Culture and Eating Disorders: Treatment Outcomes in Singapore and Hong Kong

In this last post about eating disorders in Singapore, I’ll write about the one Singapore-based retrospective outcome study in relation to a similar retrospective study conducted in Hong Kong.

In the Singapore study, researchers reviewed the charts of 94 patients diagnosed with anorexia nervosa from 1992 to 2004 at the National University Hospital, looking back from the time of the study. They didn’t contact any of the subjects for follow-up. 49 were first seen as inpatients, 34 as outpatients, and 11 were seen as outpatients but later admitted. The hospital doesn’t have a specialized ED service, so the authors relied on dietetic notes that unfortunately don’t provide a full picture of the patients’ eating disordered behaviors and cognitions.

The authors wrote about patient ‘improvement’ (not recovery!) as making a weight gain of at least 0.5 kg, or about 1 pound. 83% of their patients attended follow up appointments, which lasted for an average of about 8 months. Of these,

  • 68% ‘improved’ with an average weight gain of 6 kg.
  • 29% did not improve, losing 1.8 kg on average
  • Two patients were subsequently diagnosed with BN and BED

The authors found that patients with better outcomes had:

  • More outpatient sessions
  • Lower BMI at presentation (14.2 vs 15.7 for non-improved group), and
  • Shorter duration of disease (1 vs 3 years)

While the latter two factors would seem to indicate a more severe case, they suggest that because of this severity, clinicians devoted more attention and time to ensuring patients remained on track.

The data here is obviously very limited by its retrospective nature, and in fact the authors avoid talking about ‘recovery’ throughout. However, it’s reasonable to suggest that clinicians at NUH largely did not see their patients through the treatment process – eight months is a relatively short span of time to be in treatment, with the next study finding a median duration of 3.7 years to reach 17.5 BMI and 3 consecutive periods. Furthermore, despite some weight gain, patients at discharge (or default, i.e. they did not make further appointments) were only at a BMI of 16.4 on average. It’s also unclear whether or not many patients might have crossed over diagnostically to different behaviors. In light of the Hong Kong study I’ll look at now, a more concerted attempt at follow up would likely present a more complicated picture.

The authors of this study in Hong Kong contacted 80 patients after 4 years and assessed them on their diagnostic status and used the Morgan-Russell Outcome Assessment Schedule, a holistic evaluation tool for assessing physical markers, mental status, sexual adjustment (whatever that means), and socioeconomic status. They found that their subjects had a mean BMI of 18.5, with 67.5% experiencing regular menstruation. Diagnostically, they broke down participants down into groups as follows:

  • 45% no ED
  • 3% AN (about evenly AN-R and AN-BP; based on BMI < 16.5)
  • 20% BN
  • 8% EDNOS

A 45% recovery rate sounds pretty good – about on par with other studies – and the authors argue that 61.8% of patients had ‘good outcomes’ and only 5.3% ‘poor’ outcomes based on the Morgan-Russell scoring. This assessment rates respondents from 1 to 12, 12 being the best outcome. The scores for each diagnostic group were:

  • No ED – 9.33
  • AN – 5.41
  • BN – 7.73
  • EDNOS – 8.85

If we take the AN group as the baseline of ‘poor’ outcomes, then the other patients showed a significant amount of improvement. However, like the Singapore study, the researchers used pretty limited measures of outcome – no self-assessment of satisfaction and functioning (the diagnostic classification was made based on standardized questionnaires but no clinical interview).

Recovery seems to have occurred for about 45% of these patients, but what sort of recovery? These patients actually scored worse on the Eating Disorder Examination Questionnaire than anorexia patients (3.43 versus 2.88, and 1.41 for ‘recovered’ patients). Furthermore, a high proportion of patients continued to exhibit behaviors aside from restriction:

  • 40% exercise to lose weight
  • 3% engaged in binging
  • 20% purged
  • 10% took laxatives
  • 8% took diet pills and/or diuretics

The authors would likely argue that these issues were less likely to arise in ‘atypical’ patients who exhibited no fat phobia – 89.47% of them were classed as having a good outcome score in this study. In contrast, what the authors deemed “typical” anorexia patients, those presenting with fat phobia, showed poorer outcomes (52.63% had a ‘good’ score) and a higher tendency to cross over into bulimic behaviors. On this basis, the paper supports a continued differentiation between typical and atypical anorexia. More importantly, the authors also point out that the relatively high rates of recovery occurred despite the relative lack of formalized treatment – monthly “psychoeducational outpatient visits (roughly 10 minutes in duration” and “varying period of hospitalization with a dominant focus on weight gain” (p. 971) – rather similar to that in the Singapore study. They then raise the “controversial possibility of whether anorexia nervosa runs a ‘natural’ course independent of treatment” (p.971).

That is indeed a very controversial suggestion, and one I’m skeptical about in light of their results. Apart from questioning the validity of their measures, we don’t know how long recovered patients received treatment versus non-recovered patients and other possibly confounding factors. The authors also were unable to contact 8 patients in their follow up, three of whom had died (two committed suicide within a year of intake, while one died from emaciation after 4.7 years in treatment). Perhaps the more apparently straightforward presentation of atypical patients meant that they were more able to sustain recovery – on the other hand, they came in at intake with lower BMIs, which Lim et al suggest enabled them to receive more intensive and hence effective treatment – at least as far as weight restoration went. Furthermore, Isono (2003) highlighted the lack of understanding surrounding eating disorders, which creates an unsupportive environment for patients struggling with body image issue.

In all, both studies suggest a need to look more closely at clinicians’ ability to keep patients from defaulting on treatment and hence patients’ motivation for recovery – a motivation that treatment in both instances doesn’t seem to have emphasized very much. Unfortunately, the shortage of clinicians means that limited psychotherapy can be provided, often resulting in nutritional restoration without necessarily equipping patients with the skills to keep their recovery on track.


Lim, S. L., Sinaram, S., Ung, E. K., & Kua, E. H. (2007). The pursuit of thinness: an outcome study of anorexia nervosa. Singapore Medical Journal, 48 (3), 222-6 PMID: 17342291

Lee, S., Chan, Y., & Hsu, L. (2003). The Intermediate-Term Outcome of Chinese Patients With Anorexia Nervosa in Hong Kong. American Journal of Psychiatry, 160 (5), 967-972 DOI: 10.1176/appi.ajp.160.5.967


Jacqueline is a sociology undergraduate student from Singapore with an interest in mental health and illness, bodily work, and their intersections with the political. She is currently in recovery.