Correlation between Restless Leg Syndrome and Superficial Venous Reflux

Abstract

Restless leg syndrome (RLS) is a common cause of lower extremity discomfort. We hypothesized that patients with RLS symptoms have higher rates of deep and superficial venous reflux (SVR). Retrospective review of patients ≥18 years of age evaluated in a venous center from December 2018 to February 2019. Differences in rates of RLS symptoms, demographics, comorbidities, and clinical and radiologic presence of venous disease were analyzed. Overall, 207 patients were analyzed; 140 (67.6%) reported RLS symptoms ( n  = 25 with prior RLS diagnosis). RLS symptoms were more common with superficial or combined superficial and deep venous reflux (DVR) compared with those without reflux ( p  < 0.001). Patients with RLS symptoms as opposed to those without had similar demographics and comorbidities (all p  > 0.05) but increased rates of venous pain, phlebitis, family history of venous disease, lower extremity swelling and SVR, and combined SVR and DVR (all p  < 0.05). Our multivariable logistic regression found presence of SVR, and family history of venous reflux was associated with RLS symptomatology (all p  < 0.001). Ninety-nine patients with RLS underwent ablation; of them, 93 had duplex-proven reflux resolution of which 81 (87%) reported RLS symptom improvement. This included 13 of 16 (81.3%) with prior RLS diagnosis. SVR is associated with increased rates of RLS symptoms in a vein center population. Therefore, RLS symptoms should trigger a targeted venous evaluation. Our results suggest that venous ablation may lead to resolution of RLS symptoms in patients with SVR, but randomized prospective trials with strict RLS definition criteria are warranted to confirm these outcomes.

Chronic venous disease is a common problem with varicose veins affecting more than 25 million adults in the United States and more than 6 million with more advanced venous disease. ¹ Chronic venous disease and specifically venous reflux is caused by a failure of venous pumps and valves against gravity leading to venous reflux. Venous reflux manifests across a spectrum of conditions ranging from simple telangiectasias to more advanced stages including venous ulceration. ²

While much is known about venous disease, surprisingly little has been written about the relationship between restless leg syndrome (RLS) and venous reflux since it was first described by Dr. Karl A. Ekbom in 1944. ³ To date, only a handful of studies have investigated the association between RLS and superficial venous reflux (SVR). ^{4 5 6} Furthermore, only a few reports exist regarding successful treatment of RLS symptoms with endovenous ablation. ^{4 5}

RLS is a common condition with a reported prevalence ranging from 4% to 29% in North America and Western European populations. ⁷ RLS is a sensorimotor condition with classic symptoms that include an urge to move legs typically accompanied by an unpleasant sensation that begins or worsens during periods of rest or inactivity that are worse in the evening or night time and are not related to another medical or behavioral condition. ⁸ RLS has been shown to have a significant impact on health-related quality of life, ^{9 10 11 12} including mental health, ¹³ in addition to reduced work productivity. ^{14 15} Additionally, RLS is associated with a significant economic burden that estimates in the literature approximating excess health care expenditure spent (in dollar) ranging from $774 to $2,448 annually. ^{14 16}

Despite RLS being classified/listed as a symptom of chronic venous disease in many sources including textbooks, it has never been considered an indication for treatment. Rather, classic RLS symptoms are often used by clinicians to diminish the possibility of a venous pathology when considering the differential diagnosis for leg symptoms. With newer, safer treatments available for venous disease, past recommendations may be subject to change. Therefore, our study sought to investigate the association between RLS symptoms and SVR in a busy clinical practice to identify a potentially treatable future target. Our second end point was to identify differences in patient demographics and comorbidities between those with or without RLS in a venous clinic setting. Our final endpoint was to examine outcomes of RLS symptoms after successful ablation in those undergoing treatment for their venous disease.

Methods

Patient Criteria

A retrospective chart review was performed of prospectively collected patient data from a single Intersocietal Accreditation Commission (IAC) accredited vein center between December 2018 and February 2019. All patients ≥18 years of age who presented to the venous clinic with signs or symptoms of venous reflux disease, including venous pain, RLS, deep venous thrombosis (DVT), phlebitis, or ulcers, and subsequently evaluated by duplex ultrasound to assess vein competency were analyzed. Patients for whom no vein data had collected were excluded from the study. All data were deidentified prior to analysis. Our study followed the principles outlined in the Declaration of Helsinki. This minimal risk retrospective study was exempted by the Tufts Medical Center Institutional Review Board (IRB) and need for consent was waived by IRB.

