Oral Presentation ANZOS Annual Scientific Meeting 2021

A cluster analysis of Australians who self-manage their weight loss (#13)

Divya Ramachandran 1 , Ang Li 2 , Timothy Gill 1
  1. The Boden Collaboration, Charles Perkins Centre, Faculty of Medicine and Health, University of Sydney, Sydney, NSW, Australia
  2. Melbourne School of Population and Global Health, The University of Melbourne , Melbourne, VIC, Australia

Many people self-manage their weight-loss without accessing formal weight-loss programs. However, there is little understanding about who self-manages their weight loss and if they are successful.

Australian adults who were about to begin self-managed weight-loss were recruited through Facebook, to complete an online survey at baseline and 12-weeks follow-up which collected, socio-demographic, behavioural, attitudinal, and health data. Two-step cluster analysis(1, 2) was applied using gender, age-range English-speaking, IRSD, health status, initial BMI, self-management type, and presence of stress eating. Weight outcomes among identified clusters were compared.

Only 102 of 205 participants completed the follow-up survey and complete data was only available on 97 for cluster analysis. There was an over-representation of females (87.3%), married / partner (72.5%), and English speaking (85.3%). A majority (88%) had overweight (23%) or obesity (65%), and most were completely “unassisted” in their weight loss (85%).  Fifty-three (52.9%) percent had a chronic illness (including diagnosed depression -27.5%).    Participants had a mean weight loss of 2.07 kilograms at follow-up (2.07, SD 4.89, p =<.001), and a third (33, 32.4%) were successful in losing 3% or more of their initial body weight.  Four clusters were identified among self-managed weight losers i) the elderly, ill and stressed (29.9%), ii) younger aged and healthy, but poor and stressed (28.9%), iii wealthy but ill and stressed (26.8%) and iv wealthy, relaxed and healthy (14.4%). The wealthy, relaxed and healthy group had the highest proportion of successful weight losers (42.9%) whereas the younger aged and healthy, but poor and stressed had the least (28.6%).  While the findings are limited by sample size, it is evident that self-managed weight losers are not a homogenous group.  Cluster analysis is a useful technique to segment populations of interest useful for further research and can inform obesity management strategies to support self-managed weight-loss.

 

 

 

  1.   Kettenring JR. The Practice of Cluster Analysis. Journal of Classification. 2006;23(1):3-30.
  2. Tkaczynski A. Segmentation Using Two-Step Cluster Analysis. Singapore: Springer Singapore; 2016. p. 109-25.