A systematic review of the evidence for Canada's Physical Activity Guidelines for Adults

This systematic review examines critically the scientific basis for Canada's Physical Activity Guide for Healthy Active Living for adults. Particular reference is given to the dose-response relationship between physical activity and premature all-cause mortality and seven chronic diseases (cardiovascular disease, stroke, hypertension, colon cancer, breast cancer, type 2 diabetes (diabetes mellitus) and osteoporosis). The strength of the relationship between physical activity and specific health outcomes is evaluated critically. Literature was obtained through searching electronic databases (e.g., MEDLINE, EMBASE), cross-referencing, and through the authors' knowledge of the area. For inclusion in our systematic review articles must have at least 3 levels of physical activity and the concomitant risk for each chronic disease. The quality of included studies was appraised using a modified Downs and Black tool. Through this search we identified a total of 254 articles that met the eligibility criteria related to premature all-cause mortality (N = 70), cardiovascular disease (N = 49), stroke (N = 25), hypertension (N = 12), colon cancer (N = 33), breast cancer (N = 43), type 2 diabetes (N = 20), and osteoporosis (N = 2). Overall, the current literature supports clearly the dose-response relationship between physical activity and the seven chronic conditions identified. Moreover, higher levels of physical activity reduce the risk for premature all-cause mortality. The current Canadian guidelines appear to be appropriate to reduce the risk for the seven chronic conditions identified above and all-cause mortality.

Canada has played a leading role in the development of physical activity guidelines for individuals across the lifespan. This includes the development (in 1998) of "Canada's Physical Activity Guide to Healthy Active Living" for adults between the ages of 20 and 55 yr [21], which was followed by "Canada's Physical Activity Guide to Healthy Active Living for Older Adults" [22], and "Canada's Physical Activity Guide to Healthy Active Living for Children and Youth" [23]. The adult guidelines (which are now approximately 10 years old) state generally that 20-55 yr adults should accumulate 60 min of daily physical activity or 30 min of moderate to vigorous exercise on at least 4 days a week [18,19].
We reported recently that Canada's adult guidelines were consistent with other international guidelines and were supported by a compelling body of literature [18,19]. We revealed strong evidence that routine physical activity was effective in the primary prevention of cardiovascular disease, stroke, hypertension, breast cancer, colon cancer, type 2 diabetes and osteoporosis. Moreover, physical activity appears to play an important role in the prevention of obesity and obesity-related co-morbidities. However, implicit in the adult guidelines is the belief that there is a dose-response relationship between physical activity and the associated health benefits. Moreover, a central belief in these guidelines and most international physical activity guidelines is that the doseresponse relationship is curvilinear with the greatest health benefits seen in physically inactive individuals who become "more physically active." In fact, a consistent pattern (shown in Figure 1) has been hypothesized, wherein there are marked changes in health status with relatively minor increments in physical activity/fitness in individuals that are the least active/fit. Generally, the health benefits have been thought to level off at the upper end of the physical activity/fitness continuum ( Figure 1). However, recent work (such as that provided by Gledhill and Jamnik in the Canadian Physical Activity and Lifestyle Approach) has speculated that there are likely multiple dose-response curves for various endpoints [24].
The primary purpose of this systematic review was to examine critically the current literature to determine whether or not a dose-response relationship exists between habitual physical activity and chronic disease. In particular, we sought to determine whether the key messaging "Every little bit counts, but more is even better -everyone can do it!" of the adult physical activity guidelines is supported by a strong body of evidence.
Due to the breadth of literature, we have chosen to focus on the relationship between physical activity and all-cause mortality, and the seven chronic conditions that are thought to be reduced greatly with habitual physical activity (i.e., cardiovascular disease (excluding stroke), stroke, hypertension, colon cancer, breast cancer, type 2 diabetes and osteoporosis) (see Table 1). Owing to the nature of the physical activity guidelines, the emphasis of this paper was on primary prevention, despite the clear evidence that routine physical activity is also an effective secondary preventative strategy against many chronic conditions [16,18,19]. Accordingly, our primary objectives were to examine the evidence for a dose-response relationship between: 1) physical activity and all-cause mortality, and 2) physical activity and incidence of the following chronic conditions (cardiovascular disease (except stroke), stroke, hypertension, type 2 diabetes, colon cancer, breast cancer, and osteoporosis.

Criteria for considering studies for this review
Our research team utilized a rigorous, systematic, and evidence-based approach to examine critically the levels of evidence on physical activity and the risk for premature mortality and chronic disease. Any studies that evaluated the relationship between at least three different levels of physical activity and mortality or incidence of chronic disease were eligible for inclusion. Therefore, excluded studies included those that examined only the most active versus least active populations (e.g., sedentary/inactive vs. physically active). Any form of physical activity/exercise measurement (e.g., self-report, pedometer, accelerometer, maximal aerobic power (VO 2 max)) was eligible for inclusion. The key outcomes were mortality and incidence of chronic disease. Only published, English language studies examining adults (e.g., 19-65 yr) were included. Participants must have previously been healthy (asymptomatic) adults without established chronic disease. There was no restriction according to study design.
To examine the relative risk reductions associated with physical activity, we calculated the mean and median risk reductions across studies focusing on the highest level versus the lowest level of physical activity/ fitness. For each study we also determined whether or not a dose-response relationship was present (i.e., reflecting a progressive decrease in the risk with increasing physical activity/fitness levels).

Search strategy
Literature searches were conducted in the following electronic bibliographical databases:  The Medical Subject Headings (MeSH) were kept broad. See tables 2, 3, 4, 5, 6, 7, 8 and 9 for the complete search strategy and keywords used. The electronic search strategies were created and carried out by researchers experienced with systematic reviews of the literature (DW and LN). The citations and applicable electronic versions of the article (where available) were downloaded to an online research management system (RefWorks, Bethesda, Maryland, USA).

Screening
Two reviewers (LN and SC) screened independently the title and abstract of the citations to identify potential articles for inclusion. Duplicate citations were removed. The reviewers were not blinded to the authors or journals. Biographies of key studies and reviews in the field were also cross-referenced for further articles. For those articles that appeared relevant, the full text was obtained and data was extracted using a common template. In cases of disagreement, discussion with a third reviewer (DW) was used to achieve consensus. Full (100%) consensus was achieved. All studies that were excluded during the citation and full-article screening processes were recorded along with the reasons for exclusion.

Data Extraction
Two reviewers (LN and SC) completed standardized data extraction forms, which were verified by two other reviewers (DW and SB). We extracted information regarding the study design, the country where the study was conducted, the participant characteristics, the sample size, the objectives of the study, the methodologies employed, the major outcomes (i.e., mortality, incidence of chronic disease, physical activity levels/classifications), and the comments and conclusions made based on the findings of the study. The reviewers were not blinded to the journal or the author names when extracting information from the articles.

Level of Evidence
The approach used to establish the level and grade of evidence was consistent with that used during creation of the "Canadian clinical practice guidelines on the management and prevention of obesity in adults and children" [25]. The level of evidence provides information regarding the strength of the evidence in favour of physical activity/exercise in the primary prevention of premature mortality and the seven chronic diseases of primary interest. This evaluation process is based on a pre-defined and objective criteria (see Table 10).
The grade for each article provides information regarding whether physical activity is effective in the primary prevention of the varied conditions evaluated (Table 10). Where applicable this grade informs the reader about the potential risk of the physical activity. A study that receives the highest grading would indicate  that the benefits clearly outweigh the risks and receive a strong recommendation.

