Self-reported sleep disturbances in renal transplant recipients

Design, setting and sample

This study used a sequential cross-sectional multicenter design with a sample of 249 adult home-dwelling RTx patients, all of whom were participating in a larger study on sleep and daytime sleepiness. The inclusion criteria were: (1) RTx took place at one of the three participating Swiss transplant centers, (2) a functioning renal graft at least 6 months post-Tx, (3) the ability to understand and read German, (4) 18 years of age or older, and (5) participation in the preceding study with poor SQ (PSQI >5 [23]) and/or DS (ESS > 6 for increased DS [24]) scores. Candidates were excluded if they were undergoing dialysis or had not signed the written informed consent form.

The stage sampling approach used was based on candidates’ PSQI and ESS scores, both of which were assessed as a part of the larger study [8]. The PSQI is a self-rated questionnaire consisting of 19 items, assessing a wide variety of factors related to sleep quality over a 1 month period, including estimates of sleep duration and latency, and of the frequency and severity of specific sleep-related problems. These 19 items are grouped into seven component scores, each weighted equally on a 0–3 scale. The seven component scores are then summed to yield a global PSQI score, which has a range of 0–21; higher scores indicate worse sleep quality. A cut-off of > 5 points is used to classify patients as having poor sleep quality [23]. The ESS is a validated eight-item questionnaire to measure a subject’s expectation of dozing (falling into a light sleep) in eight hypothetical situations. Dozing probability ratings range from 0 (no probability) to 3 (high probability). An ESS total score ≥ 6 indicates DS [25]. A score ≥ 10 indicates that a person tends to become very sleepy and should seek medical advice [25]. All 249 provided self-reported Survey of Sleep (SOS) data; a sub-sample (n = 164) additionally participated in a sleep assessment interview (83 declined participation).

Variables and measurements

Age (in years), gender, years since transplantation, body mass index (kg/m²), creatinine (μmol/l), hemoglobin (g/l) and drugs (including sleep drugs) were retrieved from the participants’ hospital medical charts. Comorbidity data were also extracted from patients’ charts and categorized using the Charlson comorbidity index [26], which assigns various weights to specific conditions. Each of the 19 noted conditions was assigned a score of 1, 2, 3, or 6, depending on the associated mortality risk. For each patient the scores were summed to provide his or her overall comorbidity score [26]. Sleep quality and daytime sleepiness was extracted from the preceding study and categorized in three groups: 1) PSQI ≤ 5 (good SQ) & ESS ≥ 6 (DS); 2) PSQI > 5 (poor SQ) & ESS < 6 (no DS); 3) PSQI > 5 (poor SQ)& ESS ≥ 6 (DS).

Survey of sleep (SOS)

The self-reported Survey of Sleep (SOS) questionnaire was developed at the University of Pittsburgh and translated into German by the second author. It is often used to report sleep symptoms in insomnia patients, [22] and studies often employ it as a preparatory step before carrying out sleep assessment interviews [27, 28]. The questionnaire consists of 7 sections: (1) sleep overview (existence of problem(s) (yes/no), general sleep problem (main complaint); duration (less/more than 1 year), course (getting worse, same, better, irregular), and frequency of the sleep problem (once/month, several times/week, nightly)); (2) sleep habits (including bedtime, get-up time and sleep latency in hours and minutes, whether the subject sleeps better in another location (yes/no), regularity of bedtimes (yes/no); (3) sleep disturbances (sleep-related symptoms and a list of 45 potential disturbances); (4) daytime function (typical feelings on getting up (energetic, optimistic, refreshed, low energy, irritable, depressed, confused, anxious); nap behavior (intentional or unintentional naps, dreaming during the naps (yes/no), feeling more alert after the nap (yes/no), daytime function (sleepiness (not at all, slightly, moderately, extremely), accidents because of sleepiness (yes/no), fatigue (not at all, slightly, moderate, extremely), having to close eyes during the day to relax (yes/no), impaired daytime function (yes/no), most functional period of the day (early or late morning, afternoon or evening; night; no particular time), (5) health habits (use of sleeping drugs (Yes/No), caffeine (amount in cups), nicotine (number of cigarettes per day), alcohol use (glass unit per day), (6) sleep history (select the main complaint); and (7) medical history (diagnoses, drugs) [29].