Patient Variables

All information was retrospectively retrieved from the medical record. We collected patient demographics (age and gender), and comorbidities (diabetes with and with complication, coronary artery disease, myocardial infarction, hypertension, cerebrovascular disease/stroke/transient ischemic attack, neurocognition dysfunction, movement disorders [i.e., Parkinson's disease], insomnia, prior history of RLS and rate and type of prior pharmacotherapy, nondiabetic neuropathy, iron-deficiency anemia, liver dysfunction, end-stage renal disease on dialysis, depression, tobacco use [none, current, or former], narcotic dependence, marijuana dependence, and alcohol abuse). Next, we collected venous factors (history of DVT, family or personal history of venous disease including venous pain [pain worse with standing with improvement with elevation of legs] or history of lower extremity venous ulcers, as well as presence of RLS symptoms [defined as presence of nocturnal or recumbent leg pain or cramping at rest]). Baseline venous clinical severity scores (VCSS) for both legs were identified, as well as presence of physical examination findings of phlebitis and swelling, comprehensive classification system for chronic venous disorder (CEAP) scores, and objective duplex ultrasound findings. Presence of superficial and deep reflux was defined as reflux present on ultrasound greater than 0.5 seconds in the superficial venous system.

Finally, we collected details about superficial ablation including method of ablation (radiofrequency ablation, chemical, surgical, or combination therapy), duplex results (no response, resolved reflux in superficial, deep, or both venous system), as well as whether there was documented subjective resolution of RLS symptoms after ablation noted at postoperative follow-up.

Statistical Analysis

Analysis was performed comparing prevalence of RLS symptoms in those with and without SVR in our patient cohort. Next patient factors in those with or without RLS were analyzed. Categorical variables were analyzed using Chi-square test of independence and Fisher's exact test when appropriate. Continuous variables were analyzed with Wilcoxon's rank-sum test. Statistical significance was determined by p  < 0.05. Univariable and multivariable logistic regressions were constructed to examine the association between our covariates considered to be clinically relevant with presence of RLS symptoms. Inclusion into our multivariable model was based on significance in our univariable model with final inclusion confirmed using stepwise logistic regression. Those outcomes that led to unstable estimates or had no events were excluded from the logistic regression. All analysis was performed in STATA (StataCorp. 2019. Stata/MP 16, College Station, TX; StataCorp LLC) and R Package (Version 3.53; R Foundation for Statistical Computing, Vienna, Austria).

Results

In total, 207 patients met the inclusion criteria and were enrolled in the study. Of them, 137 patients (66.2%) were found to have evidence of SVR by duplex, 70 did not. Of the 137 patients with any SVR by duplex, 78.8% reported RLS symptoms as compared with 45.7% of those without SVR p  < 0.001 ( Fig. 1 ). When further stratified by type of SVR, symptoms were present in 83.7% of those with deep and superficial reflux and 76.1% of those with isolated SVR as compared with 45.7% without duplex proven venous disease ( p  < 0.001). There were no cases of isolated deep venous reflux without superficial venous system involvement.

Fig. 1.

Prevalence rate of restless leg syndrome in patients with or without superficial venous reflux/insufficiency. Net difference between groups 33.1% (95% confidence interval: 18.5–47.7%; p  < 0.001).

Table 1 describes univariate testing of patient differences in those with ( n  = 140) or without evidence of RLS ( n  = 67). There were no significant differences in demographics or comorbidities between the two groups ( p  < 0.05) apart from 25 (17.9%) of those with RLS symptoms having had a previous diagnosis of RLS. Of those 25 patients, only 2 were currently on pharmacotherapy for their RLS which consisted of gabapentin in 1 case, while the other was unknown.

Table 1. Baseline demographics and comorbidities of overall cohort and between those with or without restless leg syndrome symptoms.

Variable	Overall cohort ( n  = 207)	RLS ( n  = 140) (67.63%)	No RLS ( n  = 67) (32.37%)	p -Value
Demographics
Age (y)  Median (range)	62 (25–89)	62 (28–89)	62 (25–89)	0.524
Gender (female)  n (%)	174 (84.1)	122 (87.1)	52 (77.6)	0.080
Comorbidities n (%)
Any diabetes	30 (14.5) a	24 (17.14) a	6 (9.0) a	0.142
Coronary artery disease	8 (3.9)	6 (4.3)	2 (3.0)	>0.999
Myocardial infarction	2 (1.0)	0 (0.0)	2 (3.0)	0.104
Hypertension	53 (25.6)	39 (27.9)	14 (20.9)	0.312
CVA/stroke/TIA	7 (3.4)	3 (2.1)	4 (6.0)	0.217
Neurocognitive dysfunction	0	0	0	–
Insomnia	26 (12.6)	19 (13.6)	7 (10.5)	0.656
Prior history of RLS	25 (12.1)	25 (17.9)	0 (0.0)	<0.001
If yes, on pharmacotherapy	2 (1.0) b	2 (1.4) b	0 (0.0)	–
Nondiabetic neuropathy	8 (3.9)	7 (5.0)	1 (1.5)	0.442
Movement disorders	2 (1.0)	2 (1.4)	0 (0.0)	>0.999
Iron-deficiency anemia	1 (0.5)	1 (0.7)	0 (0.0)	>0.999
Liver dysfunction	0	0	0	–
End-stage renal disease on dialysis	0	0	0	–
Depression	9 (4.4)	8 (5.7)	1 (1.5)	0.277
Tobacco use
None	182 (88.4)	121 (86.4)	61 (92.4)	0.411
Former	13 (6.3)	11 (7.9)	2 (3.0)
Current	11 (5.3)	8 (5.7)	3 (4.6)
Narcotic dependence	1 (0.5)	1 (0.7)	0 (0.0)	>0.999
Marijuana use	0	0	0	–
Alcohol abuse	1 (0.5)	0 (0.0)	1 (1.5)	0.324

Abbreviations: CVA, cerebrovascular accident; RLS, restless leg syndrome; TIA, transient ischemic attack.

^a Five cases complicated diabetes; n  = 3 for those with restless leg syndrome; n  = 2 no restless leg syndrome.

^b In n  = 2 on prior medication for restless leg syndrome; n  = 1 includes gabapentin n  = 1 unknown.

Table 2 describes the baseline venous factors between those with or without RLS symptoms. Those who reported RLS symptoms had higher rates of venous pain (99.3 vs. 82.1%; p  < 0.001), phlebitis (39.3 vs. 25.4%; p  = 0.049), were more likely to have a family history of venous reflux (86.7 vs. 62.7%; p  < 0.001), and finally had higher rates of overall SVR (77.1 vs. 43.3%), isolated SVR (47.9 vs. 31.3%), and combined SVR and deep venous reflux (29.3% vs. 11.9%; all p  < 0.05).

Table 2. Vascular-specific comorbidities, symptoms, and clinical and radiologic findings between those with or without restless leg syndrome symptoms.

Variable	Overall cohort ( n  = 207)	RLS ( n  = 140) (67.63%)	No RLS ( n  = 67) (32.37%)	p -Value
Relevant vascular history n (%)
Venous pain present	194 (93.7)	139 (99.3)	55 (82.1)	<0.001
History of DVT	7 (3.4)	6 (4.3)	1 (1.5)	0.432
History PVD	31 (15.0)	24 (17.1)	7 (10.5)	0.298
Prior ablation	20 (9.7)	15 (10.7)	5 (7.5)	0.485
History of PAD	3 (1.5)	2 (1.4)	1 (1.5)	>0.999
Family history of venous insufficiency	163 (78.7)	121 (86.4)	42 (62.7)	<0.001
Hx of lower extremity venous ulcer	32 (15.5)	19 (13.6)	13 (19.4)	0.307
Exam findings n (%)/median (range)
Phlebitis present	72 (34.8)	55 (39.3)	17 (25.4)	0.049
Swelling present	130 (62.8)	99 (70.7)	31 (46.3)	0.001
VCSS score, right	9 (2–18)	9 (2–18)	9 (2–18)	0.183
VCSS score, left	9 (4–18)	9 (4–14)	9 (7–18)	0.854
CEAP score left	4 (1–6)	4 (1–6)	3.5 (1–6)	0.587
CEAP score right	4 (1–6)	4 (1–6)	2.5 (1–6)	0.331
Duplex Findings n (%) a
Any SVR	137 (66.2)	108 (77.1)	29 (43.3)	<0.001
Isolated SVR	88 (42.5)	67 (47.9)	21 (31.3)	0.025
Both deep venous reflux and SVR	49 (23.7)	41 (29.3)	8 (11.9)	0.006

