Health Reports
Trends in physical fitness among Canadian adults, 2007 to 2017

by Caroline Y. Doyon, Rachel C. Colley, Janine Clarke, Ian Janssen, Brian W. Timmons, Grant R. Tomkinson, Mark S.Tremblay and Justin J. Lang

DOI: https://www.doi.org/10.25318/82-003-x202101100001-eng

Abstract

Background

The fitness levels of Canadian adults declined substantially between 1981 and the years 2007 to 2009, suggesting a reduction in population health. This paper updates the fitness trends of Canadians aged 20 to 69 years by extending the time period to 2017

Data and methods

The Canadian Health Measures Survey is a repeated cross-sectional survey that is conducted to produce nationally representative health estimates. Descriptive statistics are presented for fitness measures in 2016 and 2017 by age and sex, and trends in fitness were calculated spanning a period of 10 years (2007 to 2017). The associations between fitness measures and meeting the 2020 Canadian physical activity recommendations were also assessed.

Results

From 2007 to 2017, there were few statistically significant changes in the fitness levels of Canadian adults. When all ages were combined, there were declining trends in predicted cardiorespiratory fitness, from 39.5 to 36.7 mL•kg^-1•min^-1 among men and 34.0 to 32.2 mL•kg^-1•min^-1 among women. Trends indicated declining flexibility among men. In general, meeting the current Canadian moderate-to-vigorous physical activity recommendation was associated with better fitness, particularly in the categories of predicted cardiorespiratory fitness and body composition.

Interpretation

The periodic assessment of fitness in Canadians provides valuable insight into population health. The present update provides evidence that fitness levels among adults have generally stabilized over the past 10 years. Taken with the reported declines in fitness that occurred from 1981 to the 2007-to-2009 period, this study shows that the fitness of Canadian adults remained low between 2007 to 2009 and 2016 to 2017. It is necessary to explore new ways to help improve the fitness levels of the Canadian population.

Keywords

anthropometry, body composition, cardiorespiratory fitness, flexibility, muscular strength, muscular power

Authors

Caroline Y. Doyon and Janine Clarke are with the Centre for Population Health Data, Statistics Canada. Rachel C. Colley is with the Health Analysis Division, Analytical Studies and Modelling Branch, Statistics Canada. Ian Janssen is with the School of Kinesiology and Health Studies and Department of Public Health Sciences, Queen’s University. Brian W. Timmons is with the Child Health and Exercise Medicine Program, Department of Pediatrics, McMaster University. Grant R. Tomkinson is with the Department of Education, Health and Behavior Studies, University of North Dakota, Grand Forks, North Dakota, United States and Alliance for Research in Exercise, Nutrition and Activity (ARENA), School of Health Sciences, University of South Australia, Adelaide, South Australia, Australia. Mark S. Tremblay is with the Healthy Active Living and Obesity Research Group, Children’s Hospital of Eastern Ontario Institute, Ottawa, Ontario. Justin J. Lang (justin.lang@canada.ca ) is with Centre for Surveillance and Applied Research, Health Promotion and Chronic Disease Prevention Branch, Public Health Agency of Canada and the School of Mathematics and Statistics, Department of Health Sciences, Carleton University.

Introduction

Health-related physical fitness refers to an integrated series of characteristics, including cardiorespiratory fitness (CRF), musculoskeletal fitness (strength, power, endurance and flexibility) and adiposity, that collectively describe an individual’s ability to perform physical activity.^{Note 1} A large body of evidence has shown that aspects of physical fitness, especially CRF and adiposity, are associated with cardiometabolic health and are strong predictors of future morbidity and mortality.^{Note 2Note 3Note 4Note 5} Since the early 2000s, research has also identified important associations between musculoskeletal fitness and health outcomes. In addition to being used as a proxy for overall and upper limb strength,^{Note 6Note 7} several studies support the importance of handgrip strength as an indicator of current and future health.^{Note 8Note 9Note 10Note 11} Research also suggests that handgrip strength is a correlate of bone mineral density and a potential predictor of frailty and fall-related fractures among older adults.^{Note 12Note 13}