The estimated time necessary to complete the SOS is 30 minutes. There is no sum scoring of the items and as of the time of writing no validity or reliability measures are available for it, as it was developed as a guide for an sleep assessment interview and not as a diagnostic tool [22]. The complete Survey of Sleep (SOS) questionnaire is available on request from the second author.

Sleep assessment interview

Data from the SOS were used to prepare and structure the sleep assessment interview. All responses indicating possible sleep disturbances were addressed and elaborated on in the interview, which was structured to follow the 7 SOS sections, and lasted approximately one hour. The information generated by the interview helped to exclude some sleep disorders; however, as no follow-up visits took place and no further sleep diagnostic measurements or tools were used, the given diagnoses according to the ICSD criteria [20] should be regarded as preliminary.

The interviewer (first author) was trained to perform sleep assessment interviews by a certified sleep specialist and somnologist at the Hirslanden Sleep Disorders Center in Zollikon, Switzerland. This training included an overview of sleep disorders and of the techniques used to diagnose them. The second author checked a random sample of the completed interview transcripts and evaluated the comprehensive justification (to provide inter-rater reliability) of the preliminary sleep diagnoses. He also provided back-up assistance in view of resolving difficulties in assessment or categorization of sleep disorders.

Data collection

Patients were informed at the start of the research project [8] that they might be invited for a further screening and assessment if their initial data indicated poor SQ and/or DS (see flowchart, Figure 1). Each such patient received a package containing an information letter, informed consent documents, a pre-stamped return envelope and the Survey of Sleep questionnaire (SOS). Candidates were included in the study if they signed the informed consent form, completed the SOS and returned the documents.

Data collection started in June 2011 at the first transplant center and ended in June 2012 at the third. Patients who had not responded within 2 months of the document mailings were contacted by phone to ask whether they had received the material and would still be willing to complete the questionnaire. Each eligible patient (N = 249) was contacted to set up a sleep assessment meeting, which could be conducted either in person or via telephone. Only 48 agreed to in-person interviews; 116 agreed to a phone interview. After 10 unsuccessful call attempts, the patients were categorized as unreachable or it was noted that they had declined to participate (n = 85). According to each participant’s wishes, the first author either met him/her at a predetermined place or called at the predetermined time.

The study was approved by the ethics committees of all three transplant centers (Ethikkommission beider Basel; Kantonale Ethikkommision Bern; Kantonale Ethikkommision Zürich). Data were anonymized following the interview and stored in an electronic databank. Participants given preliminary diagnoses were encouraged to consult their nephrologists regarding their sleep problems. Any patient who wished also received a list of certified sleep disorder centers in Switzerland for further examination and treatment.

Statistical analysis

Descriptive statistics (means, standard deviations (SD), medians, quartiles, and frequencies) were used as appropriate, based on measurement levels and variable distributions. Likewise, comparisons between respondents and non-respondents were performed via t-test, Goodman and Kruskal's gamma test, or Mann–Whitney U test. Missing values were left blank and analysis was performed on the values given. SPSS® Statistics software (Version 19.0.0, IBM Corporation, Somers NY) was used for statistical analysis, with all critical probability levels set to 5%.

