Health at a Glance

Warning View the most recent version.

Archived Content

Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please "contact us" to request a format other than those available.

Adjusting the scales: Obesity in the Canadian population after correcting for respondent bias

Statistics Canada Catalogue no. 82-624-X
by Tanya Navaneelan and Teresa Janz

[Full article in PDF]

Skip to text

Text begins

Start of text box


  • One in four adult Canadians, or about 6.3 million people, were obese in 2011–2012. Since 2003, the proportion of Canadians who were obese has increased 17.5%.
  • More men than women were obese, and obesity has increased more for men than women over the past eight years.
  • The lowest proportions of obese people were found in Canada’s three largest cities (Toronto, Montréal, Vancouver) and in areas of southern British Columbia; the highest levels were found in Atlantic Canada, the Prairies, and the Territories, and smaller cities in northern and southwestern Ontario.

End of text box

Obesity is best described as a condition in which excess body fat has accumulated to such an extent that a person’s health may be adversely affected. Obesity has become one of the world’s greatest health concerns and threatens to undo gains made in life expectancy during the 20th century.Note 1,Note 2 An extensive body of research has found associations between excess body weight and numerous chronic conditions, including type 2 diabetes, hypertension, cardiovascular disease, gallbladder disease and certain types of cancer. Nevertheless, the amount of excess fat, its distribution throughout the body, and the associated health consequences, can vary considerably between individuals.Note 3,Note 4 Despite cultural norms that stigmatize excess weight, and strong evidence of its adverse health effects, the prevalence of obesity continues to rise.Note 5

This paper presents obesity estimates adjusted for certain biases in self-reported data. Adjusted estimates for adult Canadians by age, sex, and geography, that have not been previously reported, are provided using data from the Canadian Community Health Survey (CCHS).Note 6

Why adjust self-reported data?

At Statistics Canada, obesity is determined in health surveys using the body mass index (BMI), a relative measure of weight and height (see About the body mass index). BMI can be computed using self-reported values, where the respondent is asked their height and weight, or by directly measuring respondents’ height and weight.

Although directly measured data provide more accurate estimates of obesity, it is more costly and time-consuming to gather. Gathering measured data means interviewers require special training, and people may be less likely to participate because they find it more intrusive.

Self-reported data is less expensive and easier to gather than measured data: this is beneficial when sampling large numbers of people. However, self-reported data is subject to respondent biases—people may not know their height or weight or their response may reflect perceived social and cultural norms about the ideal height and weight. Consequently, people tend to underestimate their weight and overestimate their height, resulting in an underestimation of the prevalence of obesity.Note 7,Note 8,Note 9

Correction equations were developed so that self-reported data, which offers the advantage of large sample size, could be adjusted for respondent biases to approximate measured estimates, which are more accurate.Note 10

Start of text box

The correction equations

The correction equations used in this article were developed using the 2005 Canadian Community Health Survey. This survey included a sample of respondents whose height and weight were collected using both self-reported and measured data. These results were then compared to assess the level of bias between self-reported and measured data. The resulting correction equations were published in:

“The feasibility of establishing correction factors to adjust self-reported estimates of obesity” by Sarah Connor Gorber, Margot Shields, Mark S. Tremblay and Ian McDowell, Health Reports, September 2008, Statistics Canada Catalogue no. 82-003-X

The Gorber et al. article presents four possible correction methods. Because the bias differs between the sexes, each possible method produced different equations for males and females. This paper uses the ‘Reduced Model 4’ equations, as recommended by Gorber et al.

End of text box

Measured, unadjusted self-reported, and adjusted self-reported estimates of obesity are shown in Chart 1. Measured data resulted in the highest estimates of obesity. Unadjusted self-reported data yielded the lowest estimates—seven to eight percentage points lower than the measured estimates. Adjusting this self-reported data produced national estimates more in line with the measured estimates.

Chart 1 Prevalence of obesity, by type of estimate: unadjusted self-reported, adjusted self-reported, and measured; household population aged 18 to 79, Canada, 2003 to 2011–2012

Description for Chart 1

Highest levels of obesity found in men and in the middle-aged

The adjusted prevalence of obesity among CanadiansNote 11 aged 18 and over was 24.8% in 2011–2012. This means that one in four adult Canadians, or approximately 6.3 million people, were obese, 17.5% more than in 2003.

During 2011–2012, overall obesity levels were higherNote 12 for males, 26.1%, than for females, 23.4% (Chart 2). Males 35 and older had higher levels of obesity than females in that age range. However, among those aged 18 to 34, there were no differences in obesity between the sexes.

Over time, obesity has increased more for men than women. Between 2003 and 2011–2012, the prevalence of obesity rose 17.9% among men and 16.8% among women.

Chart 2 Prevalence of obesity, adjusted self-reported, by age group and sex, household population aged 18 and over, Canada, 2011–2012

Description for Chart 2

Age is also related to obesity. Considering both sexes together, those aged 18 to 34 were significantly less likely to be obese than any other age group. The middle-aged, those aged 35 to 64, were the most likely to be obese.