Given the link between measures of physical fitness and health, temporal trends in these measures provide insight into overall population health. A recent systematic review identified meaningful declines in CRF among 2,525,827 adults from eight high-income and upper-middle-income countries since the 1980s, which have progressively increased in magnitude in more recent years.^{Note 14} A similar systematic review reported negligible to small declines in handgrip strength among 2,592,714 adults from 142 high-income and upper-middle-income countries since 2000.^{Note 15} It is also well established that trends in overweight and obesity among adults have been increasing internationally since the 1980s,^{Note 16} although there is evidence that this trend has begun to stabilize in industrialized countries such as Canada.^{Note 17} Between 1981 and 2007 to 2009, the fitness of Canadian adults also declined substantially, while body mass index (BMI), waist circumference and skinfold measurements increased.^{Note 18Note 19} Updating previous fitness surveillance data could help identify whether more recent trends in the physical fitness of Canadian adults track with reported international trends in physical fitness and adiposity.

The overarching objective of the present study was to provide an update to the Shields et al.^{Note 18} and Craig et al.^{Note 19} nationally representative fitness estimates and trends by extending the time period from the years 2007 to 2009 to the years 2016 to 2017. The specific objectives for this study included describing (1) the most recent (2016 to 2017) levels of physical fitness among Canadians aged 20 to 69 years, (2) the temporal trends in fitness among Canadian adults from the years 2007 to 2009 to the years 2016 to 2017, and (3) the relationship between fitness and meeting the Canadian moderate-to-vigorous physical activity (MVPA) recommendation published in 2020.^{Note 20}

Methods

Data source

Data for this study are from cycles 1 (2007 to 2009), 2 (2009 to 2011) and 5 (2016 and 2017) of the Canadian Health Measures Survey (CHMS). Fitness measures were not obtained in cycles 3 and 4 because of the rotation of content within the survey. The CHMS is a repeated cross-sectional survey conducted by Statistics Canada that produces nationally representative data on the health of Canadians.^{Note 21} The survey is representative of approximately 96% of the Canadian population, excluding residents of Indigenous reserves or Crown lands, residents of institutions, residents of some remote regions, and full-time members of the Canadian Forces.^{Note 22} Combined response rates of 51.7% for cycle 1,^{Note 23} 55.5% for cycle 2^{Note 24} and 48.5% for cycle 5^{Note 25} were observed. Informed written consent was obtained from participants prior to testing. Ethics approval was obtained from the Health Canada and Public Health Agency of Canada Research Ethics Board.^{Note 26}

The CHMS took place in two parts that occurred within eight weeks. First, CHMS respondents participated in a household interview where a detailed health questionnaire was administered. Next, respondents were invited to a mobile examination centre, where fitness and physical measurements were obtained, and an accelerometer was attached to be worn for seven consecutive days to obtain movement behaviour data, including physical activity. The physical fitness assessments were conducted by a specialist certified by the Canadian Society for Exercise Physiology. All respondents were asked about their current health and any use of prescription medications prior to testing. They were also asked to complete the Physical Activity Readiness Questionnaire (PAR-Q) to further assess eligibility for fitness testing. Respondents were asked to adhere to pre-exercise recommendations related to food, alcohol, nicotine, exercise and blood donations. More details on respondent screening for the CHMS can be found elsewhere.^{Note 23Note 24Note 25}

The present study was limited to respondents aged 20 to 69 years who completed at least one fitness measure during cycle 1, 2 or 5 of the CHMS (n = 9,137). For the analysis, respondents were combined into four groups: all ages, 20 to 39 years, 40 to 59 years and 60 to 69 years. Overall, 2,191 adults (24% of the total sample) aged 20 to 69 years did not complete the modified Canadian Aerobic Fitness Test (mCAFT) because of the stringent inclusion criteria. The reasons for exclusion were an affirmative answer to the PAR-Q (n = 933), a contraindication with medication (n = 605), an acute or chronic condition (n = 287), high blood pressure (n =165), a forgotten breathing aid (n = 68), an appointment as a home visit (n = 31), a pregnancy of more than 12 weeks (n = 61), an elevated heart rate (n = 19), or other reasons (n = 22). An additional 290 respondents were unable to complete one full stage of the mCAFT, resulting in an incomplete score. Those who were screened out from the mCAFT were similar across all three cycles of the CHMS (Appendix Table A).