Prevalence and percentage of sleep problems and sleep habits [SOS part 1 & 2]

Prevalence and percentage of sleep symptoms [SOS part 3]

Prevalence and percentage of daytime function [SOS part 4]

Prevalence and percentage of preliminary sleep diagnoses according to the ICSD

Prevalence and percentages of sleep problems and sleep habits [SOS part 1]

The most prevalent sleep problem was difficulty staying asleep, followed by problems falling asleep. Both are characteristic of insomnia [20]. Other characteristics of insomnia common in this group were the extended duration of the sleep problem (61.4% reported durations greater than 2 years), the severity of the sleep problem (26.9% called their problems severe), the high prevalence of nightly sleeping pill use (32.9%), sleep latency of 28 ± 19.3 minutes, a high number of awakenings (2.8 ± 1.8) per night, long sleep latency after awakening (21.9 ± 16.4 minutes), and high ratios of time in bed to hours of sleep (8.3 ± 1.3 hours) to hours of sleep 6.4 ± 1.5. These results corroborate those of Moul et al. (2002), [30] who reported that 68% of insomnia patients exhibited long-term sleep problems (more than 1 year), severe sleep problems (81%), high nightly use of sleeping pills (89%), long sleep latency (53.3 ± 51.8 minutes), high numbers of awakenings (2.7 ± 1.7) per night, long sleep latency after awakening (56.0 ± 64.7 minutes), and high ratios of time in bed to hours of sleep (8.2 ± 1.9 hours in bed: 5 ± 1.7 hours of sleep). The average sleep duration of 6.4 ± 1.5 is very low, as studies have shown that chronic restriction of sleep to 6 h or less per night produces cognitive performance deficits equivalent to up to 2 nights of total sleep deprivation [31]. Sleep deficits seriously impair waking neurobehavioral functions (lapses in behavioral alertness) in healthy adults [31].

Prevalence and percentages of sleep habits [SOS part 2]

One third of participants (n = 82) reported using sleeping pills; however, the medical chart data showed that very few (n = 9) had informed their nephrologists regarding their sleep problems or use of sleep medication. During their post-transplant hospitalization, all RTx recipients receive education regarding over-the-counter medication and medication prescriptions from other physicians, during which they are advised always to consult their nephrologist about possible interactions with their immunosuppressive drugs [32]. This discrepancy may indicate that patients are reluctant to bring up sleep problems, that they do not see sleep disorders as a topic that nephrologists can deal with, or that the nephrologists themselves simply consider sleep disorders a normal side effect of RTx immunosuppressive regimens. Compared to the general population, our prevalence of 32.9% self-reported sleep medication use is very plausible: sleep medications are used regularly by 3.2% of subjects 44 or younger, 13.3% of subjects between 45 and 64, 22% of those between 65 and 74 and 32% of individuals 75 or older [33].

Prevalence and percentage of sleep symptoms [SOS part 3]

The most prevalent night-time symptom was nocturia. The frequency of its occurrence is key to further diagnosis. Nocturnal polyuria (nocturnal urine overproduction) and diminished nocturnal bladder capacity [34] require further testing to exclude urinary tract infections and prostate hyperplasia [35]. Also very prevalent were leg cramps and frequent turning in bed, indicating muscle fatigue, nerve dysfunction or electrolytic imbalances [36]. However, these symptoms could also be indices of restless leg syndrome, periodic limb movements, myositis, or peripheral neuropathy [36]. Similarly, turning or rocking in bed could indicate parasomnia (undesirable physical or behavioral phenomena occurring during the sleep period) [37]. For the diagnosis of parasomnias a careful physical examination is crucial and often a polysomnogram, including an expanded electroencephalographic montage, is necessary to distinguish between parasomnias (non-REM or REM) and nocturnal seizures [37].

Leg cramps during sleep were the second most prevalent sleep symptom (37.8%), followed by frequent tossing and turning in bed (37.1%). These two symptoms could be related to restless leg syndrome and/or periodic limb movements. The prevalence of restless leg syndrome in RTx recipients overall is 4.5% [38]. For periodic limb movements the overall prevalence is unknown, although there is an improvement from pre- to post-Tx [39]. Nocturnal leg cramps are often associated with vascular disease, lumbar canal stenosis, cirrhosis and hemodialysis [36], however no prevalence is known for RTx recipients. The sensorimotor symptoms of restless leg syndrome and/or periodic limb movements can be treated with dopamine agonists, gabapentin and its derivatives, and opioids [40]. To summarize, in-depth assessment of all these listed symptoms is crucial for the right treatment choice.