Quality Assessment
The quality of each study was also established using the procedures of Gorber et al. [26]. Owing to the fact that only observational study designs were included in our systematic review, we used the Downs and Black [27] scale to assess the quality of non-randomized investigations. Similar to the work of Prince et al. [28] we chose to include the most relevant components of the scoring tool. Therefore, a modified version of the Downs and Black checklist was used with the final checklist consisting of 15 items with a maximum score of 15 points. Higher points reflected a superior quality of investigation.

Physical Inactivity and All-Cause Mortality
A total of 2040 citations were identified during the electronic database search ( Figure 2). Of these citations, 288 were identified in MEDLINE, 222 in EMBASE, 496 in Cochrane, and 1034 in the CINAHL/SportDiscus/Psy-chInfo search. A total of 167 duplicates were found, leaving a total of 1873 unique citations. A total of 1696 articles were excluded after scanning, leaving a total of 177 articles for full review. From these articles 130 were excluded after full review leaving 47 articles for inclusion in the systematic review. An additional 23 articles were added to the review based on the authors' knowledge of the area. The reasons for exclusion included review articles (n = 26), commentary (n = 10), did not report 3 levels of physical activity (n = 24), no objective measure of physical activity (n = 2), report (n = 15), not a formal study (n = 11), not related to all-cause mortality (n = 27), the participants were too young (n = 1), not able to retrieve articles (n = 7), and other (n = 7). Therefore, a total of 70 articles were included in the  systematic review of the literature regarding the relationship between physical activity and premature mortality.
The majority of the studies included in our systematic review were prospective cohort investigations (Table  11). These studies involved a total of 1,525,377 participants; averaging 21,791 participants per study (range 925). There were a total of 111,125 reported cases of premature all-cause mortality (ranging per study from 952). The total length of study followup for the prospective cohort studies averaged 11.1 yr (ranging from 0.5-28 yr). The articles were published over a 22 yr period ranging from 1985 to 2007. These studies involved large samples of men and women from regions throughout the world.
We observed a mean 31% lower risk for all-cause mortality in the most active individuals. The median risk reduction was 32%. It is important to highlight that many of these studies included women, with sub-analyses that revealed similar risk reductions between sexes. Our findings are consistent with previous reports [15,16,[29][30][31]. The majority (90%) of the studies supported the health benefits of physical activity demonstrating a significant risk reduction in physically active individuals. The level of evidence would be considered to be a Level 2A based on the presence of overwhelming evidence from observational trials. The studies examined were generally of a good quality with a mean (and median) score of 12 out of 15 (range [10][11][12][13][14].  A clear dose-response relationship was also observed with marked reductions in the risk for all-cause mortality occurring with relatively small increments in physical activity ( Figure 3). To examine more closely the temporal relationship between physical activity and all-cause mortality we calculated the (unadjusted) relative risks associated with incremental levels of physical activity/fitness using the reported cases of all-cause mortality and the number of participants (per group) in each investigation. In some instances, we were required to calculate the number of participants based on the reported incidence rates and person years, or based on data obtained directly from the authors (2 investigations). We were not able to obtain this information in 18 investigations, and as such this analysis was restricted to the remaining 52 investigations. There was considerable variability in the methods of classifying the physical activity/fitness levels of the participants. Accordingly, Figure 3 illustrates the mean relative risk reduction according to three separate study types including those that subdivided participants into tertiles, quartiles and quintiles, respectively. This figure demonstrates clearly the doseresponse relationship between physical activity and allcause mortality. Collectively, the literature is consistent indicating that the current Canadian guidelines (approximately 4.2 MJ/wk, 1000 kcal/wk) are associated with a 20-30% lower risk for premature all-cause mortality, with greater health benefits with high volumes and/or intensities of activity. In our analyses it was apparent that the greatest differences in risk occurred between the lowest adjacent activity/fitness categories, suggesting that sedentary individuals can markedly reduce their risk for all-cause mortality with relatively minor increments in physical activity. This is consistent with the current messaging of Canada's physical activity guidelines.
The strength of the relationship between physical fitness and premature mortality has been well-established [6,32,33]. In our analyses there were greater risk reductions in studies that took objective measures of physical fitness. We observed an average risk reduction of approximately 45%, which was consistent between men and women. A risk reduction of greater than 50% was not uncommon in these studies. For instance, Myers et al. (2004) reported that being fit or physically active was associated with greater than 50% lower mortality risk in men. They also noted that a 4.2 MJ/wk (1000 kcal/wk) increase in physical activity, or a 1 metabolic equivalent (MET) higher physical fitness level was associated with a mortality benefit of around 20%. It is also important to highlight that longitudinal studies evaluating changes in physical activity or fitness have revealed a lower premature mortality risk [16,[34][35][36][37][38][39][40][41]. As we previously reported, routine physical activity or elevated physical fitness also appears to reduce the risk for premature mortality in individuals with risk factors for chronic disease [42,43].

Implications
Since the seminal work of Morris and colleagues (in the 1950s [44,45]) and the early work of Paffenbarger (in the 1970s [46,47]) there has been considerable research (especially epidemiological evidence) documenting the health benefits of engaging in routine physical activity and/or being physically fit [17,48]. Both physical activity (a behaviour) and physical fitness (an attained state) appear to be related to health status in a dose-dependent fashion, with physical fitness demonstrating the strongest relationship [18,19]. Numerous reports indicate that physical inactivity and/or low physical fitness are associated with an increased risk for chronic disease and premature all-cause and disease-specific mortality [2,43,[49][50][51]. Some of the most compelling research includes the relationship between physical activity/ fitness and all-cause mortality. As demonstrated below and in Table 11 and Figure 1, this literature is extensive. The assessment of the relationship between all-cause mortality is complicated by the inclusion of deaths related to suicides, homicide, and accidents [18,19,52]. Nonetheless, the available evidence is incontrovertible; individuals who are habitually physically active and/or physically fit are at a markedly reduced risk for premature all-cause mortality [15,16,18,19]. In Canada, physical inactivity is a major cause of premature mortality from diseases of the cardiovascular system (33.3%), cancers (29.1%), and type 2 diabetes (3.5%) [53]. Globally, physical inactivity has been linked with 2 million premature deaths per year, including 22% of cases of coronary heart disease, and 10-16% of cases of breast cancer, colon cancer, rectal cancer and type 2 diabetes [54]. As such, the promotion of the health benefits of physical activity is of paramount importance for the effective prevention of chronic disease and premature mortality on a national and international scale.
In summary, there is a clear dose-response relationship between physical activity and premature all-cause mortality. Physically active individuals have an approximate risk reduction of 31% in comparison to physically inactive individuals. When objective measures of aerobic fitness are taken the risk reductions are even greater approximating 45%.
Recommendation #1 For a reduced risk for premature mortality, it is recommended that individuals should participate in 30 min or more of moderate to vigorous exercise on most days of the week. Greater health benefits appear to occur with higher volumes and/or intensities of activity. [Level 2, Grade A]