Measures

Fitness measures

Standing height was recorded to the nearest 0.1 cm using a ProScale M235 digital stadiometer (Accurate Technology Inc., Fletcher, United States). Body weight was measured to the nearest 0.1 kg using a Mettler Toledo CW-90/90X terminal scale (Mettler Toledo Canada, Mississauga, Canada). BMI was calculated as weight in kilograms divided by height in metres squared (kg•m-²)

In cycle 1, waist circumference was measured at the midpoint between the iliac crest and the last rib, following the World Health Organization protocol.^{Note 27} In cycles 2 and 5, the waist circumference protocol changed to the U.S. National Institutes of Health protocol, which used a measurement on a horizontal line at the superior border of the iliac crest at the end of a normal exhalation.^{Note 28Note 29} A published equation was used to adjust the cycle 1 data to match the measurement protocol from cycles 2 and 5 (men R2 = 0.99; women R2 = 0.98).^{Note 30} All waist circumference measurements were taken using a Gulick II tape measure with a calibrated tension indicator device (Country Technology, Gays Mills, United States) and recorded to the nearest 0.1 cm.

The mCAFT was used as a submaximal test to predict a respondent’s CRF as $\dot{V}{O}_{2max}$in mL•kg^-1•min^-1. Details on the mCAFT protocol are provided elsewhere.^{Note 19Note 31} In brief, during the mCAFT, respondents were asked to step up and down on two 20.3 cm steps following a set cadence from an audio track that increased with each consecutive three-minute stage. The last stage for women and the last two stages for men were performed using a single 40.6 cm step. The respondent’s starting stage was determined based on age and sex. Heart rate was recorded immediately after each stage, and the test was terminated once a participant’s heart rate reached 85% of their age-predicted maximum heart rate (220 – age). Heart rate was measured using a heart-rate monitor (Polar Electro Canada Inc., Lachine, Quebec, Canada). CRF ($\dot{V}{O}_{2max}$  in mL•kg^-1•min^-1) was predicted using estimated oxygen cost during the final stage, body weight in kilograms and age in years.^{Note 32}

Flexibility was assessed using the sit-and-reach test.^{Note 28} Respondents were first asked to follow a standardized warm-up. Respondents were then asked to sit on a mat with their legs fully extended and the soles of their feet flat against the flexometer (Fit Systems Inc., Calgary, Canada). They stretched as far forward as possible in a steady motion without bending their knees while stretching their arms toward their toes. The results of two trials were recorded to the nearest 0.1 cm, with the better of two kept for analyses.

Muscular strength was assessed as isometric handgrip strength using a Smedley III analog hand dynamometer (Takei Scientific Instruments, Tokyo, Japan).^{Note 28} Both hands were measured twice, while alternating attempts between each hand. Results were recorded to the nearest kilogram. The best result for each hand was selected and combined to obtain the total handgrip strength score in kilograms. Relative handgrip strength was calculated by dividing the total handgrip strength score by body weight (kg/kg).

Jumping performance (i.e., muscular power) was tested for the first time in cycle 5, using a Leonardo Mechanography® Ground Reaction Force Plate (Novotec Medical GmbH, Pforzheim, Germany). Details on the jumping mechanography procedures and an overview of the predicted variables are available elsewhere.^{Note 33} In brief, respondents completed a single two-leg jump test, which consisted of a single countermovement jump with arm swing performed in a fluid movement with both feet leaving and landing on the plate simultaneously to achieve maximal jumping height. The respondents repeated the test until three valid trials were completed, with a maximum of five attempts. The trial with the best performance was used in the analysis. The Leonardo Mechanography Ground Reaction Force Plate Research Edition® software (v.4.2.b06.10f) uses force and time data to predict jumping height (m), absolute peak power (kW) and relative peak power (kW per kg of body weight).^{Note 33} The force plate signal was sampled at a frequency of 400 or 800 Hz. The validity of each trial was initially assessed by the software or manually assessed by a trained kinesiologist. Subsequently, the validity of all trials was confirmed by an external reviewer specialized in jumping mechanography.