Prevalence and percentage of daytime function [SOS part 4]

Table 4 shows the high prevalence of daytime sleepiness, tiredness and impaired daytime functioning, highlighting the importance for affected patients to use reminders (e.g., pillbox alarms, SMS reminder functions, or other cues) to ensure punctual intake of their immunosuppressive drugs. An earlier study showed correlations between DS and impaired immunosuppressive medication adherence [Burkhalter H, Wirz-Justice A, Cajochen C, Weaver T, Steiger J, Fehr T, Venzin R, De Geest S: Daytime sleepiness is associated with immunosuppressive non-adherence in renal transplant recipients: a cross-sectional multi-center study. Submitted]. However, it is possible that compensating behaviors such as increased use of mild stimulants (e.g., caffeine, nicotine) (Table 1) account for the lower prevalence of non-adherence (16%) than of DS (52%) [41].

Napping behavior and sleep duration depends on cultural, environmental, occupational and health factors [42]. In this study, 47.4% of participants reported intentional napping, a behavior shown to be protective against mortality [42]. However, a nap lasting several hours [43] might interfere with nighttime sleep–a point which would have to be borne in mind while counseling patients regarding sleep hygiene. The ideal nap duration for adults is about 10–20 minutes and the timing depends on the quality of sleep duration the preceding night, amount of prior wakefulness and morningness-eveningness tendencies [44].

Prevalence and percentages of preliminary sleep diagnoses

This study’s most prevalent sleep diagnosis was chronic insomnia, followed by circadian rhythm sleep disorders. The prevalence of insomnia in the general population is 15-20% [45] and prevalence of circadian rhythm sleep disorders ranges from 3.1% in adults aged 40–64 to 16% in adolescents [46]. Our prevalence of 42.6% insomnia and 20.1% CRSD is only partially comparable based on our group’s pre-selection criteria (RTx recipients having poor SQ and/or DS). Various publications suggest RTx recipients’ sleep disorders are related to medications (e.g., β-blockers [47], nonsteroidal anti-inflammatory drugs [48], corticosteroids [49] and mycophenolic acid [50]) and other clinical conditions [51, 52]. Molnar et al. [53] list numerous potential causes of sleep disorders in this group, including pre-existing sleep disorders, transplant surgery, hospitalization, anxiety and uncertainty, fear of organ rejection, immunosuppressive medication, deteriorating kidney function and co-morbid medical conditions, psychosocial problems, psychiatric and neurological disturbances, lifestyle, diet, environmental factors and aging. With so many possible contributing factors, the most appropriate course of action might be a referral to a sleep expert, who could counsel the patient on the full range of behavioral and medical interventions available, and help them to choose those best suited to their needs [54]. Sleep interventions for RTx recipients are the same as for the general population, apart from the risk of interaction with immunosuppressive therapy and the need to consider the long-term side effects of their therapy (e.g., osteoporosis, new onset of diabetes, pain).

Limitation of this study

Since only 249 RTx recipients filled in the questionnaire, of which only 164 (65.9%) gave interviews, the generalizability of this study’s findings are limited. In addition, the high prevalence of RTx recipients in the “poor SQ (PSQI > 5) & and excessive DS (ESS ≥ 6)” group showing an increasing proportion along the study steps, limits the significance and comparability of the presumed sleep diagnoses.

Suggested further research

Further research will be necessary to develop safe interventions for RTx recipients with sleep-wake disturbances, taking into account their impaired renal function (limited organ survival), high risk of skin cancer (a side-effect of immunosuppressive treatment) and need to adhere to their medication regimens (high risk of acute graft rejection). These interventions should include education [55] regarding sleep disorders and their negative health impacts. Apart from established cognitive and behavioral interventions for insomnia, new chronotherapeutics treatments, particularly bright light therapy and melatonin supplementation [56] should be investigated. For RTx recipients, who already have a high number of medications to ingest daily, light therapy might be a realistic intervention to stabilize sleep-wake rhythms compared to melatonin supplementation (one more drug to ingest).