Obesity can be explored further by dividing it into three classes: Class I – BMI of 30.0 to 34.9; Class II – BMI of 35.0 to 39.9; and Class III – BMI of 40.0 or more (see About the body mass index). Like overall obesity, the prevalence of Class III obesity, the level associated with the highest level of health risk, has increased, from 1.8% in 2003 to 2.5% in 2011–2012.

While a greater proportion of men were obese, women were more likely to be Class III obese: 3.0% of obese women were Class III compared to 2.0% of obese men. This reflects the fact that obese women, on average, had higher BMIs than obese men. The average BMI among obese women was 34.8; among obese men, 33.9.

Obesity in British Columbia and Quebec lower than national level

A major advantage of adjusting self-reported data is that it enables Statistics Canada to gather observations from larger samples of individuals. Larger samples are needed to produce estimates for smaller geographic areas, such as provinces and health regions.Note 13 It is not practical to collect such large samples of measured data,Note 14 but adjusting self-reported data can yield results that approximate measured data.

Across the country, the prevalence of obesity in the provinces varied greatly in 2011–2012 (Chart 3). Two provinces stand out for having the lowest levels of obesity—British Columbia, 20.4%, and Quebec, 22.8%.

Provinces/territories where obesity levels were higher than the national average, were

  • Northwest Territories, 35.3%
  • Newfoundland and Labrador, 35.2%
  • New Brunswick, 33.2%
  • Nunavut, 33.0%
  • Prince Edward Island, 32.4%
  • Nova Scotia, 32.3%
  • Saskatchewan, 31.6%
  • Manitoba, 27.7%.

The prevalence of obesity in Ontario, Alberta and Yukon did not differ from the national average. Previous research has found similar patterns of obesity distribution throughout the country.Note 15,Note 16

Broken down by sex, the data showed this same provincial pattern with one exception: the prevalence of obesity among women in Quebec, 22.3%, was not statistically different from the prevalence for all Canadian women. The lower level of obesity in Quebec appears to be the result of less obesity among the province’s men, 23.3%, compared to Canadian men in general.

Chart 3 Prevalence of obesity, adjusted self-reported, by province/territory, household population aged 18 and older, Canada, 2011–2012

Description for Chart 3

Less obesity in health regions containing major cities and in southern British Columbia

Provinces can be further divided into smaller geographic areas such as health regions. At this level of geography the variation in obesity was even greater, ranging from a low of 11.3% to a high of 40.8% (see Appendix A for a list of obesity prevalences by health region).

Start of text box

Health regions are health administrative areas defined by provincial ministries of health. For complete Canadian coverage, each northern territory represents a health region. There were 110 health regions in 2012.Note 17

End of text box

The lowest levels of obesity were in health regions that contain Canada’s three largest cities or their surrounding areas: Région de Montréal, 19.9%; York Regional Health Unit, 19.0%; City of Toronto Health Unit, 17.7%; and Vancouver Health Service Delivery Area, 11.3%. These obesity levels were all lower than the national prevalence of 24.8%. The remaining health regions with lower prevalences were all in British Columbia: South Vancouver Island Health Service Delivery Area, 20.1%; Richmond Health Service Delivery Area, 13.0%; and North Shore/Coast Garibaldi Health Service Delivery Area, 12.4%.

In contrast, 51 health regions had obesity levels that were higher than the national average (Figure 1 and Appendix A). The highest levels tended to be found in mostly rural health regions in the Atlantic and Prairie Provinces. The five health regions with the highest estimates were

  • Zone 7 (Miramichi area), New Brunswick, 40.8%
  • Mamawetan/Keewatin/Athabasca, Saskatchewan, 40.3%
  • Sunrise Regional Health Authority, Saskatchewan, 39.9%
  • Cape Breton District Health Authority, Nova Scotia, 39.7%
  • Northern Regional Health Authority, Manitoba, 38.9%.

Lower than average obesity in Canada’s three largest CMAs and other CMAs in British Columbia and Quebec

Obesity levels also varied considerably when using another type of geography—census metropolitan areas (CMAs)—to analyze patterns of obesity (see Table 1). These trends tended to mirror the results found at the provincial and health region level.

Start of text box

Census metropolitan areas (CMAs) are formed by one or more adjacent municipalities centered on a population core. A CMA must have a total population of at least 100,000 of which 50,000 or more must live in the core. To be included in the CMA, other municipalities must have a high degree of social and economic integration with the core.  There were 33 CMAs in Canada in 2012.

End of text box

Levels of obesity lower than the 24.8% national average were found in the three largest CMAs—Toronto, Montreal, Vancouver—and other smaller CMAs in British Columbia and Quebec, specifically

  • Kelowna, British Columbia, 17.0%
  • Vancouver, British Columbia, 17.4%
  • Victoria, British Columbia, 19.6%
  • Toronto, Ontario, 20.2%
  • Québec, Quebec, 20.9%
  • Montréal, Quebec, 21.5%.

