cohort study included both male and female subjects above 24 years of age. In agreement with previous studies [9–12], we confirmed that age and BMI were strong and independent risk factors for both total and unprovoked VTE. Obese individuals (BMI ≥ 30 kg m−2) had a 2-fold increased risk of VTE, and subjects above 70 years of age had an elevenfold increased risk of VTE compared to subjects below 50 years of age.
In our study, subjects with self-reported diabetes had a 2.5-fold higher incidence rate of VTE, and a non-significantly 1.4-fold increased risk of VTE. Previous reports from prospective studies are conflicting, with regard to the impact of diabetes on risk of VTE [3,11]. The apparent inconsistency between studies may to some extent rely on different diagnostic criteria for diabetes.
The main strengths of our study are its prospective design, long-term follow-up, and large number of participants and validated VTE events. The study cohort represents a general population, with a wide age structure, because of the high rate of attendance among people in the Tromsø community. All hospital care in the region is exclusively provided by a single hospital, which enhances the possibility of a complete VTE register. However, the study has some limitations. Analyses were restricted to subjects who had provided information on family history of MI. However, the HRs for VTE by the traditional cardiovascular risk factors were essentially identical, when analyzing the complete cohort (data not shown), implying that the selected cohort was representative for the total population. Another possible limitation of the study regards misclassification of family history of MI among those who reported family history. Both under-reporting and over-reporting of affected relatives are possible. In a previous validation study on the reliability of reported family history of MI, Kee et al.  demonstrated a high specificity (97%) and a lower sensitivity (68%) of a positive family history of MI. Thus, underestimation of the risk associated with family history of MI is more likely. In our study, several risk factors were modifiable. In general, modifiable risk factors are a potential limitation of cohort studies, especially when the time between exposure and disease manifestation is very long. This type of non-differential misclassification generally leads to underestimation of the true associations. Furthermore, information on concomitant treatment was not available in our study, and thus could not be taken into consideration.
In conclusion, family history of MI as an independent risk factor for VTE provides further evidence for the concept of a link between arterial and venous thrombosis. Traditional cardiovascular risk factors did not seem to be underlying determinants for this association, suggesting that family members share yet unknown genetic or environmental risk factors.