Primary Prevention of Cardiovascular Disease
In our systematic search of the literature, a total of 9408 citations were identified during the electronic database search (Figure 4). Of these citations, 5973 were identified in MEDLINE, 2561 in EMBASE, 193 in Cochrane, and 681 in the CINAHL/SportDiscus/PsychInfo search. A total of 923 duplicates were found, leaving a total of 8485 unique citations. A total of 8138 articles were excluded after scanning, leaving a total of 347 articles for full review. An additional 20 articles were added through cross-referencing. From these articles 319 were excluded after full review leaving 33 articles for inclusion in the systematic review. The reasons for exclusion included non-experimental studies (n = 45), only effect on cardiovascular disease risk factors (n = 115), did not report 3 levels of physical activity (n = 12), subjects less than 18 yr of age (n = 4), reviews, summaries, dissertations, thesis, and abstracts (n = 30), clinical population (n = 14), not on cardiovascular disease or did not fit

Comments and Conclusions
Blair et al 1989 [7] To study physical fitness (PF) and risk of all-cause mortality in men and women.
• n = 13,344 ( Myers et al 2004 [32] To determine the effects of PF and physical activity (PA) on allcause mortality.
• n = 6,213 Baseline and mean 5.5 ± 2.0 year follow-Up Being fit or active is associated with >50% reductions in mortality risk.
Prospective cohort   Blair et al 1995 [36] To evaluate the relationship between changes in PF and risk of mortality in men.
• n = 9,777 4.9 year mean follow-up • 223 deaths Men who maintained or increased adequate PF had a reduced risk for all-cause mortality than individuals who were consistently unfit. Groups based on changes in PF G1 = unfit to unfit G2 = unfit to fit G3 = fit to unfit G4 = fit to fit Bijnen et al 1999 [37] To examine the association of PA at baseline and 5 years                       • 690 deaths The data support the hypothesis that an active and fit way of life delays death.        [194] To examine the relationship between maximal oxygen uptake and overall mortality.  The study provides some support for recommendations that emphasize MPA. A benefit of VPA is also evident. Lee et al 1995 [196] To examine the independent association of vigorous and nonvigorous PA with longevity.
• n = 17,321   [197] To investigate the effect of various PA patterns on all-cause mortality.
• n = 8,421 Baseline 1988 and follow-up 1993 • 1,234 deaths The results suggest that regular PA generating 1000 kcal/wk or more should be recommended for lowering mortality rates. Among those with no major risk factors, even 1-2 episodes per week generating 1000 kcal or more can postpone mortality.   Trend p = <0.001 Cox proportional HR Leon et al 1997 [199] To examine the long-term association of LTPA and risk of death from coronary heart disease and all-causes.
• n = 12,138 16 year follow-up The data suggest that a relatively small amount of daily moderate intensity LTPA can reduce premature mortality in middleaged and older men at high risk for CHD.     [202] To determine the effects of exercise and non-exercise PA on mortality.
• n = 67,143 Baseline and an average of 5.7 year follow-up  Trend p = 0.008 Cox proportional hazard models Menotti and Seccareccia 1985 [203] To investigate the relationship between OPA and all-cause mortality.
• n = 99,029 Baseline and 5 year follow-up The results suggest that PA may play a role in the prediction of fatal events.
• Sex: Men        The study demonstrates the protective effect of LTPA on mortality.
• Sex: Men   The study shows the strong predictive effect of PA on mortality is independent of education level.    Table 11: Studies examining the relationship between physical activity and all-cause mortality.

(Continued)
Sundquist et al 2004 [216] To study the association between varying levels of PA and all-cause mortality in the elderly.     Weller and Corey 1998 [220] To study the relationship between PA and mortality in women.
• n = 6,620 Baseline and 7 year follow-up • 449 deaths PA is inversely associated with risk of death in women.   • n = 1,975 Baseline and 10 year follow-up The study found a strong inverse association between heavy LTPA and all-cause mortality. definition of cardiovascular disease (n = 78), and other (n = 19). Therefore, a total of 49 articles were included in the systematic review of the literature regarding the relationship between physical activity and the incidence of cardiovascular disease. The majority of the studies included in our systematic review were prospective cohort investigations (Table  12). These studies involved a total of 726,474 participants; averaging 12,313 participants per study (range 680-88,393). There were a total of 34,815 reported cases of cardiovascular disease (ranging per study from 42-2,596). The total length of study follow-up for the prospective cohort studies averaged 14.1 yr (ranging from 2-29 yr). The articles were published over a 32 yr period ranging from 1975 to 2007. These studies involved large samples of men and women from regions throughout the world.

UK
Similar to the all-cause mortality data, the risk for cardiovascular disease demonstrates a graded inverse doseresponse relationship to physical activity and fitness. The relative reduction in the incidence of cardiovascular disease averages 33% (median risk reduction of 36%), with greater risk reductions in studies that employed objective measures of aerobic fitness. It is not uncommon for studies to demonstrate a 50% or higher risk reduction when an objective measure of physical fitness was taken (Table 12). The importance of physical activity may actually be underestimated owing to multivariate control for many confounding factors (as discussed previously) and the fact that effects of within-person variation in physical activity are often not considered [55]. The relative risk reduction appears to be similar for men and women, and also appear to extend to non-Caucasian populations [56]. Some evidence also exists indicating that small amounts of physical activity are associated with lower cardiovascular-disease related mortality [57,58]. Similar to all-cause mortality, physical activity confers health benefits independent of other known risk factors [42,59]. Collectively, the level of evidence would be considered to be Level 2A based on the presence of overwhelming evidence from observational trials. The quality of the investigations was generally high with a mean (and median) Downs and Black score of 12 (range 9-14).

Implications
Research in the field began with the landmark work of Morris and colleagues, which demonstrated that men in physically demanding occupations (bus conductors and postmen) had a significantly lower risk of heart disease than individuals who worked in less demanding jobs (bus drivers and office workers) [45]. Since then considerable research has examined the relationship between physical activity and the risk for cardiovascular disease. In fact, several systematic reviews of the literature have been developed regarding the role of habitual physical activity in the primary and secondary prevention of cardiovascular disease [33,[60][61][62]. The research to date has been consistent and compelling, habitual physical activity reduces markedly the risk for cardiovascular disease.
Based on the available literature, there is compelling evidence that the recommendation of 30 min of moderate intensity exercise on most days of the week (equivalent to 4.2 MJ/wk or 1000 kcal/wk) reaches a threshold associated with significant reductions in cardiovascularrelated mortality [32,63]. Brisk walking has also been shown to be preferable to a slower pace [64]. However, weekly exercise volumes of less than 4.2 MJ (1000 kcal) may be cardio-protective [14,59,[65][66][67]. For instance,  found that as little as 1 hr/wk of walking was associated with a 50% lower cardiovascular disease mortality in one sample of women. Wisloff et al. [58] reported that a single weekly bout of self-reported high intensity exercise was associated with a lower risk of cardiovascular death relative to those reporting no activity in both men (RR = 0.61, 95% CI = 0.49-0.75), and women (RR = 0.49, 95% CI = 0.27-0.89). Moreover, no additional benefit was seen with higher durations or frequency of exercise sessions [58]. The authors stated that this evidence challenges "current recommendations that require at least 1000 kcal of caloric expenditure per week to achieve exercise-induced protection against premature cardiovascular death." However, this research is in fact supportive of the Canadian guidelines which recognize the potential health benefits of low volumes of physical activity as reflected by the statement "Every little bit counts, but more is even better -everyone can do it!" It however should be noted that the statement "more is even better" is supported by a strong evidence base.
Recommendation #2 For a reduced risk for cardiovascular diseaserelated events and mortality, it is recommended that individuals participate in 30 min or more of moderate to vigorous exercise on most days of the week. Greater health benefits appear to occur with high volume and/or intensities of activity. Health benefits may also occur with as little as one hr of brisk walking per week. [Level 2, Grade A]

