Epidemiology and risk factors for varicose veins among older people: cross-sectional population study in the UK

Abstract

Background

There are many hypotheses concerning risk factors for the development of varicose veins based mostly on pathophysiological plausibility. Population studies have been carried out mostly on the middle aged with relatively few on elderly populations.

Objectives

To investigate epidemiological risk factors for varicose veins in an elderly population in the UK.

Methods

The South Wales Skin Cancer study – an examination survey undertaken between 1988 and 1991 of a random sample (n = 792) drawn from all patients aged 60 and over registered with a general practitioner in South Glamorgan. Exposure variables were obtained from a structured administered questionnaire combined with clinical examination. Unadjusted and adjusted odds ratios were estimated using logistic regression.

Results

The response rate was 71% with an average age of 71 years (range 60–97). The age-adjusted prevalence of trunk varices was 63.2% (95% confidence interval [CI] 57.9–68.4%) in men and 57.0% (95% CI 50.6–63.4%) in women. In a multiple logistic regression the significant risk factors for varicose veins were increasing age (P value = 0.001), obesity (odds ratio [OR] 3.28, 95% CI 1.25–8.63, P = 0.042), self-reported history of deep vein thrombosis (DVT) (OR 3.19, 1.16–8.78, P = 0.024) and history of hypertension (OR 0.58, 0.38–0.89, P = 0.013). The results for gender suggested that women were at greater risk than men, but this was not statistically significant (OR 1.53, 0.99–2.38, P = 0.056).

Conclusion

Trunk varices occur very commonly in older age groups with increasing age, obesity and possibly female sex as risk factors. Associations found with DVT and hypertension were based on history alone and must be interpreted with caution.

Keywords

varicose veins risk factors epidemiology

Introduction

Venous disease, which includes varicose veins as one manifestation, has been estimated to be one of the most commonly reported chronic conditions in the Western world. Estimates of the prevalence of varicose veins vary greatly, from 2% to 56% in men and from less than 1% to 73% of women,¹ although only a few of these estimates have been derived from population-based studies. The range of prevalence rates in population-based epidemiological studies is less extreme² with a range of 6.8–39.7% in men and 24.6–41.7% in women. Varicose veins are associated with considerable morbidity – including the major problem of chronic venous ulceration – and venous disease has been estimated to consume approximately 2% of national health-care resources in the UK.³

Previous epidemiological studies (reviewed by Beebe-Dimmer et al. ¹) found evidence of a positive association between the prevalence of varicose veins and increasing age; family history of varicose veins; having an occupation requiring standing; obesity and a history of phlebitis or blood clots. They report conflicting evidence concerning constipation, smoking, oral contraceptive use, hormone replacement therapy (HRT), hypertension, physical activity and injury. Conflicting evidence exists concerning the association between the prevalence of varicose veins and gender with a number of studies finding a significantly increased prevalence in women, others the opposite or no difference. For example, Evans et al. ⁴ report that men have a greater risk at all ages whereas Carpentier et al. ⁵ report that women have a great risk. Beebe-Dimmer et al. ¹ conclude that further studies are needed to clarify the relationship with a number of factors, including smoking, body mass index (BMI), physical activity, oral contraceptives and HRT, and a history of diabetes, hypertension and traumatic injury to the extremities. Furthermore, some of the stated established risk factors¹ have not been confirmed in all cases. The relationship with BMI is controversial, for example with some evidence for an association in both sexes,⁶ women only^7,8 or no association at all.⁵

Many studies are based on convenience samples, such as factory workers, rather than true random samples of the general population. Further research has been widely recommended.^1,2,4 This study derives from an examination survey undertaken primarily to investigate non-melanoma skin cancer and its precursors, which opportunistically documented the presence and potential risk factors for a number of other dermatological conditions, including superficial venous disease.

Methods

The methods of the South Wales Skin Cancer Study have been fully described elsewhere.⁹ Briefly, during 1988 and 1989 a random sample of 792 subjects aged 60 years or over was drawn from the then South Glamorgan Family Health Services Authority register. Those agreeing to participate were visited in their homes where a questionnaire was administered and limited skin examination undertaken by a research registrar in dermatology.

