Recently, several authors have argued that assisted dying may be ethically appropriate when requested by a person who suffers from serious depression unresponsive to treatment. We here present four arguments to the contrary. First, the arguments made by proponents of assisted dying rely on notions of "treatment-resistant depression" that are problematic. Second, an individual patient suffering from depression may not be justified in believing that chances of recovery are minimal. Third, the therapeutic significance of hope must be acknowledged; when mental healthcare opens up the door to admitting hopelessness, there is a danger of a self-fulfilling prophecy. Finally, proponents of assisted dying in mental healthcare overlook the dangers posed to mental-health services by the institutionalization of assisted dying.

Background: Studies of the use and effects of physical restraint in anorexia nervosa (AN) treatment are lacking. The purpose of this study was to describe the frequency of physical restraint in a specialized program for adolescents with AN, and to examine if meal-related physical restraint (forced nasogastric tube-feeding) was related to 5-year outcome. Method: Thirty-eight (66% of 58) patients with AN (mean age 15.9, SD = 1.9) admitted to a regional, specialized adolescent eating disorders (ED) inpatient unit. Patient data, including restraint episodes, were obtained from hospital records, and outcome was assessed at a 5-year follow-up. Results: A total of 201 restraint episodes occurred over 5513 days of inpatient treatment, including 109 meal-related episodes and 56 episodes to avoid self-harm. Twelve (32%) patients experienced at least one restraint episode during the admission, of which eight (21%) experienced meal-related restraint. Four patients represented 91% of all restraint episodes, experiencing 10 or more episodes during admission. Meal-related restraint was significantly associated with a higher rate of persisting ED diagnosis, but not with weight gain during admission, EDE-Q global score or BMI at follow-up. Conclusions: Restraint episodes occurred rather infrequently. A small number of patients (n = 4) accounted for a high proportion of episodes (91%). More knowledge is important to reduce the need for restraint in treatment for AN.