The Primary Prevention of Stroke
Stroke affects a significant proportion of Canadian society with approximately 50,000 new cases each year [68]. The relationship between physical activity and the risk for stroke is compelling, supporting at least a 25-      • G1 = 1.00 (referent) • Women's Health Study     (Continued) Bijnen et al 1998 [166] To describe the association between the PA pattern of elderly men and CHD mortality.
• n = 802 Length of Follow-up: 10 • Number of Cases: 90 PA did not show a protective effect on death from CHD.  [174] To examine the association between two measures of physical activity (LTPA and usual walking pace) with cause specific mortality (CHD).

yrs
• Number of Cases: 955 Inverse associations of both LTPA and walking pace with mortality from CHD were seen.
• Sex: Men     [199] To study the relationship of PA to CHD in a well defined population at above average risk for CHD over a 16 yr observation period.  There was an inverse and significant dose-response association between LTPA and CHD-related mortality.
Adjusted RR (95% CI) Whole group Weller et al 1998 [220] To examine the relationship between PA and mortality.

MI Logistic regression analysis
Yu et al 2003 [221] To examine the optimal intensity of LTPA to decrease the risk of CHD mortality in middle aged British men.
• n = 1,975 10 year follow-up          Cox proportional HR Physical demand of work and CHD incidence: Kaprio et al 2000 [232] To examine the contribution of genetic and other familial factors to the relationship between LTPA and CHD.

(Continued)
Manson et al 1999 [242] To assess the comparative roles of walking and vigorous exercise in the prevention of coronary events in women.
• n = 72,488 PA assessment: • Number of Cases: 645 coronary events Both walking and VPA are associated with a substantial reductions in incidence of CHD. Risk reductions for each were similar hen total PAy was similar. Walking 3 or more hours per week could reduce the risk of CHD by 30-40%.   Observed a strong and dose dependent inverse association between LTPA and non fatal CHD.       30% risk reduction in the most active individuals [31]. In fact, in a review of the literature Katzmarzyk and Janssen [20] reported that lack of physical activity carried a relative risk of 1. In our systematic review of the literature, a total of 1104 citations were identified during the electronic database search ( Figure 5). Of these citations, 405 were identified in MEDLINE, 183 in EMBASE, 227 in Cochrane, and 289 in the CINAHL/SportDiscus/PsychInfo search.
A total of 13 duplicates were found, leaving a total of 1091 unique citations. A total of 1011 articles were excluded after scanning, leaving a total of 80 articles for full review. An additional 9 articles were retrieved through cross-referencing and the authors' knowledge of the field. From these articles 64 were excluded after full review leaving 25 articles for inclusion in the systematic review. The reasons for exclusion included non-experimental/weak design (poor execution introducing bias) (n = 16), did not contain three levels of physical activity or not possible to determine dose-response relationship (n = 14), reviews, summaries, meta-analyses (n = 17), dissertations, thesis, abstracts (n = 8), and other (n = 9). Therefore, a total of 25 articles were included in the systematic review of the literature regarding the relationship between physical activity and the primary prevention of stroke (Table 13).
The data providing dose-response information is all observational in nature, involving both case control and cohort investigations. These studies (predominantly prospective cohort designs) included a total of 479,336 participants; averaging 17,753 subjects per study (range 428-73,265). There were a total of 12,361 reported cases of stroke (ranging per study from 32-2,863). The total length of study follow-up for the prospective cohort studies averaged 13.2 yr (ranging from 6-26 yr). The articles were published over a 14 yr period ranging from 1993 to 2007. These studies involved large samples of men and women from regions throughout the world including studies from the USA (11), UK (2), Iceland (1), Denmark (2), Norway (4), Netherlands (1), Finland (2), Japan (1), Australia (1) and Greece (1). Very few studies [69,70] examined non-Caucasian participants.
We found strong evidence that physical activity was associated with a reduced risk for stroke. The level of evidence was consistent with a Level 3A classification. We observed an average risk reduction of 31% across all studies (median = 29%). In comparison to cardiovascular disease, there was more variability in the risk reductions in stroke in the highest activity/fitness group. The quality of the investigations was also generally quite good with a mean (and median) Downs and Black score of 13 (range [11][12][13][14][15]. The risk reductions appear to be even greater in studies that assessed physical fitness directly. For instance, in data from the Aerobics Center Longitudinal Study [71] the high fitness group (estimated peak METs = 13.1) and the moderate fitness group (estimated peak METs 10.5) had significantly lower risks of stroke mortality (68 and 63%, respectively) than the least fit men (estimated peak METs 8.5).
A dose-response relationship did emerge when examining the literature. However, as illustrated by others this was extremely variable amongst studies and varied

Men
Both high and low-intensity exercise may be associated with a reduced risk of stroke in both men and women. Norway

(Continued)
Lee et al 1999 [74] To examine the association between exercise and stroke risk.  Although RR for of death from stroke was below 1 for both moderate and high compared with low PA, this association did not reach the level of statistical significance.
• G1 = 1.00 (referent)              Simonsick et al 1993 [268] To examine the association between recreational PA among physically capable older adults and incidence of selected chronic diseases and mortality over 3 and 6 years.

After 3 years Iowa
No consistent relationship between PA and stroke was found after 3 or 6 years across all 3 population cohorts.   according to the type of stroke (ischemic or haemorrhagic) [52]. For instance, 12 studies (46%) revealed a doseresponse relationship in one or more measures of occupational and/or leisure-time physical activity and the risk for stroke. It is difficult to determine the minimal and optimal physical activity dosage for the prevention of stroke. Brisk walking has been associated with a lower risk of total and ischemic stroke [72]. In the Harvard Alumni study, the risk of stroke was lower at a weekly energy expenditure of 4. In summary, the results of these studies (taken as a whole) indicate that occupation-and leisure time-related physical activity are inversely related to the risk for stroke. Both physically active men and women have a lower risk of stroke, and it appears that this benefit may be present for both ischemic and haemorrhagic stroke [74]. The relationship between physical activity and stroke appears to be consistent between men and women. Unfortunately, relatively limited data exists in non-Caucasian populations.
Recommendation #3 For a reduced risk of stroke, it is recommended that individuals should participate in 30 min or more of moderate to vigorous exercise on most days of the week. Brisk walking appears to be protective against the development of stroke. It remains to be determined whether lower volumes of physical activity lead to a reduced risk for stroke. [Level 3, Grade A]

