This study evaluated whether sleep disorders positively associate with selected index mental disorders; if a positive association between a sleep disorder and mental disorder differs according to whether there was one versus multiple mental disorders; and whether psychotropic drug use increases the prevalence of sleep disorders, after adjusting for selected variables.

Rates of insomnia and sleep apnea increased with age, more so for sleep apnea. Higher rates of these disorders with increasing age are consistent with previous studies [18, 19, 41]. Sex differences for sleep disorders vary, depending on the specific sleep disorder. Consistent with previous research, insomnia is more common in women, and sleep apnea is more common in men [20,21,22, 41]. Higher sleep apnea rates among married versus single adults may be because having a sleeping partner increases the chance of detecting the disorder.

Support for our first hypothesis exists because each mental disorder is positively associated with the sleep disorders. This result is consistent with previous studies [9,10,11,12,13,14,15]. Insomnia had the strongest association with bipolar disorder and schizophrenia. Insomnia may be more common in these mental disorders because of psychotic symptoms causing intense fear and anxiety. Sleep apnea had the strongest association with depression and bipolar disorder. Sleep apnea may be more strongly associated with bipolar and depression because of its bi-directional relationship [42].

We next considered whether these associations were further increased for the index mental disorder if it was associated with comorbid mental disorders. Each mental disorder had a high percentage of comorbid mental illness, as consistent with previous studies [43,44,45,46]. The mental disorders that are generally considered to have a greater impact on daily life, such as schizophrenia, OCD, and bipolar disorder, had higher levels of comorbid mental disorders, as also consistent with previous studies [24,25,26,27,28]. Supporting our second hypothesis, individuals with both index mental disorders associated and comorbid mental disorders tended to have even greater levels of insomnia and sleep disorders, as consistent with previous studies [47, 48].

Finally, in support of our third hypothesis, each psychotropic drug was positively associated with insomnia and sleep apnea, even after adjusting for age, sex, and comorbid mental disorders. This is consistent with previous research indicating a positive association directly between psychotropic drugs and sleep disorders [29,30,31,32,33,34,35,36]. Psychotropic drugs are generally more strongly associated with insomnia than sleep apnea. It is not clear from our data the extent to which the drug resulted in the sleeping disorder or the sleeping disorder led to the prescribed drug.

Psychostimulants are the most commonly used medication for treating mental disorders, except for ADHD, where amphetamine preparations are most common. Both findings are consistent with the literature [49, 50]. Anticonvulsants are almost as commonly used in bipolar patients as psychostimulants. Previous studies found that psychostimulants are most commonly prescribed for bipolar disorder, ranging from 70%-72% [41, 51,52,53]. Ataractics, also known as antipsychotics, are more commonly prescribed than psychostimulants for schizophrenia. This is consistent with current medical practice [16, 54].

Each of the mental disorders, except OCD, positively associated with insomnia, after adjusting for demographics and other mental disorders. Further adjustment for psychotropic drugs significantly decreases the association between the mental disorders anxiety, depression, and bipolar disorder with insomnia. Depression and bipolar disorder are no longer associated with insomnia in the fully adjusted model. Hence, insomnia among people with anxiety, depression, or bipolar disorder appear to be higher because of the psychotropic drugs used to treat these mental health problems. Except for CNS stimulants, all the medications are implicated, primarily sedatives (non-barbiturate) followed by psychostimulants. The relatively strong positive association of insomnia with sedatives (non-barbiturate) is a unique finding. However, the weak association with amphetamine preparations has been shown previously [31]. There is little change in the association between stress, ADHD, OCD, and schizophrenia with insomnia, indicating that medication use has little influence on the increased risk of insomnia for these mental disorders. The positive association that remains between stress, ADHD, and schizophrenia with insomnia indicates that the increased risk of insomnia is almost entirely attributed to these mental disorders and not their associated medications.

Each of the mental disorders is positively associated with sleep apnea, except OCD and schizophrenia (which have insufficient numbers to assess), after adjusting for demographics and other mental disorders. Further adjustment for psychotropic drugs significantly decreased the association between the mental disorders anxiety, depression, ADHD, and bipolar disorder with sleep apnea, to the point that the association becomes insignificant for anxiety and ADHD. Hence, psychotropic drugs explain some of the positive association for depression and bipolar disorder and all the association for anxiety and ADHD. Except for CNS stimulants, all the medications are implicated in the positive association, especially psychostimulants and then anticonvulsants. The positive association between stress and sleep apnea is not affected by psychotropic drugs.

A mental health disorder or sleep disorder diagnosis would not lead to a loss of insurance, making selective under-reporting of claims unlikely. One or more claims were used to identify the presence of each of the mental disorders. However, less serious cases, which did not lead to medical attention and resulting claims, are not represented. Hence, measures of association between mental disorders and sleep disorders may be biased downward. Finally, the current study was limited to identifying associations and not causal relationships. The causal direction between mental disorders and sleep disorders was two-way, but the extent of this bi-directional relationship could not be determined in our data. However, beyond stress, anxiety, and depression, it is likely that the mental disorders led to the sleep disorders.