Intended for healthcare professionals


Seasonal variation in deep vein thrombosis

BMJ 1996; 312 doi: (Published 03 February 1996) Cite this as: BMJ 1996;312:284
  1. Henri Bounameaux, senior lecturera,
  2. Luc Hicklin, residenta,
  3. Sylvie Desmarais, research fellowa
  1. a Division of Angiology and Haemostasis, Department of Internal Medicine, University Hospital of Geneva, CH 1211 Geneva 4, Switzerland
  1. Correspondence to: Dr Bounameaux.
  • Accepted 1 November 1995

In many temperate countries coronary events and stroke, as well as related deaths, are more common in winter than in summer.1 Fatal pulmonary embolism has also been reported more often in winter.2 3 This might be linked to changes in coagulation factors4 and to peripheral vasoconstriction, leading to reduced blood flow in the legs.5 Alternatively, venous stasis due to vasodilatation is greatest in summer. These uncertainties prompted us to study the seasonal distribution of deep vein thrombosis of the legs by reviewing the month of presentation of all patients with this diagnosis referred to our vascular laboratory over six years.

Subjects, methods, and results

All files coded as “suspected” deep vein thrombosis (n=7303, mean (SD) age 63 (19), 58% women, 85% medical patients) or “confirmed” deep vein thrombosis (n=1905, age 65 (20), 54% women, 83% medical patients) were selected from the database of our vascular laboratory over the period 1989-94. Diagnosis was made by a combination of continuous wave Doppler and venous occlusion plethysmography or, from 1992, by B mode venous compression ultrasonography. Inconsistency of the extracted data was found in only seven out of 300 cases (2.3%) by comparison with randomly selected original charts.

The monthly number of confirmed events over the six year period ranged from 27 to 37 in January, 13-29 in February, 23-33 in March, 16-29 in April, 16-38 in May, 20-36 in June, 14-39 in July, 18-40 in August, 24-32 in September, 22-32 in October, 21-31 in November, and 24-37 in December. The monthly variation was similar for suspected events. The figure shows the number of suspected and confirmed events by season over the six years studied, with seasons defined as follows: winter (December, January, February); spring (March, April, May); summer (June, July, August); autumn (September, October, November). Time series analysis with the moving average technique did not suggest any regular or irregular variation in the occurrence of the events. In addition, no differences were found for the monthly or seasonal distribution of the suspected events (Kruskal-Wallis, P=0.43 and 0.99, respectively) or confirmed events (P=0.27 and 0.14, respectively).


This survey shows that in Geneva, a city with a continental climate and quite wide temperature differences between summer and winter, there is no seasonal variation in the incidence of suspected and confirmed deep vein thromboses. This conclusion is based on a large population and a long period.

Fig 1
Fig 1

Number of suspected and confirmed deep vein thromboses related to seasons during 1989-94 in Geneva

Two previous reports had suggested an increased incidence in winter of fatal pulmonary embolism after total hip arthroplasty2 or among subjects who died suddenly,3 but the conclusions were derived from small numbers of events. Moreover, other confounding factors might theoretically be responsible for an increased incidence of fatal pulmonary embolism in winter. Although the embolic source, leg thrombosis, is no more frequent, winter related comorbidities might contribute to a reduced tolerance to small emboli.

Our survey has limitations. Firstly, it is retrospective, but the data used were all registered prospectively and their quality was ascertained in a randomly selected sample. Secondly, the diagnosis was not made by means of venography in most cases, but objective, non-invasive diagnostic means with recognised diagnostic performances were used. Moreover, if we extrapolate our 318 yearly events to the 380000 inhabitants of Geneva, the yearly incidence of confirmed thromboses was eight per 10000, similar to the five per 10000 found in population based studies. Lastly, we found no seasonal variations in the number of suspected cases of deep vein thrombosis, a condition which is not dependent on a diagnostic procedure.

We thank Ms R Avvanzino and Ms J Hirt for maintaining the database and Ms B Mermillod for statistical advice.


  • Funding Dr Desmarais was supported by a grant from the Royal College of Physicians and Surgeons of Canada.

  • Conflict of interest None.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.