Questionnaire

Smoking history was recorded as never having smoked, ex-smoker or current smoker. Social class was based on the main occupation held during the subject's working life. For married or widowed women the social class was based on their husband's last main employment. Social class was classified according to the Registrar General's scale of Social Class and Socio-economic groups into I: Professional occupations; II: Intermediate occupations; IIInm: Skilled non-manual occupations; IIIm: Skilled manual occupations; IV: Partly skilled occupations; and V Unskilled occupations. The subject's age and gender were also obtained. A history of hip surgery (including surgery for trauma) was also recorded by self-reporting; similarly histories of pelvic surgery (including any surgery involving general or spinal/epidural anaesthetic) and lower limb surgery were also self-reported. A history of deep vein thrombosis (DVT) was based on asking each subject if they had ever been told by a doctor that they had suffered a clot in the deep veins of either leg; similarly a history of hypertension and diabetes was based on the respondents ever being told by a doctor about the condition.

Examination

Examinations were undertaken in the subject's own home in warm and well-lit surroundings by a trained research registrar in dermatology. The definition of varicose veins used in this study includes trunk varices and reticular varices, but not hyphen webs/venectasias. This corresponds to C1 or above using the clinical, aetiological, anatomical and pathological elements scale.¹⁰ Varicose veins were recorded as absent, slight, moderate or severe on each leg depending on the visual extent and diameter of the varicosities, following that of Widmer in the Basle study.¹¹ Height and weight were recorded using a simple stadiometer and bathroom scales; BMI was calculated as weight (kg) divided by height squared (m²). The degree of ability to move the ankle was also recorded for each ankle as none, poor, average or good, depending on the range of movement without application of resistance.

Statistical analysis

For each subject varicose veins was defined as being present if either leg was classified as having slight, moderate or severe varicose veins; and absent if both legs were classified as having no varicose veins. Age was classified into five groups of four five-year age bands (60–64, 65–69, 70–74 and 75–79) and one unlimited age band (80+). BMI classified as underweight (≤20), normal weight (20 to 25), overweight (>25 to 30) and obese (>30). Ankle impairment was classified as present if movement was coded as poor or none on either or both ankles.

Unadjusted odds ratios (OR) and 95% Wald confidence intervals (CIs) were estimated for each possible risk factor using logistic regression. The overall effect of ordinal risk factors was tested by comparing the deviance of the model including the risk factor in a linear relationship to deviance of the model not including the risk factor. Adjusted ORs and 95% CIs were estimated using multiple logistic regression adjusting for all other risk factors. In order to avoid bias due to incomplete covariate data, multiple imputation was used, with 10 imputations, using the iteratively-chain equations package in Stata.¹² All other analyses were carried out using Stata 9.1 SE (TX, USA).

Results

Of the 762 people invited to participate, 560 (70.1%) entered the study. The non-responders were slightly more likely to be women, older and living in the inner city than the general population aged 60 or over in the area. It was only possible to classify 524 of the subject's social class due to some women being ‘economically inactive’ and were thus treated as missing data; no other covariate information was missing. Of the 560 participants, 58% were women and 42% were men. Overall, the prevalence (95% CI) of varicose veins was 60.4% (56.3–64.3%), with men having a prevalence of 56.4% (50.0–62.6%) and women 63.2% (57.8–68.2%). The mean (SD) age for men was 70.4 (7.3) years and for women was 71.2 (7.9) years; age-adjusted prevalence was 63.2% for men and 57.0% for women. The prevalences of varicose veins in men and women according to age, social class, smoking status and BMI are given in Table 1. Prevalences were higher in women than in men in almost all the subgroups, although there was considerable overlap in the 95% CIs. However, in very elderly subjects (80 years + ) and in those with obesity (BMI > 30) there was no suggestion of a female preponderance.

