Taller women are at greater risk of recurrent venous thromboembolism: The Iowa Women’s Health Study

Summary

Venous thromboembolism (VTE) recurs frequently. Greater height is associated with increased risk of incident VTE, but it is unclear if height is related to risk of VTE recurrence. Recurrent VTE is associated with substantial morbidity and mortality, thus identifying individuals at greatest risk of experiencing a recurrent event, who may benefit from extended anticoagulant therapy, is vitally important. Using data from the Iowa Women’s Health Study we explored whether greater height was associated with increased risk of VTE recurrence. Among 1691 women who experienced an initial VTE event 286 (16.9%) experienced a recurrent event. Risk of recurrence was 76% (95% CI: 16%–186%) higher among women ≥66 inches [~168 centimeters] tall relative to those ≤62 inches [~158 centimeters] tall, after adjustment for age and waist circumference. Future research should evaluate whether body height improves clinical prediction of VTE recurrence risk.

Venous thromboembolism (VTE) frequently recurs, even long after the incident VTE. Approximately one-third of individuals with an initial VTE event will experience a recurrence within the subsequent decade [1–3], though the highest risk for recurrence is within the first year ( 8%) [4]. Recurrent VTE events are fatal in approximately 5% to 9% of patients [5]. Given this substantial morbidity and mortality, it is of clinical importance to be able to understand who will go on to have a recurrent VTE, as patients at elevated risk of recurrence may benefit from extended anticoagulant therapy [6]. In fact, determining the optimal duration of anticoagulation for unprovoked VTE has been described as the most important unanswered question in VTE management [7–9]. Older age, male gender, obesity, hereditary thrombophilia, elevated D-dimer levels, and persistent comorbidities are associated with greater risk of recurrence [6, 10, 11], though it is impossible to predict with certainty who will experience a recurrent event.

Greater body height has been associated with an increased risk of incident VTE in several epidemiologic prospective cohort studies [1, 12–16]. Longer legs have also been associated with greater VTE risk independent of total body height [17]. Taller individuals, and those with longer legs, may be subject to greater venous damage as a consequence of greater hydrostatic pressure [18–20]. Additionally, given their longer veins, taller individuals have a greater surface area where a thrombus may occur, and may have more venous valves which could also predispose to a greater incidence of VTE.

Although height and leg length have been linked to risk of incident VTE, the relation of height and leg length to VTE recurrence is unexplored. We used data from the Iowa Women’s Health Study (IWHS) to test the hypothesis that taller individuals are at greater risk of recurrent VTE.

Among 1691 women who had an incident VTE and were still enrolled in Medicare 6 months later, 286 (16.9%) went on to experience a recurrent VTE over a median of 4.11 years (range: 0.50 to 18.8 years) of follow-up. Distributions of covariates and reasons for censoring, stratified by height category, are presented in the Supplemental Table. Relative to women 62 in [~158 cm] or shorter, women 66 in [~168 cm] or taller were at 76% (95% CI: 16%, 186%) greater risk of recurrent VTE, after adjustment for age and waist circumference (Table). Overall, associations were somewhat stronger for unprovoked initial VTE than for provoked initial VTE, though the trend of taller women being at greater risk of recurrence was present for both.

Table.

Relation of height to risk of venous thromboembolism recurrence among 1691 women with a prior venous thromboembolism event^*: The Iowa Women's Health Study, 1986–2004

	≤ 62 in	63–65 in	≥ 66 in	P-trend
	[≤ 157.5 cm]	[160.0–165.1 cm]	[≥ 167.6 cm]
Total VTE
N events	52	195	39
N women total	354	1,162	175
Person years	1,969	6,255	811
IR (95% CI)	2.62 (1.98, 3.46)	3.10 (2.69, 3.57)	4.68 (3.42, 6.42)
HR (95% CI)	1.00	1.19 (0.87, 1.62)	1.76 (1.16, 2.68)	0.01
Unprovoked VTE**
N events	15	56	15
N women total	85	300	45
Person years	532	1,774	243
IR (95% CI)	2.71 (1.61, 4.54)	3.18 (2.45, 4.13)	5.97 (3.58, 9.95)
HR (95% CI)	1.00	1.17 (0.66, 2.08)	2.20 (1.07, 4.53)	0.04
Provoked VTE**
N events	37	139	24
N women total	269	862	130
Person years	1,437	4,482	568
IR (95% CI)	2.59 (1.86, 3.60)	3.08 (2.60, 3.65)	4.12 (2.76, 6.16)
HR (95% CI)	1.00	1.20 (0.83, 1.74)	1.59 (0.95, 2.68)	0.09