Measurement of physical activity

Participants with at least one valid fitness measure and valid accelerometer-measured physical activity (n = 7,136) were included in a subsample analysis to compare the fitness levels of respondents who meet the MVPA recommendation from the new 2020 Canadian 24-hour movement guidelines (150 minutes of MVPA per week, with no bout length requirement) with those who do not meet it.^{Note 20}

Following the visit to the mobile examination centre, eligible respondents were provided with an Actical activity monitor (Phillips Respironics, Oregon, United States). They were instructed to wear the activity monitor on an elasticized belt over their right hip during waking hours for seven consecutive days. Respondents were unable to see their results during the data collection period. The Actical measures and records time-stamped acceleration in all directions, providing an estimate of movement intensity as a count value per minute. A valid day was defined as an accumulation of 10 or more hours of a respondent wearing their activity monitor, and a valid respondent was defined as someone who accumulated a minimum of four valid days.^{Note 34} To determine daily wear time, non-wear time in minutes was subtracted from 1,440 minutes (24 hours). Non-wear time was defined as at least 60 consecutive minutes of 0 counts, with allowance for one to two minutes of counts between 0 and 100. Average daily and weekly minutes of MVPA were derived in two ways: first, for minutes accumulated in bouts of at least 10 minutes, and second, for all minutes of MVPA (i.e., all accelerometer epochs identified as MVPA). Respondents were classified as having met or not having met the 2020 Canadian 24-hour movement guideline physical activity recommendation of 150 minutes per week of non-bouted MVPA.^{Note 20} Further details on data reduction procedures and accelerometer data analysis have been published elsewhere.^{Note 35}

Analytical techniques

Age- and sex-specific descriptive statistics for fitness measures were calculated for the most recent cycle of the CHMS (2016 and 2017). Age was divided into three groups, representing young (20 to 39 years), middle-aged (40 to 59 years) and older (60 to 69 years) adults. Pairwise contrasts were used to assess age and sex differences across cycles (i.e., temporal trends) and to examine differences in fitness for those meeting and not meeting the Canadian physical activity guidelines for all three cycles, combined.^{Note 20} Significance was set at p < 0.05 when comparing the results of those who met the guidelines and those who did not, and it was set at p < 0.01 for all other analyses to account for multiple comparisons. The standardized effect size anchored to cycle 1 (2007 to 2009) was also plotted to help visualize the trends in fitness. The jumping mechanography data were available only for cycle 5, and thus were not included in the fitness trends analysis. All analyses were weighted using survey weights produced by Statistics Canada to account for non-response bias. Bootstrap weights using the balance repeated replication method were used to calculate the 95% confidence intervals (CIs) to account for the complexity of the CHMS sampling design.^{Note 23Note 24Note 25} Analyses were conducted with SAS version 9.4 (SAS Institute, Cary, United States) and SUDAAN version 11.0.1 (RTI International, Research Triangle Park, United States).

Results

Age and sex differences in physical fitness measures, 2016 and 2017

Descriptive statistics for cycle 5 fitness variables are included in Table 1. All measures of fitness declined with age, except for measures of adiposity, which increased with age. Compared with women, men in all age groups had higher levels of predicted CRF, absolute and relative handgrip strength, jumping height, jumping power, and waist circumference. Women in all age groups had better sit-and-reach flexibility than men. There were no differences between men and women for BMI, except for the group of respondents aged 20 to 39, where men had higher values than women.