Primary Prevention of Hypertension
A total of 6287 citations were identified during the electronic database search ( Figure 6). Of these citations, 4054 were identified in MEDLINE, 1360 in EMBASE, 253 in Cochrane, and 620 in the CINAHL/SportDiscus/ PsychInfo search. A total of 40 duplicates were found, leaving a total of 6247 unique citations. A total of 6167 articles were excluded after scanning, leaving a total of 80 articles for full review. An additional five articles were found through cross-referencing and the reviewers' personal files. From these articles 72 were excluded after full review for the following reasons: weak design (n = 4), did not contain three levels of physical activity or not possible to determine dose-response relationship (n = 19), reviews, summaries, meta-analyses (n = 8), not dealing with hypertension (n = 2), only reported on changes in blood pressure (n = 27), clinical population (n = 7), and other (n = 6). Therefore, a total of 12 articles were included in the systematic review of the literature regarding the relationship between physical activity and the primary prevention of hypertension. The majority of the literature examining the dose-response (for at least three levels of physical activity/fitness) involved prospective cohort analyses (83%).
As shown in Table 14, 12 investigations examined the dose-response (i.e., three or more levels) relationship between physical activity and the incidence of hypertension. This involved a total of 112,636 participants, averaging 10,240 subjects per study (range 1,243-41,837).   Increased EE during LTPA and increased intensity of these activities were associated with reduced risk for incident hypertension (age adjusted) in men but not women.     There were a total of 11,441 reported cases of hypertension (ranging per study from 118-2,936). The total length of study follow-up averaged 8.6 yr (ranging from 0-16 yr). The articles were published over a 24 yr period ranging from 1983 to 2007. All studies reviewed demonstrated positive effects of physical activity on the risk for hypertension. Of these studies all (7; 58%) revealed an inverse and graded relationship between hypertension and at least one measure of physical activity or fitness. Across all studies, when comparing the most active/fit group versus the least active/fit group we found an average RR of 0.68 (median = 0.70, range 0.37 to 0.90). Therefore, we observed that physical activity/fitness was associated with an average risk reduction of 32% for hypertension. It should be noted that the study [75] demonstrating the largest risk reduction (63%) evaluated cardiorespiratory fitness directly during a maximal treadmill test. This supports research (as discussed previously) which indicates that physical fitness is a better predictor of chronic disease than physical activity [6,18,19,32,33]. Taken as a whole, the level of evidence can be classified as Level 3A. The quality of studies was generally good with a mean Downs and Black score of 11 (median = 11, range = 10-12).
Five studies showed variable results (i.e., no clearly defined dose-response) while generally supporting the inverse relationship between physical activity/fitness and hypertension [76][77][78][79][80]. The variability in the response appears to be the result of different activity/fitness classifications and/or differing subject populations. For instance, some studies revealed that the dose-response relationships differed between genders and/or ethnicities [76,77]. Pereira et al. [76] revealed a 30% reduction in the risk for hypertension in the most active white men. There were graded dose-response relationships between indices of both leisure and sport activities in the white men.
However, there was a lack of association between physical activity and hypertension in white women and African American men and women. Similarly, Haapenen et al. [77] revealed a stronger association in men than in women. However, it should be noted clearly that other studies included in this systematic review evaluated women demonstrating a graded response [81]. Moreover, several studies were conducted with non-Caucasian populations and demonstrated a dose-dependent benefit [82][83][84][85]. In fact, data was obtained from varied regions of the world including USA (7), Japan (2), China (1), and Finland (1). Therefore, there is evidence to suggest that the protective effects of physical activity with respect to hypertension are transferable to women and non-Caucasian populations. However, further research is clearly warranted that examines the relationship between physical activity and hypertension in persons of different ethnicities. Moreover, further research is needed to determine the effects of impact of socio-economic status on the observed relationships.
Some studies have indicated that vigorous activity is required to reduce the risk for hypertension. For instance, Paffenbarger [78] revealed that Harvard Alumni who did not engage in vigorous sports play were at a 35% higher risk for developing hypertension. However, there was no difference in the risk for hypertension in men who climbed >50 stairs per day, walked more than 5 city blocks daily, or engaged in light sports only. Similarly, the Paffenbarger and Lee [79] study revealed that moderately vigorous sports play was associated with a lower risk for hypertension, but physical activity (kcal/wk), walking distance (km/wk) and the amount of stairs climbed (floors/wk) were not significant predictors of the risk for hypertension. Collectively, this research group concluded that these findings highlighted the importance of the intensity of effort.
However, it should be noted that many of the studies in our systematic review observed the protective effect with moderate intensity physical activities. Findings from randomized controlled trials have also provided strong evidence that moderate intensity aerobic exercise is sufficient to reduce blood pressure and the risk for hypertension, particularly in at risk individuals [86,87]. The American College of Sports Medicine [88] recently advocated that to prevent hypertension, individuals should exercise on most, and preferably all, days of the week at a moderate intensity, for 30 min or more per day (continuous or accumulated). They also recommended supplementing endurance type activities with resistance exercise. This is supported by research indicating that moderate intensity resistance training can reduce blood pressure [89]. Collectively, this research and our current summary of the dose-response literature indicates that physical activity levels that are of a moderate to vigorous intensity are sufficient to lead to marked reductions in the risk for hypertension.

Implications
The impact of hypertension on North American society is enormous. In the US, 31% of non-institutionalized adults over the ages of 20 are currently thought to have hypertension [90]. In Canada, approximately 20% of adults report a diagnosis of hypertension including over 4 million Canadians [91][92][93]. It has been estimated that a 55 yr old Canadian with normal blood pressure has a greater than 90% chance of developing hypertension before the age of 80 yr [92]. The primary prevention of hypertension is of paramount importance to the attenuation of the risks and costs associated with hypertension and related comorbidities.
There is clear evidence that routine physical activity and/or increased physical fitness reduce greatly the risk for hypertension in both normotensive and hypertensive individuals [18,19]. Extensive research has been conducted in the area including numerous prospective trials and various randomized controlled trials. Numerous reviews of the literature (of epidemiological and randomized controlled trials) have supported an inverse relationship between physical activity/fitness and in the incidence of hypertension [20,87,89,[94][95][96][97][98][99][100][101][102]. In a recent systematic review of the prospective literature, Katzmarzyk and Janssen (2004) calculated that physically inactive individuals were at a 30% higher risk for hypertension (RR = 1.30 (95% CI = 1.16-1.46)) with a population attributable risk of 13.8% in Canada [20]. Acute bouts of exercise have also been shown to lead to transient changes in blood pressure that are potentially of health benefit [98]. For instance, blood pressure is often reduced after a single exercise session for 12-22 hr [88,103].
It is clear that routine physical activity is effective in both the primary and secondary prevention of hypertension. However, the optimal dosage of physical activity/ exercise remains somewhat unclear. Our review of the literature examined critically the relationship between multiple levels of physical activity/fitness and the incidence of hypertension (in individuals without diagnosed hypertension). As identified above this evidence was compelling supporting the protective effects of habitual physical activity in the primary prevention of hypertension.
Recommendation #4 For a reduced risk for hypertension, it is recommended that individuals should participate in 30 min or more of moderate to vigorous exercise on most days of the week. [Level 3, Grade A]

