1 Healthy Populations Institute, Dalhousie University, PO Box 150000, Halifax, B3H 4R2, Canada.
2 School of Health and Human Performance, Faculty of Health, Dalhousie University, PO Box 150000, Halifax, NS, B3H 4R2, Canada.
3 Faculty of Rehabilitation Medicine, Department of Occupational Therapy, Corbett Hall, University of Alberta, 11405 87 Avenue NW, Edmonton, AB, T6H 2G4, Canada.
4 Research and Policy Consultant, Rönnebergsvägen 116, 29891 Tollarp, Sweden.
5 Department of Health, Kinesiology and Applied Physiology, Concordia University, 7141 Sherbrooke Street West, Montreal, Quebec, H4B1R6 Canada.
6 Werklund School of Education, University of Calgary, 3280 Hospital Drive NW, Calgary, AB, T2N 4Z6, Canada.

Background: Stigmatization of persons living with obesity is an important public health issue. In 2015, Obesity Canada adopted person-first language in all internal documentation produced by the organization, and, from 2017, required all authors to use person-first language in abstract submissions to Obesity Canada hosted conferences. The impact of this intentional shift in strategic focus is not known. Therefore, the aim of this study was to conduct a content analysis of proceedings at conferences hosted by Obesity Canada to identify whether or how constructs related to weight bias and obesity stigma have changed over time.

Methods: Of 1790 abstracts accepted to conferences between 2008-2019, we excluded 353 abstracts that featured animal or cellular models, leaving 1437 abstracts that were reviewed for the presence of five constructs of interest and if they changed over time: 1) use of person-first versus use of disease-first terminology, 2) incorporation of lived experience of obesity, 3) weight bias and stigma, 4) aggressive or alarmist framing and 5) obesity framed as a modifiable risk factor versus as a disease. We calculated and analyzed through linear regression: 1) the overall frequency of use of each construct over time as a proportion of the total number of abstracts reviewed, and 2) the ratio of abstracts where the construct appeared at least once based on the total number of abstracts.

Results: We found a significant positive correlation between use of person-first language in abstracts and time (R2 = 0.51, p < 0.01 for frequency, R2 = 0.65, p < 0.05 for ratio) and a corresponding negative correlation for the use of disease-first terminology (R2 = 0.48, p = 0.01 for frequency, R2 = 0.75, p < 0.001 for ratio). There was a significant positive correlation between mentions of weight bias and time (R2 = 0.53 and 0.57, p < 0.01 for frequency and ratio respectively).

Conclusion: Use of person-first language and attention to weight bias increased, while disease-first terminology decreased in accepted abstracts over the past 11 years since Obesity Canada began hosting conferences and particularly since more explicit actions for expectations to use person-first language were put in place in 2015 and 2017.