Temporal trend in fitness, 2007 to 2017

Fitness did not change substantially among Canadian adults over the 10-year period from 2007 to 2017 (Table 2 and Figure 1). For predicted CRF, there was a significant decline when comparing cycle 5 (2016 and 2017) with cycle 1 (2007 to 2009) for all age groups combined (20- to 69-year-olds) for both men and women. However, when examined separately by age group, the temporal trend of decreasing predicted CRF disappeared. Among men, there were significant declines in sit-and-reach flexibility for all ages combined and for 40- to 59-year-olds when comparing both cycles 2 (2009 to 2011) and 5 (2016 and 2017) with cycle 1 (2007 to 2009). There were no significant temporal changes in the measures of adiposity (BMI and waist circumference) among men and women.

Physical activity recommendations and fitness

Associations between fitness measures and meeting the 2020 Canadian MVPA recommendation are presented in Table 3.^{Note 20} Associations between fitness measures and the 2011 Canadian physical activity recommendation are presented in Supplemental Appendix Tables B-1 and B-2.^{Note 36} All three CHMS cycles were combined because there were no significant differences between them for the variables of interest. In the combined cycles, 42% of respondents met the 2020 recommendation of 150 minutes per week of MVPA with no bout length requirement. For all ages combined, meeting the recommendation was associated with better overall fitness for all measures except absolute handgrip strength and jumping peak power among men and women. For men aged 60 to 69, meeting the recommendation was associated with reduced absolute handgrip strength.

Discussion

The present study provides an update on fitness levels and trends among Canadian adults.^{Note 18Note 19} This study supports previously documented characteristics in adult physical fitness that indicate that measures of predicted CRF, handgrip strength and jumping power are generally higher in men than in women and decline with age, whereas flexibility is greater in women than in men and remains relatively stable with age for women, while it declines somewhat with age for men.^{Note 37} Overall, the fitness levels of Canadian adults have not changed substantially in the past 10 years (2007 to 2017), with only small declines in predicted CRF, and declines in flexibility for men only. These results are concerning given the documented declines in fitness that occurred from 1981 to the 2007-to-2009 period.^{Note 18Note 19} Furthermore, those who met the Canadian physical activity recommendations^{Note 20} generally had better levels of fitness than those who did not meet the recommendations, regardless of age or sex.

Temporal trends in physical fitness

In their studies on the fitness trends among Canadian adults from 1981 to the years 2007 to 2009, Craig et al.^{Note 19} reported a 2.8% to 11.5% decline in predicted CRF, and Shields et al.^{Note 18} reported increases of 7.1% to 15.1% in BMI and 7.1% to 15.3% in waist circumference among men and women aged 20 to 69 years. These reductions in fitness among Canadian adults since 1981 were concerning because of the potential future health risks associated with low fitness levels.^{Note 18} The present study identified few substantial changes in fitness from 2007 to 2017, with observed changes in the adverse direction. Although the lack of further decline may be viewed as equating to a public health victory, it must be highlighted that the fitness levels of Canadian adults have not improved in the last decade and remain low compared with fitness levels from 1981. In comparison with health-related cut-points, the mean estimate for predicted CRF from 2016 and 2017 is considered “at risk” for men aged 40 to 69 and women aged 40 to 59.^{Note 38} In addition, all the mean BMI values from 2016 and 2017 are classified in the overweight category for men and women, and all except for men aged 20 to 39 are considered to have a high waist circumference. Thus, the poor fitness of Canadian adults remains troubling, considering the strong link between low fitness and poor health outcomes among adults.^{Note 5Note 11}

A systematic review by Lamoureux et al.^{Note 14} reported an international decline in CRF among adults of 1.6% per decade from 1967 to 2016, with all eight included countries demonstrating a decline ranging from 0.6% (Republic of Korea) to 5.5% (Singapore) per decade. The review reported more pronounced declines among men and young adults. The current study identified similar declines that were greater in magnitude among men (7.1%) than women (5.3%) when looking at all age groups combined. However, when looking at specific age groups, this study identified general declines that did not reach statistical significance that may be the result of a reduced sample size. Nonetheless, making meaningful comparisons of the magnitude of the changes observed across individual trend studies remains difficult because of the heterogeneity in the type of test used to assess CRF (i.e., run or walk test, treadmill, or cycle ergometer) and how performance is reported (i.e., predicted or directly measured $\dot{V}{O}_{2max}$ in mL•kg^-1•min^-1, running time, or running distance). There is certainly a need to work toward an international standard for CRF testing given the strong link between CRF and health outcomes, and its potential utility for population and clinical health surveillance.^{Note 5}

