Restless Legs Syndrome: Prevalence and Associated Factors

Epidemiologic studies have shown conflicting evidence regarding whether the prevalence of RLS increases with age^2–20 (Table 24-1). Interpretation of age effects in such population-based work is complicated by the fact that the upper age limit varies among studies, and various studies use different definitions of RLS, with some employing face-to-face or telephone interviews and full International Restless Legs Syndrome Study Group (IRLSSG) criteria or mail questionnaires with either single or multiple items. The large variability in prevalence among studies is evident in Table 24-1. Most, but not all, suggest a female predominance of the condition, and the large variability across studies could also reflect the geographically and genetically diverse populations studied, including those from North America, northern and southern Europe, and Asia. The issue as to whether RLS shows age dependence (defined as higher prevalence simply as a function of chronological age) remains unclear. Conversely, the aging effect might also be characterized as age related (defined as a specific age span of vulnerability).²¹

Within the limits of descriptive research, epidemiologic studies such as those in Table 24-1 also allow for an appreciation of how age interacts with risk factors for RLS. Conditions commonly associated with RLS include hypertension and cardiovascular disease, relationships seen in many,^3,6,12,22 but not all,^2,14,17,18 as well as diabetes and/or peripheral neuropathy, similarly seen in many,^5,6,12,22,23 but not all,^7,10,17,18 studies. These relationships were demonstrated statistically to be independent from age, but because cardiovascular disease and deficient glucose metabolism both increase with aging, these generally positive associations have considerable relevance to the elderly population. Another condition common in old age, arthritis, has been associated with RLS in several studies^6,22 but also with some conflicting evidence.⁷ Similar conflicting evidence in population-based data has been noted for hypothyroidism^7,12 and renal disease,^7,14,17 and at least one study has reported associations between RLS and gastroesophageal reflux²² and daytime sweating²⁴ in both older men and women.

Lifestyle factors may also hold relevance as risk factors for RLS in the elderly population. For example, older persons are more likely to have a sedentary lifestyle and a number of studies have documented that lack of physical activity is associated with RLS,⁵ although curiously there are also data to suggest that more frequent exercise exerts a risk for RLS⁶ or that exercise was unrelated.^14,18 Smokers have been shown to incur a greater risk for RLS in most studies examined^{5,6,10,12,17,18,22} with relatively few conflicting data.¹⁴ Caffeine⁶ and alcohol⁶ may also represent risks for RLS, but alcohol use was also reported to be protective^5,18 or exert no influence.¹⁷ In all of these studies, it must be recalled that the assignment of putative risk is based on cross-sectional, observational findings. Longitudinal data would be required to appreciate whether a preexisting risk factor, operating in individuals without any history of RLS, might predict eventual development of the condition. It should also be stressed that although most of these studies showing relationships between RLS and a putative risk factor controlled for age, none attempted to demonstrate that the association between age and RLS was due to a particular risk factor. Such an approach would require demonstration that chronological age was no longer a significant risk when particular risk factor(s) were controlled statistically. No study has yet to present such data.

Issues of causality are nowhere more apparent than when describing associations between depressed mood and RLS.²⁵ Such findings would be particularly relevant for the geriatric population who endorse items frequently related to depression at higher rates than populations of middle-aged or younger adults. Relationships between RLS and general indices of lower mental health^5,6 or depressive symptoms^{2,3,7,10,13,17,22,26,27} have been shown in many population-based studies. The most parsimonious assumption of causality in these exclusively cross-sectional studies is not that depression predisposes for RLS but rather that RLS symptoms, if left untreated, may lead to depression in the older person. Paradoxically, another interpretative possibility is that, because virtually none of these studies has controlled for simultaneous use of psychoactive medications, including tricyclic antidepressants, selective serotonin reuptake inhibitors, or neuroleptics, which are themselves risk factors for RLS^6,7,10 and/or periodic leg movements in sleep (PLMS),²⁸ the associations could be secondary to the use of such medications.