Primary Prevention of Colon and Breast Cancer Colon Cancer
In our systematic search of the colon cancer literature, a total of 252 citations were identified during the electronic database search (Figure 7). Of these citations, 83 were identified in MEDLINE, 44 in EMBASE, 25 in Cochrane, and 100 in the CINAHL/SportDiscus/Psy-chInfo search. A total of 15 duplicates were found, leaving a total of 237 unique citations. A total of 164 articles were excluded after screening, leaving a total of 73 articles for full review. From these articles 47 were excluded after full-text review leaving 26 articles for inclusion, and an additional 7 articles were added from the authors' personal files. The reasons for exclusion included non-experimental/weak design (n = 8), reviews, summaries, meta-analyses (n = 13), editorial/comment (n = 3), not dealing specifically with colon cancer (n = 4), did not contain three levels of physical activity or not possible to determine dose-response relationship (n = 9), and other (n = 10). Therefore, a total of 33 articles were included in the systematic review of the literature regarding the relationship between physical activity and the primary prevention of colon cancer.
These studies involved a total of 1,433,103 participants; averaging 43,427 participants per study (range 142-413,044). There were a total of 17,959 reported cases of colon cancer (ranging per study from 93-1,993). The total length of study follow-up for the prospective cohort studies averaged 10.7 yr (ranging from 4-26 yr). The articles were published over a 23 yr period ranging from 1985 to 2008. These studies involved large samples of men and women from regions throughout the world.
A dose-dependency of this relationship was present in the majority of the studies. When comparing the most  active/fit group versus the least active/fit group we found a mean risk reduction of 30% (median = 32%) across all studies. The most compelling literature was that which evaluated the relationship between moderate-to-vigorous leisure time physical activity. Based on the literature reviewed and the volume of activity assessed it would appear that Canada's guidelines for physical activity are sufficient to lower the risk for the development of colon cancer in asymptomatic adults. The level of evidence would be considered to be Level 2A. The studies were generally of a higher quality with a mean Downs and Black score of 13 (median = 14, range = 11-15).
It should be noted that there was considerable variability in the findings and conclusions of the studies (Table 15). As discussed later, the literature was further confounded by the fact that the relative risks associated with physical activity were often controlled (through multivariate analyses) for various potential confounding factors, which may actually inappropriately decrease the level of risk reduction associated with physical activity [31]. Moreover, similar to other chronic conditions this literature was limited greatly by the lack of consistent physical activity assessment and description. In many instances, it was difficult to determine the actual absolute volume and/or intensity of activity for each category of comparison. However, despite these limitations the results of these studies (taken as a whole) indicate that both occupation-and leisure time-related physical activity are inversely related to the risk of colon cancer.

Breast Cancer
As reviewed eloquently by others, the epidemiological evidence relating physical activity to a decreased incidence of breast cancer is persuasive. A recent systematic review of the literature found that more than 60 observational trials have examined the relationship between physical activity and breast cancer [31]. Previous reviews of the literature have revealed compelling and consistent findings indicating that habitual physical activity is associated with a reduced risk for breast cancer ranging from 20-80% [31,104].
Various investigations have attempted to evaluate the dose-response relationship between physical activity and the incidence of breast cancer (Table 16). Despite the volume of evidence available questions still remain regarding the minimal and optimal volume of exercise required to reduce the risk for breast cancer. As discussed by others [31,104] the findings are as varied as the investigations.
In our systematic search of the literature, a total of 571 citations were identified during the electronic database search (Figure 8). Of these citations, 228 were identified in MEDLINE, 89 in EMBASE, 56 in Cochrane, and 198 in the CINAHL/SportDiscus/PsychInfo search.
A total of 46 duplicates were found, leaving a total of 571 unique citations. A total of 411 articles were excluded after scanning, leaving a total of 114 articles for full review. From these articles 77 were excluded after full review leaving 37 articles for inclusion in the systematic review. An additional 6 articles were found through the reviewers' personal files. The reasons for exclusion included not containing three levels of physical activity or not possible to determine dose-response relationship (n = 1), reviews, summaries, meta-analyses (n = 20), report (n = 5), editorial/comment (n = 21), not a research article (N = 11), not dealing specifically with breast cancer (n = 4), not relevant (n = 5), not primary prevention (n = 3), and other (n = 10). Therefore, a total of 43 articles were included in the systematic review of the literature regarding the relationship between physical activity and the primary prevention of breast cancer.
The data providing dose-response information is all observational in nature, involving both case control and cohort investigations. These studies involved a total of 1,861,707 participants averaging 44,326 subjects per study (range 526-680,000). There were a total of 80,247 reported cases of breast cancer (ranging per study from 986). The total length of study follow-up for the prospective cohort studies averaged 10.5 yr (ranging from 4-31 yr). The articles were published over a 14 yr period ranging from 1993-2007. These studies involved large samples of men and women from regions throughout the world.
The literature with respect to the primary prevention of breast cancer is as compelling as that found with respect to colon cancer. There is strong evidence that routine physical activity is associated with a reduced risk for the development of breast cancer. However, this literature is also confounded by many shortcomings (similar to other cancer literature) including considerable variability in the statistical analyses employed, the physical activity measurement tools used, and the experimental designs.
The overall risk reduction for breast cancer for individuals that are habitually physically active (at or above Canada's guidelines for physical activity) is thought to approximate 20-40% [31,105]. In our analyses, we found very similar findings. When comparing the most active group versus the least active group we found a mean (and median) risk reduction of 20% across all studies. The level of evidence would be considered to be Level 2A. Generally, the articles were of high quality with a mean Downs and Black score of 13 (median = 13, range = 9-14).
A dose-dependency of this relationship is also generally present in the majority of the studies. For instance, greater than 50% studies revealed a dose-response                                                               Table 15: Studies examining the relationship between physical activity and colon cancer.               The study suggests that energy intake, BMI and physical inactivity are each independently and positively associated with BC risk.        • Ethnicity: Hispanic and non-Hispanic White PA assessment: Self-reported non-OPA (MET hr/wk score) • G1 = 1.00 (referent) • Characteristics: Case Group            The results indicate a weak negative association between PA and risk of BC in middle-aged women.  Table 16: Studies examining the relationship between physical activity and breast cancer.

(Continued)
Adjusted OR (95% CI) by category of total EE in adulthood, aged        Rintala et al 2002 [337] To obtain an estimate of BC incidence in association with selfrated OPA.

17,986 cases of BC
The results support the hypothesis that OPA, if high enough, markedly reduced BC risk.          Similarly, Sesso et al. (1998) revealed that there was an 8% reduction in the risk for breast cancer with a relatively small energy expenditure of 500-999 kcal/wk. Further risk reductions were observed with higher energy expenditures (= 1000 kcal/wk = 51% reduction in the risk). As discussed above, Monninkhof et al. revealed a 6% decrease in breast cancer risk for each additional hour of physical activity per week [104]. Taken as a whole, it would therefore appear that Canada's guidelines for physical activity are more than appropriate for reducing the risk for breast cancer. Further research however is required to determine the minimal volume of exercise that is effective in the primary prevention of breast cancer.