Table 1

Prevalence of varicose veins in men and women by age, social class, smoking status and BMI

Characteristic	Prevalence % (95% CI)
Characteristic	Male (n = 234)	Female (n = 326)
Age (years)
60–64	48.3 (36.2, 60.7)	51.7 (41.4, 61, 9)
65–69	59.2 (47.5, 69.8)	63.9 (53.1, 73.4)
70–74	47.4 (32.9, 62.5)	66.7 (53.4, 77.8)
75–79	63.4 (48.1, 76.4)	74.0 (61.2, 85.1)
80+	72.7 (51.8, 86.8)	66.7 (53.3, 77.8)
Social class*
I	38.9 (20.3, 61.4)	50.0 (15.0, 85.0)
II	64.3 (49.2, 77.0)	64.8 (51.5, 76.2)
IIINM	57.9 (42.2, 72.1)	63.6 (54.8, 71.7)
IIIM	59.5 (48.8, 69.4)	70.8 (50.8, 85.1)
IV	50.0 (35.5, 64.5)	61.4 (48.4, 72.9)
V	50.0 (23.7, 76.3)	56.7 (39.2, 72.6)
Smoking status
Never	52.0 (33.5, 70.0)	62.5 (54.8, 69.6)
Ex-smoker	59.2 (50.9, 66.9)	59.6 (49.2, 69.1)
Current smoker	52.2 (34.2, 74.2)	68.9 (35.4, 87.9)
BMI
≤20	54.6 (34.7, 73.0)	61.7 (47.4, 74.2)
21–25	55.4 (46.8, 63.7)	62.6 (55.3, 69.3)
26–30	55.7 (44.1, 66.8)	62.3 (51.1, 72.3)
>30	75.0 (46.8, 91.1)	73.9 (53.5, 87.5)
Total	56.4 (50.0, 62.6)	63.2 (57.8, 68.2)

BMI = body mass index; CI = confidence interval

*Thirty-six were economically inactive

The results of the univariate and multivariate logistic regression models are given in Table 2. Gender was not significantly associated with the presence of varicose veins in either the unadjusted or adjusted analysis. However, it was very close to significance in the adjusted analysis (P = 0.056), thus providing weak evidence of an association. Age was significantly associated with the presence of varicose veins in both the unadjusted and adjusted analyses, with those who were in the older age groups (75–79 and 80+ years) having a greater than two-fold odds of having varicose veins than those aged 60–64 years. BMI was associated with varicose veins only after adjustment (P = 0.042) with those subjects who were obese (BMI > 30) at a significantly greater risk than those with a BMI of less than 20 (adjusted OR 3.28, 95% CI 1.25–8.63). Social class and smoking status were not significantly associated with varicose veins.

Table 2

Unadjusted and adjusted OR of varicose veins by gender, age, social class, smoking, BMI and clinical history

