Original article
Original Article Underweight, overweight and obesity: relationships with mortality in the 13-year follow-up of the Canada Fitness Survey

https://doi.org/10.1016/S0895-4356(01)00356-0Get rights and content

Abstract

The purpose of this study was to determine the risk of all-cause mortality in the Canadian population across the new WHO/NIH BMI categories for the classification of overweight and obesity. The sample includes 10,725 adult participants (20–69 years) in the 1981 Canada Fitness Survey. A total of 593 deaths occurred during 13 years of follow-up. Hazard ratios (HR) for mortality were estimated using Cox proportional hazards models. Compared to normal weight individuals, there is an increased risk of mortality in the underweight category (HR 1.63, 95% CI 0.93–2.85) in addition to increasing levels of risk across the overweight (HR 1.16, 95% CI 0.96–1.39), obese class I (HR 1.25, 95% CI 0.96–1.65) and obese class II and III (HR 2.96, 95% CI 1.39–6.29) categories. Similar patterns were observed in sex-specific analyses. Underweight, overweight and obese Canadians are all at increased risk of mortality compared to those who are normal weight.

Introduction

Obesity is a major public health concern in North America. It has recently been estimated that approximately 4.3% of the total health care costs in the United States are attributable to obesity, after adjustment for early mortality among obese people [1]. There is considerable evidence that extreme obesity is related to an increased risk of chronic disease and death 2, 3; however, there is less of a consensus on the effects of overweight or moderate obesity on mortality [4]. Many studies have reported a J- or U-shaped relationship between body mass index (BMI) and mortality 5, 6, such that there is an increased risk of death at both low and high levels of BMI, relative to normal weight individuals. The U.S. National Institutes of Health (NIH) [3] and the World Health Organization (WHO) [7] have called for the standardization of the classification of overweight and obesity based on categories of BMI (kg/m2): underweight (<18.5), normal weight (18.5–24.9), overweight (or pre-obese) (25–29.9), obesity class I (30–34.9), obesity class II (35–39.9), and obesity class III (⩾40). The premise behind the new guidelines is that increasing categories of BMI are related to increasing health risks associated with obesity and its co-morbidities.

A recent follow-up of over 1 million participants in the Cancer Prevention Study II in the United States recovered a U-shaped relationship between BMI and mortality; however, the risk of death across the specific U.S. NIH and WHO categories of BMI were not estimated [8]. A study by Seidell et al. [9], using the BMI cut-offs recommended by the WHO [10], reported a U-shaped relationship between BMI and mortality in men, but not in women. A more recent report from the Seven Countries Study, based on a sample of men, reported that the BMI categories of underweight and obesity were both related to increased mortality [11]. Another recent analysis from the Alameda County Study examined the relationship between the new BMI categories and mortality [4]. An increased risk of death in the underweight and obese class II and III (combined) categories was evident, but there was no increased risk for overweight or class I obesity. In fact, the risk of mortality in the pre-obese group was significantly less than the normal weight group (RR = 0.82, 95% CI 0.72–0.92). Thus, there is a need to address the issue of mortality related to the recommended clinical guidelines for the classification of overweight and obesity in large population-based samples that include both men and women.

The purpose of this study was to examine the relationship between BMI and mortality in the Canadian population. The all-cause mortality associated with the U.S. NIH and WHO categories of underweight, normal weight, overweight and obesity were examined using data from the 13-year mortality follow-up of the 1981 Canada Fitness Survey (CFS), a representative sample of the Canadian population.

Section snippets

Sample

The Canada Fitness Survey was based on a nationally representative sample of the Canadian population [12]. The total population excluded in the sample was 3%; which included First Nation people living on reserves, students living in school dormitories, armed forces personnel living on bases and residents of the Territories and remote areas. The sample was a survey of households from urban and rural areas of every province, and included all ethnic groups (with the exception of First Nation

Results

The distribution of participants across categories of BMI at baseline are presented in Table 1, along with the estimated hazard ratios (HR) for all-cause mortality. In the total sample, approximately 3% are underweight, 58% are normal weight, and 39% are overweight or obese (BMI ⩾ 25 kg/m2). There are a total of 61,800 person-years of follow-up in males and 70,715 person-years of follow-up in females. The pattern of hazard ratios is similar in males, females, and the combined sample. There is

Discussion

The new recommendations of the U.S. NIH and WHO for the classification of overweight and obesity based on the BMI are based on the premise that the cut-offs used to define overweight and obesity are related to health status 3, 7. In other words, increasing levels of overweight or obesity should be related to increasing morbidity and mortality risks. In the Cancer Prevention Study II, the nadir of the curve for BMI and mortality fell within the recommended normal weight category (BMI 18.5–24.9

Acknowledgements

This study was partially supported by the Heart and Stroke Foundation of Ontario (Grant NA#4400). Claude Bouchard is funded, in part, by the George A. Bray Chair in Nutrition. Funding for the record linkage at Statistics Canada was provided through Health Canada.