Implications
There is a preponderance of data linking physical inactivity to site-specific cancers, particularly of the breast and colon [31,[104][105][106][107][108][109]. The protective effects of physical activity also appear with other forms of cancer (such as endometrial cancer) [110]. In an important review of the literature Lee revealed that physically active women have a 20-30% lower risk of breast cancer, and physically active men and women have a 30-40% lower risk of colon cancer [105]. A more recent systematic review of the literature revealed a 20-80% lower risk of breast cancer in post-menopausal women [104], with a weaker association in pre-menopausal women. Considering data from both pre-and post-menopausal women the authors demonstrated that physically active individuals had a 15-20% lower risk of breast cancer. Monninkhof et al. also reported a 6% lower risk of breast cancer for each additional hour of physical activity per week [104]. This level of risk reduction was also supported by the U.S. Department of Health and Human Services during its recent evaluation of the literature [31]. Our current reviews of the literature support previous work in the field including the finding of a doseresponse relationship between physical activity and cancers of the breast and colon [104,105,109]. It would appear that 30-60 min/day of moderate-to-vigorous physical activity is associated with a lower risk of breast and colon cancer.
Recommendation #5 For a reduced risk for site specific cancers (such as colon cancer and breast cancer), it is recommended that individuals should participate in 30 min or more of moderate to vigorous exercise on most days of the week. [Level 2, Grade A]

Primary Prevention of Type 2 Diabetes
In comparison to other chronic conditions, there is relatively limited literature examining the relationship between multiple levels of physical activity/fitness and the incidence of type 2 diabetes. All of the literature  examining the dose-response (for at least three levels of physical activity/fitness) involved prospective cohort analyses. A total of 3655 citations were identified during the electronic database search (Figure 9). Of these citations, 2038 were identified in MEDLINE, 1116 in EMBASE, 118 in Cochrane, and 372 in the CINAHL/ SportDiscus/PsychInfo search. A total of 614 duplicates were found, leaving a total of 3041 unique citations. A total of 2865 articles were excluded after scanning, leaving a total of 176 articles for full review. From these articles 156 were excluded after full review leaving 20 articles for inclusion in the systematic review of the literature regarding the relationship between physical activity and type 2 diabetes. The reasons for exclusion included non-experimental/weak design (N = 18), three levels of physical activity not reported (N = 16), reviews, summaries, or meta-analyses (N = 41), not related to type 2 diabetes (N = 71), and other (N = 10).
As shown in Table 17, 20 investigations examined the dose-response (i.e., three or more levels) relationship between physical activity and the incidence of type 2 diabetes. This involved a total of 624,952 subjects, averaging 32,892 subjects per study (range 1,543-87,907). There were a total of 19,325 cases of type 2 diabetes (ranging per study from 78-4,030). The total length of follow-up averaged 9.3 yr (ranging from 3 -16.8 yr). The articles were published over a 16 yr period ranging from 1991 to 2007.
Of these studies 100% revealed an inverse relationship between type 2 diabetes and levels of physical activity or fitness. When comparing the most active/fit group versus the least active/fit group we found an average risk reduction of 42% (median = 44%). Therefore in our analyses the most physically active/fit had a 42% lower risk of developing type 2 diabetes. The majority (84%) of these studies revealed incremental reductions in the risk for type 2 diabetes with increasing activity/fitness levels. Therefore, the health benefits with respect to type 2 diabetes prevention appear to continue across the physical activity/fitness continuum. Similar to other clinical conditions, the dose-response relationship is such that small changes in activity levels yield marked reductions in the risk for type 2 diabetes. The health benefits of exercise appear to be particularly prevalent in individuals at high risk for developing type 2 diabetes (e.g., those with a high body mass index, the metabolic syndrome, a history of hypertension and/or a family history of type 2 diabetes). The level of evidence relating physical activity to the primary prevention of type 2 diabetes would be considered to be Level 2A. The quality of the investigations was generally high with a mean (and median) Downs and Black score of 13 (range [11][12][13][14]. As with other conditions is it difficult to separate the effects of volume and intensity of exercise. However, small changes in activity levels clearly can have a large effect on the risk for and incidence of type 2 diabetes. For instance, Hu and coworkers [111] revealed that nurses (n = 68,497) who engaged in 1 hr/day of brisk walking had 24% less obesity and 34% less type 2 diabetes (over a 6-year follow-up). These authors estimated that approximately 30% of new cases of obesity and 43% of new cases of type 2 diabetes could be prevented by adopting an active lifestyle including less than 10 hr/wk of television watching and ≥ 30 min/d of brisk walking. Similarly, over a 5-year period, male physicians who             D & B score = 12 • Characteristics: Healthy asymptomatic Never (score 0) Sometimes (score 1) Often (score 2) Non-Diabetes (n = 2017) PA score: 2.34 ± 1.7 'Some' PA: 84% Pearson product moment correlation coefficient and Cox proportional HR Hu et al. 1999 [347] To quantify the doseresponse relationship between total PA and incidence of T2D in women.   Although BMI and physical inactivity are independent predictors of incident diabetes, the magnitude of the association with BMI was greater than with PA in combined analyses. These findings underscore the critical importance of adiposity as a determinant of T2D.   exercised vigorously at least once weekly had a 29% lower incidence of type 2 diabetes than individuals who did not exercise regularly [112]. These authors also revealed that physical activity that was sufficient to cause sweating was associated with a lower incidence of type 2 diabetes. Other research has also demonstrated that moderate-to-vigorous physical activity (≥ 5.5 METs for at least 40 min per week) and/or aerobic fitness levels above 31 mL·kg -1 ·min -1 are associated with a lower risk of type 2 diabetes in middle-aged men [113] with the effect being the greatest in high-risk individuals. Therefore, it would appear that Canada's recommendations for physical activity are sufficient to reduce the risk for type 2 diabetes. In 2001, Hu et al. [114] reported very interesting and compelling research regarding the role of lifestyle factors in the development of type 2 diabetes. Using data from the Nurses' Health Study, they defined a low-risk group according to five lifestyle factors including BMI, a healthy diet, the participation in moderate-to-vigorous physical activity for at least 30 min per day, no current smoking, and the consumption of an average of at least one-half serving of an alcoholic beverage per day. They revealed that the women in the low risk group had a RR for type 2 diabetes of only 0.09 (CI 0.05-0.17) in comparison to the rest of the cohort. They also found that 91% of the cases of type 2 diabetes in this cohort (CI 83-95%) could be attributed to the five lifestyle factors. This research provided compelling evidence that the majority of type 2 diabetes could be prevented through healthy living [115].
As reviewed in Table 17 there is evidence that leisuretime, occupational and commuting-related leisure time activities significantly reduce the risk for the development of type 2 diabetes. For instance, a recent study by Sato and colleagues [116] revealed that the walking distance to work was directly related to the incidence of type 2 diabetes in 8,576 Japanese men followed for 4 years. The risk reduction was approximately 27% in the participants who walked to work for ≥21 min compared to those who did so for ≥10 min. These findings are similar to that found by Hu et al. who reported that moderate occupational, commuting and leisure-time physical activities all had a significant inverse relationship to risk in middle-aged men and women [117].
Although ethnicity is often not reported in the current research, the studies examined in our systematic review came from a variety of countries and regions. Data was obtained from studies from the USA, Canada, UK, Japan, and Finland. For instance, Hsia et al. (2005) conducted a prospective 5-year study of 87,907 post-menopausal women, finding a strong graded inverse relationship between physical activity and type 2 diabetes. The relationship was stronger in "Caucasian" than in minority (African-American, Hispanic or Asian) women. The authors postulated this finding might reflect less precise risk assessments in minority women [118]. As we have outlined previously, further research is clearly warranted that examines the relationship between physical activity and type 2 diabetes in persons of different ethnicities. Moreover, further research is needed to determine the effects of socio-economic status on the observed relationships. Nonetheless, the research is compelling, habitual physical activity appears to be highly effective in the primary prevention of type 2 diabetes.