Characteristic	Unadjusted		Adjusted*
Characteristic	OR (95% CI)	P value	OR (95% CI)	P value
Gender
Male	1		1
Female	1.33 (0.94, 1.87)	0.106	1.53 (0.99, 2.38)	0.056
Age (years)
60–64	1	0.003^†	1	0.001^†
65–60	1.59 (1.00, 2.51)		1.78 (1.09, 2.89)
70–74	1.39 (0.82, 2.35)		1.69 (0.97, 2.94)
75–79	2.27 (1.30, 3.95)		2.63 (1.46, 4.74)
80+	2.14 (1.19, 3.83)		2.42 (1.28, 4.58)
Social class ^†
I	1	0.602^†	1	0.392^†
II	2.51 (0.98, 6.41)		2.14 (0.79, 5.75)
IIINM	2.27 (0.92, 5.59)		1.79 (0.68, 4.72)
IIIM	2.25 (0.88, 5.72)		2.02 (0.76, 5.37)
IV	1.80 (0.70, 4.59)		1.46 (0.54, 3.92)
V	1.66 (0.57, 4.78)		1.15 (0.37, 3.59)
Smoking status
Never	1	0.977^†	1	0.152^†
Ex-smoker	0.93 (0.63, 1.38)		1.16 (0.73, 1.84)
Current smoker	1.00 (0.64, 1.57)		1.45 (0.87, 2.42)
BMI
≤20	1	0.320^†	1	0.042^†
21–25	1.01 (0.59, 1.71)		1.23 (0.69, 2.17)
26–30	0.99 (0.55, 1.77)		1.32 (0.69, 2.50)
>30	1.97 (0.80, 4.85)		3.28 (1.25, 8.63)
Impaired ankle movement	2.38 (0.95, 6.01)	0.065	1.93 (0.73, 5.12)	0.185
Medical history
DVT	3.47 (1.31, 9.21)	0.013	3.19 (1.16, 8.78)	0.024
Diabetes	0.59 (0.25, 1.36)	0.215	0.53 (0.21, 1.33)	0.175
Hypertension	0.66 (0.44, 0.97)	0.035	0.58 (0.38, 0.89)	0.013
Pelvic surgery	1.11 (0.71, 1.73)	0.647	0.96 (0.59, 1.55)	0.862
Hip surgery	0.57 (0.20, 1.59)	0.278	0.40 (0.13, 1.22)	0.107
Lower limb fracture	1.11 (0.70, 1.73)	0.674	0.98 (0.60, 1.60)	0.529

BMI = body mass index; CI = confidence interval; OR = odds ratio; DVT, deep vein thrombosis

*Adjusted for all other characteristics in table

^†Test for linear trend

Table 2 also shows that impaired ankle movement was not significantly associated with the presence of varicose veins. Furthermore, no evidence of an association between varicose veins and either a history of diabetes, hip surgery or pelvic surgery was found either in the unadjusted or the adjusted analysis. A significant association was observed between varicose veins and a history of DVT, with those having a history of DVT having over three times the odds of those who did not (adjusted OR = 3.19, 95% CI 1.16–8.78). Conversely, those with a history of hypertension were less likely to have varicose veins (adjusted OR = 0.58, 95% CI 0.38–0.89).

Discussion

In this population-based cross-sectional study, we examined the relationships between varicose veins and a number of possible risk factors in an elderly population. A significant association between varicose veins and age was found with an increasing prevalence in the older age groups. Similarly, a greater prevalence of varicose veins was observed in those with a history of doctor diagnosis of DVT compared with those without a history. Further, a lower prevalence was observed for those with a history of hypertension compared with those without. There was slight evidence of an association with gender and BMI, with obese people (BMI > 30) having almost three times the odds of having varicose veins compared with those people with a BMI of less than 20. There was no evidence of an association with a history of diabetes, a history of pelvic surgery, a history of hip surgery, a history of lower limb fracture, smoking status, and social class.

Since this is a cross-sectional study we are unable to show that factors have a causal effect, we can only show an association. Furthermore, possible risk factors identified were based on history and may have been subject to recall bias. However, the major strength of this study is the good participation rate and that this is one of the largest studies to examine risk factors for varicose veins in an elderly population.

According to Beebe-Dimmer et al ¹ a number of studies have found that female gender was a risk factor for varicose veins. Although we did not find a significant association, the adjusted significance level provided limited evidence of a difference. This is broadly in line with most studies^{5,7,8,13–17} in which an increased risk was found in women compared with men, but at odds with another study⁶ in which men were at greater risk than women. Some previous evidence of a relationship between BMI and varicose veins has been found,^1,6 particularly in women,^1,7,8 but other studies have found no association⁶ in men or women. Our study supports the hypothesis of an association between BMI, particularly obesity and varicose veins. However, we found no evidence of an interaction between BMI and gender (results not shown). Smoking was not statistically associated with varicose veins, in line with other studies^5,6,8 but at odds with one study⁷ which found an increased risk of varicose veins among male smokers, but not female smokers. This may be due to different life-style factors adjusted for in the analyses. No association was found with social class, in agreement with another study which⁶ also found no association.