Abbreviations: CEAP, clinical-etiological-anatomical-pathophysiological; DVT, deep venous thrombosis; Hx, history; PAD, peripheral arterial disease; PVD, peripheral vascular disease; RLS, restless leg syndrome; SVR, superficial venous reflux; VCSS, venous clinical severity score.

^aNo cases of isolated deep venous.

Of the 140 patients with RLS symptoms, 101 underwent superficial venous ablation most commonly done by surgery (34.7%), followed by isolated radiofrequency ablation (RFA) (27.7%) or chemical ablation (15.8%), or a mixture of RFA and surgery (11.9%); the remainder of treatment methods including combination therapies was all <5%. Postprocedure duplex details were available for 90 of the 101 patients (89.1%), with 85 (94.4%) found to have successful resolution of reflux with 52.5% experiencing resolution of their SVR, 33.7% experiencing resolution of both their superficial and deep venous reflux, and 13.9% experiencing resolution of only deep venous reflux. Of those successfully ablated, 73 (85.9%) reported RLS symptom resolution (including 12 out of 15 patients with prior RLS diagnosis). Furthermore, all patients with confirmed RLS symptom resolution were noted to have successful ablation results ( p  < 0.001). We found no statistical difference between treatment modalities, primary reflux location (isolated SVR or DVR and SVR), and resolved system (superficial, deep, and both) with RLS symptomatic resolution ( p  > 0.05).

The results of our univariable ( Table 3 ) and multivariable logistic regression ( Table 4 ) showed SVR, and family history of venous reflux was the only factor independently associated with presence of RLS symptoms (both p  < 0.05). While significant in our univariable logistic regression, deep venous reflux and venous pain were correlated with SVR and therefore excluded during our stepwise logistic regression.

Table 3. Univariable logistic regression for restless leg symptoms.

Variables	Odds ratio	95% CI	p -Value
Age (y)	0.99	0.97–1.01	0.379
Gender, male (ref. female)	0.51	0.24–1.09	0.083
Diabetes (ref. no.)	2.10	0.82–5.42	0.124
Coronary artery disease (ref. no.)	1.46	0.29–7.41	0.651
Hypertension (ref. no.)	1.46	0.73–2.93	0.284
CVA/stroke/TIA (ref. no.)	0.34	0.07–1.59	0.172
Insomnia (ref. no.)	1.35	0.54–3.38	0.527
Nondiabetic neuropathy (ref. no.)	3.47	0.32–28.82	0.249
Depression (ref. no.)	4.0	0.49–32.66	0.196
Tobacco use (baseline no.)
Former	2.77	0.60–12.90	0.194
Current	1.34	0.34–5.25	0.670
Deep Vein thrombosis (ref. no.)	2.96	0.35–25.05	0.320
Peripheral Vascular disease (ref. no.)	1.77	0.72–4.35	0.211
Peripheral Arterial disease (ref. no.)	0.96	0.09–10.74	0.871
Family history of venous reflux (ref. no.)	3.79	1.90–7.57	<0.001
Hx of lower extremity venous ulcer (ref. no.)	0.65	0.30–1.42	0.280
Phlebitis present (ref. no.)	1.90	0.99–3.63	0.051
Swelling present (ref. no.)	2.80	1.53–5.12	0.001
VCSS score left	0.91	0.78–1.07	0.257
VCSS score Right	1.05	0.91–1.21	0.500
CEAP score left	1.06	0.83–1.36	0.647
CEAP score right	1.09	0.85–1.41	0.490
Superficial venous reflux (ref. no.)	4.42	2.37–8.25	<0.001
Deep venous reflux (ref. no.)	3.05	1.34–6.96	0.008

Abbreviations: CVA, cerebrovascular accident; CEAP, clinical–etiological–anatomical–pathophysiological; CI, confidence interval; TIA, transient ischemic attack; VCSS, venous clinical severity score.