In comparison, obesity levels tended to be higher than the national average in CMAs located in the Atlantic Provinces and in northern and southwestern Ontario, specifically in

  • Saint John, New Brunswick, 38.1%
  • Greater Sudbury, Ontario, 33.8%
  • St. John’s, Newfoundland and Labrador, 33.2%
  • Brantford, Ontario, 32.1%
  • Hamilton, Ontario, 31.3%
  • Saskatoon, Saskatchewan, 31.3%
  • Thunder Bay, Ontario, 30.7%.

This paper provides a first overview of adjusted self-reported obesity estimates for various levels of geography. While these numbers do not account for differences in the demographic composition of geographic areas, previous research has found that the demographic structure of a population can be related to its prevalence of obesity.Note 18,Note 19 These adjusted estimates may be used in future research to more fully explore disparities in obesity between geographic areas.


Estimates of obesity based on self-reported data tend to be lower than estimates based on measured data because of biases in how people report their weight and height. Self-reported data, which is less expensive and easier to obtain than measured data, can be adjusted to more closely reflect measured values.

Adjusted estimates for 2011–2012 show that the people most likely to be obese were males, those aged 35 to 64, and people who lived in the Prairies, the Territories, Atlantic Canada, or smaller cities in northern and southwestern Ontario. Conversely, Canadians aged 18 and 34, and people living in the three largest CMAs (Toronto, Montréal, Vancouver), as well as in southern British Columbia, were the least likely to be obese.

Start of text box

About the body mass index

The body mass index (BMI) is a commonly used method of assessing excess weight.Note 20 BMI is a ratio that consists of an individual’s weight relative to their height. It is calculated by dividing a person’s weight in kilograms by their height in meters squared.

BMI= weight in kilos height in metres 2 MathType@MTEF@5@5@+= feaagKart1ev2aqatCvAUfeBSjuyZL2yd9gzLbvyNv2CaerbuLwBLn hiov2DGi1BTfMBaeXatLxBI9gBaerbd9wDYLwzYbItLDharqqtubsr 4rNCHbcvPDwzYbGeaGqiVu0Je9sqqrpepC0xbbL8F4rqqrFfpeea0x e9Lq=Jc9vqaqpepm0xbba9pwe9Q8fs0=yqaqpepae9pg0FirpepeKk Fr0xfr=xfr=xb9adbaqaaeGacaGaaiaabeqaamaabeabaaGcbaGaae Oqaiaab2eacaqGjbGaaeypamaalaaabaGaae4DaiaabwgacaqGPbGa ae4zaiaabIgacaqG0bGaaeiiaiaabMgacaqGUbGaaeiiaiaabUgaca qGPbGaaeiBaiaab+gacaqGZbaabaGaaeiAaiaabwgacaqGPbGaae4z aiaabIgacaqG0bGaaeiiaiaabMgacaqGUbGaaeiiaiaab2gacaqGLb GaaeiDaiaabkhacaqGLbGaae4CamaaCaaaleqabaGaaGOmaaaaaaaa aa@57E9@

BMI ranges are classified into categories based on health risk. This paper uses the classification system adopted by Health Canada and by the World Health Organization, which allows for comparisons between populations and makes it possible to identify individuals and groups at increased health risk.Note 3,Note 4

Table 2 BMI classification by health risk
Table summary
This table displays the results of Table 2 BMI classification by health risk. The information is grouped by Category (appearing as row headers), BMI (kg/m) and Risk of developing health problems (appearing as column headers).
Category BMI (kg/m2) Risk of developing health problems
Underweight Less than 18.5 Increased
NormalNote 1 18.5 to 24.9 Least
Overweight 25.0 to 29.9 Increased
Obese – Class I 30.0 to 34.9 High
Obese – Class II 35.0 to 39.9 Very high
Obese – Class III 40.0 or more Extremely high

Limitations of BMI

Although BMI is a widely used method for assessing excess weight, it has several limitations. BMI does not directly measure body fat, nor does it take into account the distribution of fat throughout the body.Note 4 Previous research has found that abdominal fat has a greater association with morbidity and mortality than fat located in other areas of the body.Note 4,Note 21,Note 22  In addition, BMI may not work as intended for certain groups of people, including those who are still growing, very lean, very muscular, very tall, or very short, as well as for certain ethnic and racial groups.Note 4

End of text box

Tanya Navaneelan and Teresa Janz are analysts with the Health Statistics Division.

The authors wish to thank Jennifer Ali and Allan Rowell for their assistance in the production of this article.


Related material for this article

Additional information

  • For more statistics and analysis on the health of Canadians and the health care system, visit the Health in Canada module. This module is accessible from our website, under Features.

Bibliographic references

Report a problem on this page

Is something not working? Is there information outdated? Can't find what you're looking for?

Please contact us and let us know how we can help you.

Privacy notice

Date modified: