We found that all of our Ethiopian PD patients reported some sleep problem, with a large minority (42.6%) reporting not sleeping well > 2 nights per week. One community based study from Norway reported that 60% of their PD patients had sleep problems .
Compared to patients from Germany  and the United Kingdom  in terms of overall sleep disturbance, there was a higher percentage of Ethiopian patients with a severe sleep disorder. Trenkwalder et al.  reported a mean PDSS score of 16.5- similarly to our mean score of 18.3 (median 17). However, they found that only 6.3% of their patients had a score > 30, whereas in our population, 23.2% of our patients had a score > 30 .
Nocturia and mobility difficulties were the most frequent sleep problems in our population. Other authors have found similar issues. Nocturia was reported by 62% of patients in the NMS Quest Study . Adler et al. reported that 80% of patients with PD have two or more episodes of nocturia per night, and 33% urinate at least three times per night . Lees and his colleagues  have reported nocturnal disturbances in 215 of 220 PD patients, including nocturia (79%) and difficulty turning over in bed (65%).
Over a third of our patients (36.1%) reported having distressing dreams. Nightmares have been reported in 30% of patients with PD and are correlated with disease severity and levodopa dose .
Insomnia occurs in about 30% of patients with PD.  Patients often develop a sleep pattern marked by excessive napping during the day and wakefulness at night . We found sleep onset insomnia ≥2 days/week in 47 PD patients (30.3%) and sleep maintenance insomnia ≥2 days/week in 54 PD patients (34.9%). This is comparable with one study from India. Kumar et al. reported the prevalence of insomnia in PD patients were 30% .
Excessive day-time sleepiness (EDS) is a common complaint of patients with PD. [28, 29] It can occur early in PD , and may predate the diagnosis . We found that 47.1% of our patients had possible or definite EDS. This is one of the highest rates reported in the world. Possible or definite EDS (ESS > 10) was seen in 15.5% of PD patients in Norway, 33% in Austria, 40.6% in New York USA, 46.2% in France, and 50.2% in Houston,USA [3, 32,33,34,35]. Adler et al. identified that advanced disease stage and age predicted EDS . We also found an association between higher PD stage and higher ESS scores.
The number of patients using trihexyphenidyl is high (23.9%) in our study. This drug is not commonly used in western countries. However, due to cost, levodopa/carbidopa and trihexyphenidyl are the only available medications to treat PD symptoms. Therefore, trihexyphenidyl is often the first medication prescribed. Although its anticholinergic property may affect sleep, we did not find an association between trihexyphenidyl use and a higher level of sleep disturbance.
We found that high ESS scores associated with patients reporting both nocturia and breathing difficulties/snoring on the PDSS. OSA is defined as intermittently absent or reduced airflow during sleep despite respiratory effort. A study from Mexico City on 120 PD patients reported obstructive sleep apnea (OSA) in 39% of patients . We found 57 PD patients (36.8%) reporting OSA symptoms at least 1 day per week. A study from France on 100 patients also reported 27% of PD patents were having obstructive sleep apnea .
One limitation of our study is that we did not assess body mass index (BMI) in our patients. High BMI is a major contributor of OSA. However, the prevalence of obesity in Ethiopia is very low. A World Bank report in June 2017 estimated the prevalence of obesity in Ethiopia to be less than 5% . This was much lower than the other countries . A summary report on risk factors for non-communicable diseases in Ethiopia from 2016 reported the prevalence of obesity (BMI > 30) for male and female Ethiopians was 0.5 and 2% respectively . Another study from 2011 estimated the prevalence of obesity to be 2.1% for males and 10.2% for females. Although this study reported a higher rate among females than the others, this rate is still low, and most of our PD patients were male .
Another limitation of our study was that we did not assess for anxiety/depression. One study found an association between depression and sleep disorders , while others did not assess for depression or found no association [16, 23, 24, 31,32,33,34]. There is no validated depression scale for PD patients in Ethiopia, so we felt this was out of scope for this study.
Our study had other limitations. In Ethiopia there is no polysomnography (PSG), the gold standard for evaluating sleep disorders. Therefore, we had to rely on the PDSS-2. The PDSS is a subjective semi quantitative scale, which attempts to provide a holistic and clinical assessment of the complex etiology of sleep problems in Parkinson’s disease.
One other significant limitation of our study was our inability to assess for REM Sleep Behavior Disorder (RBD). Only 15 (9.7%) of our patients attended their clinic visits with a reliable sleep partner, so we could not use a questionnaire to evaluate for RBD, and of course, did not have access to PSG. In one study of 19 patients with PD, 47% met the diagnostic criteria of RBD based on PSG recordings, but only 33% of these cases were detected by a questionnaire . We suspect that our percent of patients with Sleep Disorders would have been higher had we had a reliable way to assess for RBD.