Table 4. Multivariable stepwise logistic regression of variables associated with restless leg symptoms.

Variable	Odds ratio	95% confidence interval	p -Value
Superficial venous reflux (ref. no.)	3.90	1.96–7.74	<0.001
Leg swelling (ref. no.)	1.93	0.98–3.81	0.058
Family history of venous reflux (ref. no.)	4.12	1.92–8.87	<0.001

Discussion

In our patient population, in those for whom clinical suspicion of symptomatic venous disease was present and resultant venous duplex examination was performed, we found a significantly higher prevalence of RLS symptoms in patients with SVR than in those without identified reflux. Additionally, in those patients seen for venous complaints, SVR was the only objective independent factor statistically associated with RLS symptoms. Our results provide new data on rates of RLS symptoms in patients with underlying SVR while corroborating the limited studies which have previously identified a similar association. ^{4 5 17} Furthermore, the high rate of RLS symptoms in those who found to have SVR suggests an association between the two factors. Therefore, in patients presenting with RLS symptoms, we believe SVR should be actively screened for. This conclusion stands in stark contrast to the common practice of eliciting night-time symptoms in an effort to exclude venous disease as the etiology of leg symptoms.

Although no pathophysiological explanation for the association of RLS symptoms and SVR has been objectively proven, Shah et al recently noted an association of RLS and the lateral subdermal plexus of veins, with swelling and inflammation resulting in irritation of the lateral ileal-tibial band, causing nocturnal pain and cramping. ⁵ Fronek et al, in a similar report of 371 patients, examined in their vein clinic found that 92% had venous reflux and 49% had nocturnal symptoms (cramping or RLS) but they also found no pattern of refluxing segments associated with nocturnal symptoms. ⁶

Furthermore, 86% of those reporting RLS symptoms had subjective improvement after duplex proven ablation. This was notable, as it included a portion of those previously diagnosed with RLS. These findings enforce our recommendation to pursue upfront venous evaluation in those with RLS symptoms. These findings, while limited, are consistent with existing literature on RLS and SVR. ^{4 5 17}

Identifying that SVR associated with RLS symptoms is significant as it provides the possibility of more definitive interventions (i.e., venous ablation of the incompetent superficial venous system by methods including but not limited to radiofrequency, chemical, or surgical ablation) as an alternative to medical management alone. Current estimates of medical management for RLS range from $774 to 2,448 annually ^{14 16} which makes the typical one-time cost of approximately $2,040 for radiofrequency ablation, a potentially low risk, cost-effective option for treatment. ¹⁸

Limitations

The limitations of our study include the retrospective nature of our collection. Our sample represents only those with venous complaints and the true incidence of SVR in patient with RLS symptoms is unknown. Additionally, we lacked a standardized scoring system for RLS symptoms to stratify our patient cohort or their symptomatic improvement. Furthermore, we were unable to report development of complications associated with ablation, nor were we powered enough to examine differences in treatment modality and outcome. Therefore, these latter findings should be interpreted with caution.

Our results suggest future prospective studies are warranted to examine prevalence of SVR in those diagnosed with RLS by a validated metric. Furthermore, future studies should attempt to quantify the impact of ablation and different methods, such as surgery, chemical ablation, or radiofrequency ablation, on RLS-like symptoms in those with accepted indications for treatment to determine if a true net benefit in treatment exists.

Conclusion

The presence of RLS symptoms as defined in our study was correlated with underlying SVR disease. Further workup with duplex ultrasound in those with RLS symptoms may therefore be indicated as part of the initial evaluation. Future study is still required to identify the applicability of this recommendation to the overall U.S. population and to determine whether treatment of SVR benefits patients with RLS in prospective trial.

Funding Statement

Funding None.