IR = Incidence Rate per 100 person-years; HR = Hazard Ratio; 95% CI = 95% Confidence Interval

^*Adjusted for age and waist circumference

^**Classification based on presentation of initial VTE event

Our results suggest that taller people are at a potentially clinically meaningful increased risk of VTE recurrence. The height (or leg length) and incident (or recurrent) VTE association performs well when judged by Hill’s classic epidemiologic criteria for causation [21]. The relation has appropriate temporality with height preceding VTE events, the associations are strong and for incident VTE have been replicated numerous times [1, 12–16], there is evidence of a dose response relationship, the association is biologically plausible, there are no clear alternate explanations for why taller people would be at greater risk of VTE, and the association is specific to VTE (for other cardiovascular conditions taller people are lower risk) [12, 22].

Although our data has a number of strengths, including the relatively large number of recurrent VTE events, there are also limitations. Ideally, this analysis should be repeated in a population of men and women, of a broader age-distribution, with validated VTE events and recurrence carefully classified, biomarker data, and where leg length is objectively measured. In particular, generalizability to younger age-groups (<65 years) is uncertain. Despite the limitations of our data, our findings clearly suggest that greater height is a risk factor for recurrent VTE.

Recently there has been great interest in predicting risk of recurrent VTE [6–9, 11, 23], given the high prevalence of recurrence, and associated morbidity and mortality. The underlying goal of this work is to stratify individual patients on their risk of VTE recurrence, in order to determine the most clinically efficacious length of anticoagulant therapy by carefully balancing the individual-level benefits of secondary prevention against the risk of adverse bleeding events. Recently, the Vienna Prediction Model was created [11], which quantitatively estimated individual-level VTE recurrence risk over 12 and 60 months on the basis of sex, location of initial VTE, and D-dimer levels. Given the relatively strong relation between height and recurrent VTE in our sample, we believe the simple anthropometric measures of height and/or leg lengths may enhance prediction of VTE recurrence. It is also possible that greater height (or leg length) may explain the higher risk of recurrent VTE frequently observed among men relative to women [6, 10, 11]. Unfortunately, given limitations of our dataset we were not able to address these research questions.

Height (and leg length) are not modifiable exposures and therefore cannot be intervened upon. However, assessment of leg length (height is a proxy, as has been shown previously [17]) is simple and inexpensive, and may be of clinical value in stratifying VTE patients for risk of VTE recurrence. Future research should explore whether knowledge of leg length or height enhances prediction in VTE recurrence risk assessment models.

Methods

The IWHS cohort includes 41,836 women who, in January 1986, were randomly sampled from the Iowa State driver’s list and responded to an initial questionnaire. This questionnaire queried numerous characteristics, including height to the nearest inch.

As detailed previously [1], the IWHS data were linked to Centers for Medicare Services (CMS) enrollment and health care utilization data from 1986 through 2004. Medicare provides payment for all or part of health care for most U.S. residents aged 65 and older [24], and data stemming from the payment of Medicare bills is now used widely as a population-based data source for clinical occurrences [24]. Of IWHS cohort members surviving to 65 years, 98% (40,377 of 40,997) met our CMS inclusion criteria and were therefore followed for incident VTE events [1].

IWHS participants were considered to have had a VTE if any of the following VTE ICD-9 codes occurred on their Medicare MedPAR (hospitalization) discharge diagnosis records: 415.1x, 451.1x, 451.2, 451.81, 451.9, 453.0, 453.1, 453.2, 453.3, 453.4x, 453.8, 453.9 [1]. Provoked VTE was classified on the basis of ICD codes indicating malignancy at the time of discharge, surgery in the three months prior to VTE diagnosis, hospitalization of ≥4 days or trauma in the two months prior to the VTE, or a serious medical disease at the time of hospital discharge. All remaining cases were classified as unprovoked [1].

A recurrent event was defined as one that occurred subsequent to, but independently of, the incident event. We had no specific information about independence of events. Therefore, in order to distinguish index VTE events from recurrent VTE events, we required a 6-month lag period between the discharge date of the initial event and the admission date of the subsequent event [1].

For this analysis of height and VTE recurrence we included only women who experienced an incident event. Cox proportional hazards regression was used to evaluate the relation between height and VTE recurrence. As in our previous IWHS manuscript, which reported the association between height and risk of incident VTE[1], height was categorized: ≤62 in [157.5 cm], 63–65 in [160.0–165.1 cm], ≥66 in [167.6 cm]. Person-years accrued from the date of the incident event until VTE recurrence, death, disenrollment, or December 31, 2004.