Thrombosis has been suggested as a possible risk factor.¹ We found a significant association between varicose veins and a history of a doctor diagnosis of DVT, with those having had DVT having over three times the odds of having varicose veins compared with those who did not. However, given that DVT is a venous disease there may be considerable recall bias between those with and without varicose veins and also misinterpretation by subjects on the exact condition specified by the doctor. We did not have the opportunity to examine previous case-notes and so these results must be interpreted with considerable caution. Hypertension has also been suggested as a risk factor¹ and in this study a significant association with a history of hypertension was found. The only other population-based studies to examine this relationship did not find a significant one,^5,8,18 although an association with systolic blood pressure was found in women, but not men in another study.⁷ Since the presence of hypertension was also based on history, this inverse association with varicose veins must also be interpreted with caution. No association was found between the presence of diabetes and varicose veins in line with other studies,^7,8 which also found no association.

The South Wales Skin Cancer Study is one of the few and certainly the first population study in the UK to examine the prevalence of varicose veins and association with risk factors in the elderly. The association with age and BMI are compatible with findings in studies of younger groups and there was still a suggestion in the elderly of a higher prevalence in women compared with men.

References

Beebe-Dimmer

, Pfeifer

, Engle

, The epidemiology of chronic venous insufficiency and varicose veins. Ann Epidemiol 2005;15:175–84

Robertson

, Evans

, Fowkes

. Epidemiology of chronic venous disease. Phlebology 2008;23:103–11

Laing

. Chronic Venous Diseases of the Leg. London: Office of Health Economics, 1992

Evans

, Fowkes

, Ruckley

, Prevalence of varicose veins and chronic venous insufficiency in men and women in the general population: Edinburgh vein study. J Epidemiol Community Health 1999;53:149–53

Carpentier

, Maricq

, Biro

, Prevalence, risk factors, and clinical patterns of chronic venous disorders of lower limbs: a population-based study in France. J Vasc Surg 2004;40:650–9

Lee

, Evans

, Allan

, Lifestyle factors and the risk of varicose veins: Edinburgh Vein Study. J Clin Epidemiol 2003;56:171–9

Brand

, Dannenberg

, Abbott

, The epidemiology of varicose veins: the Framingham Study. Am J Prev Med 1988;4:96–101

Canonico

, Gallo

, Paolisso

, Prevalence of varicose veins in an Italian elderly population. Angiology 1998;49:129–35

Harvey

, Frankel

, Marks

, Non-melanoma skin cancer and solar keratoses. I. Methods and descriptive results of the South Wales Skin Cancer Study. Br J Cancer 1996;74:1302–7

10.

Eklof

, Rutherford

, Bergan

, Revision of the CEAP classification for chronic venous disorders: consensus statement. J Vasc Surg 2004;40:1248–52

11.

Widmer

, ed. Peripheral Venous Disorders: Prevalences and Sociomedical Importance. Bern: Hans Huber, 1978

12.

Royston

. Multiple imputation of missing values. Stata J 2004;4:227–41

13.

Sisto

, Reunanen

, Laurikka

, Prevalence and risk factors of varicose veins in lower extremities: mini-Finland health survey. Eur J Surg 1995;161:405–14

14.

Laurikka

, Sisto

, Tarkka

, Risk indicators for varicose veins in forty- to sixty-year-olds in the Tampere varicose vein study. World J Surg 2002;26:648–51

15.

Franks

, Wright

, Moffatt

, Prevalence of venous disease: a community study in west London. Eur J Surg 1992;158:143–7

16.

Abramson

, Hopp

, Epstein

. The epidemiology of varicose veins. A survey in western Jerusalem. J Epidemiol Community Health 1981;35:213–7

17.

Coon

, Willis

, Keller

. Causative factors of varicose veins: myths and facts. An epidemiological study of 610 women. VASA 1973;2:115–20

18.

Makivaara

, Ahti

, Luukkaala

, Arterial disease but not hypertension predisposes to varicose veins. Phlebology 2008;23:142–6