Given the high prevalence of anemia in elderly populations, altered iron metabolism associated with RLS and aging requires special consideration, and several surveys have noted such associations (see Chapters 9 and 10).^17,22 O’Keeffe²⁹ was the first to note that elderly individuals with RLS were likely to show low serum ferritin levels. Although overt anemia and reduced serum iron may not always co-occur with RLS, neuroimaging and cerebrospinal fluid studies have suggested that total brain iron concentrations are lowered in RLS, findings that are compatible with modified blood-brain transport.^30,31 Because anemia is highly prevalent in the elderly, it is well within the range of possibility that iron metabolism moderates the high prevalence of RLS in the aged. A surprisingly small number of studies have assessed this in elderly populations. O’Keeffe³² replicated the original finding with a small additional case series suggesting that serum ferritin levels less than 50 ng/ml were significantly more likely in elderly patients with recent-onset RLS, but other population-based studies demonstrate a more complicated situation. One study demonstrated that RLS were not accompanied by low levels of serum ferritin or by higher levels of soluble transferrin receptor, but mid-range levels of serum iron and transferrin saturation appeared protective for RLS.³³ Neither serum ferritin nor hemoglobin levels less than 2 SDs below gender-expected values were significant factors in RLS in a German study in the age range of 20 to 79 years.¹² In a northern Italian elderly population (South Tyrol), however, higher soluble transferrin receptor levels (often seen in early-stage anemia) and lower serum iron levels were correlated with RLS.¹⁴ Another study showed that cerebrospinal fluid (CSF) ferritin levels in older, late-onset RLS patients were not related to onset of RLS symptoms.³⁴ That study also showed that elderly patients had higher levels of CSF ferritin than younger patients.³⁴ Taken together, these studies imply that iron deficiency may not be a relevant risk factor for RLS manifesting in elderly populations, unless the RLS also had early onset.

The complex relationship between RLS and PLMS is covered elsewhere in this volume (see Chapters 17 and 22). From the aging perspective, it is important to bear in mind that PLMS can occur in many older persons without accompanying RLS symptoms. In elderly populations, the prevalence of PLMS (without reference to RLS) has been reported to be as high as 45%,³⁵ and over 85% of older individuals had a PLMS index over 5.0 across 5 nights of recordings.³⁶ Complaints of poor and/or altered sleep architecture were unrelated to PLMS in many studies.^37–39 One study³⁵ found that PLMS were related to nocturnal leg kicking and breathing symptoms but were unrelated to many aspects of sleep disruption (e.g., lower total sleep time, prolonged sleep onset latency); the best correlate of PLMS was the estimated number of awakenings on the recording night. Other studies of older subjects found a relationship between difficulty falling asleep and PLMS⁴⁰ or a relationship with lower total sleep times and wake after sleep onset.⁴¹ By contrast, some studies have suggested no relationships between PLMS and symptoms. Mendelson⁴² could find no relationships between PLMS arousals and symptoms. Montplaisir and colleagues⁴³ compared with controls, individuals with insomnia, and individuals with hypersomnia and found no differences in PLMS among groups, and Karadeniz and coworkers⁴⁴ were unable to show changes in macrosleep or microsleep architecture in conjunction with PLMS in insomnia patients ages 40 to 64 years. Hornyak and coworkers.⁴⁵ found no relationship between PLMS and sleep quality in insomniac patients without RLS. In another study, PLMS were unrelated to polysomnographically defined sleep quality in 70 normal subjects ages 40 to 60 years, although in men a small but significant effect was reported for lower sleep quality in relation to a PLMS index greater than 10.⁴⁶ When viewed in their totality, these results present a very mixed picture as to the correlates of PLMS in old age in the absence of frank RLS.

Of final note, there is some evidence that polysomnographic features may distinguish PLMS in the elderly. For example, night-to-night variability that has been reported in elderly individuals with PLMS who do not have apparent RLS,^36,47–49 and the number of PLMS with arousals or awakenings has been reported to be higher in older age groups.^42,50,51 Finally, reduced magnitude of heart rate variability accompanying PLMS in older, relative to younger, subjects has also been reported.⁵²

Restless Legs Syndrome in Dementia

Identification of RLS in dementia patients is challenging. Because reporting of RLS symptoms requires considerable verbal facility, typically lacking in demented patients, it becomes necessary to rely on key signs, including rubbing or massaging legs, high levels of leg motor activity (such as pacing), and indicators of leg discomfort during rest or inactivity and its improvement with activity are compatible with RLS. The presence of such signs during certain times of day (i.e., late afternoon and/or early evening) are also compatible with such a diagnosis. Although not explicitly mentioned in the National Institutes of Health Conference Report, the phenomenon of wandering, a major issue in the behavioral management of dementia patients,⁵³ may represent a manifestation of RLS of considerable import in geriatric care. As a case in point, treatments for wandering typically used in dementia patients may involve dopaminergic antagonism, which might otherwise exacerbate, rather than ameliorate, the wandering behaviors (see Treatment Considerations).