Worldwide, excess adiposity has been considered a pandemic, with a mean increase in BMI of 0.63 kg•m^-2 (95% CI: 0.53 to 0.73) among men and 0.59 kg•m^-2 (95% CI: 0.49 to 0.70) among women per decade from 1975 to 2014.^{Note 39} This is equivalent to an average increase in body weight of approximately 1.5 kg each decade. Despite these large temporal increases in BMI at the global level, there is evidence that, similar to the present findings, BMI levels have begun to plateau at high levels in many high-income countries.^{Note 17Note 40} Similar to BMI, this study has also identified a plateau at high levels in waist circumference among Canadian men and women. These findings are concerning given the strong association between high excess adiposity and an increased risk of all-cause mortality.^{Note 41}

Physical activity recommendations and fitness

The present study found that meeting the physical activity recommendation was generally associated with lower absolute handgrip strength in older adult men. This finding may reflect that physically active older men are generally lighter than physically inactive older men, and that handgrip strength tends to be higher with greater body mass.^{Note 42} No differences were generally observed in the jumping measures between participants who met or did not meet the physical activity recommendations, with the exception of statistically significant associations between the 2020 recommendation^{Note 20} and relative jumping peak power. It is possible that these reflect the aerobic nature of the physical activity recommendations, which do not typically involve activities that would increase muscular strength and power. It is also possible, at least for the handgrip strength results, that heavier participants with higher absolute handgrip strength generally did not meet the physical activity recommendations; this is consistent when comparing these findings with relative handgrip strength (i.e., generally in the expected direction). An important area of the current physical activity recommendation that is often overlooked or assessed only via self-reporting in public health surveillance is the muscle strengthening portion (i.e., performing muscle strengthening activities using major muscle groups at least twice a week).^{Note 43} Future research should investigate the potential use of handgrip strength and jumping mechanography results to help evaluate the muscle strengthening portion of the physical activity recommendation.^{Note 20} Importantly, jumping mechanography is a new fitness measure in the CHMS that could provide more insight into different health outcomes for Canadians.

Strengths and limitations

An important strength of this study is the use of objective measures of physical fitness and physical activity. The measurement protocol for these measures, with the exception of waist circumference (which was adjusted for), has not changed substantially over the 10-year period, providing a strong basis for estimating temporal trends. It is important to note that the CHMS is a repeated cross-sectional survey that precludes the evaluation of causal relationships. As a result, when assessing the associations between physical activity and physical fitness, it is not possible to determine the direction of the association. Like any national survey, non-response bias may have had an impact on the generalizability of the results. However, the survey weights were calculated to help account for some of the non-response bias. There was also bias associated with the strict screening criteria that were implemented to avoid potential risk associated with the fitness tests. This likely skewed results toward values that would be healthier than those in the general population. For instance, Shields et al.^{Note 18} found that those who were screened out of data collection for the mCAFT during cycle 1 (2007 to 2009) of the CHMS had a substantially higher BMI than those who completed the test. This was also the case for subsequent cycles of the CHMS, as the inclusion criteria had not changed substantially (see Appendix Table C).

This study provides an update on the levels of and temporal trends in the fitness of Canadian adults by examining 10 years of data from 2007 to 2017. Few substantial changes in the fitness trends of Canadians were found. Exceptions included declining trends in predicted CRF when pooling all age groups for both men and women, and a general decline in flexibility among men. Fitness measures were also generally associated with meeting the Canadian physical activity recommendations, especially for predicted CRF, BMI and waist circumference. Continued monitoring of national fitness levels provides valuable insight into the general health of the Canadian population. Given the continued low levels of fitness relative to values observed in 1981, further action is needed to help Canadian adults improve their fitness..

Appendix