References (32)

  • D.B Allison et al.

    Simulation study of the effects of excluding early deaths on risk factor-mortality analyses in the presence of confounding due to occult diseasethe example of body mass index

    Ann Epidemiol

    (1999)
  • D.B Allison et al.

    The direct health care costs of obesity in the United States

    Am J Public Health

    (1999)
  • L.V Sjostrom

    Mortality of severely obese subjects

    Am J Clin Nutr

    (1992)
  • Clinical guidelines for the identification, evaluation, and treatment of overweight and obesity in adults. Bethesda,...
  • W.J Strawbridge et al.

    New NHLBI clinical guidelines for obesity and overweightwill they promote health?

    Am J Public Health

    (2000)
  • R.P Troiano et al.

    The relationship between body weight and mortalitya quantitative analysis of combined information from existing studies

    Int J Obes

    (1996)
  • J.C Seidell et al.

    Overweight and obesity in the mortality rate datacurrent evidence and research issues

    Med Sci Sports Exerc

    (1999)
  • Obesity: preventing and managing the global epidemic. Report of a WHO consultation on obesity, Geneva, 3–5 June, 1997....
  • E.E Calle et al.

    Body-mass index and mortality in a prospective cohort of U.S. adults

    N Engl J Med

    (1999)
  • J.C Seidell et al.

    Overweight, underweight, and mortalitya prospective study of 48 287 men and women

    Arch Intern Med

    (1996)
  • Physical status: The use and interpretation of anthropometry: report of a WHO expert committee. Geneva: World Health...
  • T.L Visscher et al.

    Underweight and overweight in relation to mortality among men aged 40–59 and 50–69 yearsthe Seven Countries Study

    Am J Epidemiol

    (2000)
  • Fitness Canada

    Fitness and lifestyle in Canada

    (1983)
  • Fitness Canada

    Standardized test of fitnessoperations manual

    (1981)
  • H.S Shannon et al.

    Comparison of individual follow-up and computerized linkage using the Canadian Mortality Data Base System

    Can J Public Health

    (1989)
  • A.R Schnatter et al.

    An analysis of death ascertainment and follow-up through Statistics Canada's Mortality Database System

    Can J Public Health

    (1990)
  • Cited by (127)

    • Combined alcohol use and weight status effects on mortality risk among adults: Analysis of the National Health Interview Survey linked mortality files, 2001–2015

      2022, Drug and Alcohol Dependence
      Citation Excerpt :

      Underweight BMI status is associated with poor health outcomes such as cognitive impairment, brain damage, and higher risk of all-cause mortality (Ijaz et al., 2018; Lee et al., 2016; Nedeltcheva and Scheer, 2014; Oliveira et al., 2010). We found support for the increased risk of all-cause and cancer mortality among underweight individuals (Katzmarzyk et al., 2001). Even though light/moderate drinkers have lower BMI, excessive or problematic alcohol intake can be associated with sleep duration and disruption, affecting metabolic health and resulting in poor nutrient intake and malnourishment (Rossi et al., 2015).

    • Impact of adolescent alcohol use across the lifespan: Long-lasting tolerance to high-dose alcohol coupled with potentiated spatial memory impairments to moderate-dose alcohol

      2017, Alcohol
      Citation Excerpt :

      It is not yet clear whether the reduced body weight is due to reduced food intake, reductions in general health, or impaired nutritional absorption. However, reduced body weight in adolescents is often associated with specific health concerns such as asthma, scoliosis, delayed puberty onset, and increased risk for osteoporosis (He et al., 2001; Katzmarzyk, Craig, & Bouchard, 2001; Lake, Power, & Cole, 1997). Because of the significant difference in body weight produced by CIE exposure during adolescence, we opted to analyze body weight during the lifespan taking into account (i.e., covarying) the disparity.

    View all citing articles on Scopus
    View full text