Implications
In 1992, the consensus panel from the International Conference on Physical Activity, Fitness and Health (held in Toronto, Canada) [17] indicated that physical activity can effectively reduce the risk for, and incidence of, type 2 diabetes. Over 15 years later, the research is compelling; habitual physical activity is an effective primary preventative strategy against the development of type 2 diabetes [111][112][113][118][119][120][121][122][123]. As shown in our analyses, numerous observational studies have revealed that regular physical activity is associated with a lower risk of developing type 2 diabetes [111][112][113][118][119][120][121][122][123]. Moreover, increased aerobic fitness is inversely associated with the risk of type 2 diabetes [113,124]. It is also apparent that both aerobic and resistance type activities reduce the risk for type 2 diabetes [125,126].
Although it is difficult to determine the dose-response between physical activity and type 2 diabetes in the majority of the current randomized controlled trials, these trials have revealed important findings. Influential exercise and/or lifestyle intervention trials have demonstrated clearly the health benefits of physical activity/ exercise in the prevention of type 2 diabetes. For instance, in the Diabetes Prevention Program (US), 3,234 high-risk participants were randomly assigned to one of three groups: a) a placebo control, b) metformin drug therapy (850 mg twice daily), and c) a lifestyle intervention. The authors revealed that the lifestyle intervention (including physical activity for at least 150 minutes per week) was more effective than metformin (alone) (respective reductions in incidence 58% and 31%) [127]. Similarly, Tuomilehto et al. (2001) conducted a randomized controlled trial with middle-aged, overweight subjects with impaired glucose tolerance (172 males and 350 females). The authors reported a significant reduction in the incidence of type 2 diabetes in the intervention group (which received advice regarding moderate intensity exercise (30 min/day) and dietary control). The lifestyle intervention reduced the risk of type 2 diabetes by approximately 54% in women and 63% in men [128]. In a review of the literature, Williamson et al. revealed modest weight loss via diet and physical activity reduced the incidence of type 2 diabetes in high risk individuals by 40-60% over a 3-4 year period [129]. Collectively, the epidemiological and randomized controlled trials provide compelling evidence supporting the role of habitual physical activity in the primary prevention of type 2 diabetes.
Recommendation #6 For a reduced risk for type 2 diabetes, it is recommended that individuals should participate in 30 min or more of moderate to vigorous exercise on most days of the week. [Level 2, Grade A]

Primary Prevention of Osteoporosis
The protective effects of physical activity and exercise training on bone health are well documented. In fact, the relationship between indicators of bone health (such as bone mineral density or bone mineral content) and physical activity have been evaluated extensively (see Table 18). Numerous exercise intervention trials have revealed that aerobic and resistance activities have a beneficial effect on bone mineral density across the lifespan [16]. In fact, several systematic reviews of the literature [130][131][132][133][134][135] and major consensus statements [136] have shown clearly the potential benefits of both aerobic and resistance training on bone health (particularly in post-menopausal women). It has been estimated that exercise interventions prevent or reverse at least 1% of bone loss per year in the lumbar spine and the femoral neck of pre-and post-menopausal women [130,137].
Exercise has also been shown to significantly reduce the risk and/or number of falls in comparison to inactive controls [138][139][140][141][142]. Moreover, fracture risk and/or incidence has been shown to be reduced in active individuals [143][144][145]. Case-control studies of older persons who suffered a hip fracture have revealed that these individuals had significantly lower physical activity levels throughout adulthood [136,146]. Observational studies have also revealed an inverse relationship between the incidence of fractures and physical activity [147][148][149]. For instance, Joakimsen et al. revealed lower fracture rates in individuals who performed more weight-bearing activities [148]. Similarly, Kujala et al. [147] in a 21-year prospective study revealed that intense activity was associated with a lower incidence of hip fracture (Hazard Ratio = 0.38, 95% CI = 0.16-0.91). Feskanich et al. (2002) also revealed that moderate physical activity was inversely related to the risk of hip fracture in postmenopausal women [149]. In a review of observational trials, Katzmarzyk and Janssen [20] revealed that the fracture risk was markedly higher in habitually inactive individuals (RR = 1.59 (95% CI = 1.40-1.80)) with a population attributable risk of 24% in Canada.
There is clear evidence that exercise training is of benefit for bone health and accordingly reduces the risk for osteoporosis. However, remarkably limited research has actually examined the relationship between routine physical activity and the prevalence and/or incidence of osteoporosis ( Figure 10). In our systematic search of the osteoporosis literature, a total of 3655 citations were identified during the electronic database search ( Figure  7). Of these citations, 1888 were identified in MEDLINE, 236 in EMBASE, 82 in Cochrane, and 481 in the CINAHL/SportDiscus/PsychInfo search. A total of 276 duplicates were found, leaving a total of 2411 unique citations. A total of 2059 articles were excluded after screening, leaving a total of 352 articles for full review. From these articles all 352 were excluded after full-text review. The reasons for exclusion included non-experimental/weak design (n = 87), did not contain three levels of physical activity or not possible to determine dose-response relationship (n = 38), reviews, summaries, meta-analyses (n = 39), not dealing specifically with osteoporosis (n = 21), only on change in bone mineral density (N = 123), clinical population (N = 10), bone metabolism (N = 13), fractures (N = 3), population < 18 yrs (N = 11), and other (N = 7). An additional 2 articles were found through the authors' knowledge of the field.
As identified in our systematic search, the majority of the literature has dealt with the relationship between physical activity and indicators of bone health and/or the incidence of fractures. However, a recent observational trial [150] has provided evidence supporting the ability of physical activity to reduce the incidence of osteoporosis. For instance, Robitaille et al. revealed a dose-response relationship between physical activity level and the prevalence of reported osteoporosis in 8073 women aged ≥ 20 yr in the National Health and Nutrition Examination Survey, 1999-2004 [150]. Those performing no physical activity were at a higher risk than those who engaged in moderate (<30 MET hr/wk) and high (>30 MET hr/wk) levels of physical activity. There was a dose-response relationship with the highest physical activity group having the lowest prevalence of osteoporosis. Similarly, Keramat et al. [151] in a casecontrol investigation revealed that physical activity was protective against the development of osteoporosis.
At this time it is difficult to define clearly the actual dose-response required to cause a reduction in the incidence of osteoporosis. It is clear that bone adaptations to exercise are load dependent and site specific [9,10,16,152]. As such, physical activities that involve significantly loading/impact are often advocated for the prevention of osteoporosis. It is has been shown that running 15-20 miles per week is associated with bone mineral accrual or maintenance. Longer distances however may be associated with reduced bone mineral density [136].
Feskanich et al. reported that the risk of hip fracture was lowered by 6% for each increase of 3 MET-hours  • G1 = 7.8 (6.9-8.7) • G2 = 6.7 (3.8-11.8) • G3 = 9.5 (7.6-11.9)  per week of activity (or 1 hr/wk of walking at an average pace) [149]. There was a linear reduction with increasing physical activity level. Walking for at least 4 hr/wk was also associated with a 41% lower risk of hip fracture compared to less than 1 hr/wk [149]. The work of Robitaille et al. also indicated that moderate levels of physical activity are sufficient to reduce the prevalence of osteoporosis [150]. In summary, there is preliminary evidence to indicate that the current Canadian physical activity guidelines are sufficient to maintain and/improve bone health. However, further research is clearly required, in particular research that examines the relationship between physical activity and the incidence of osteoporosis in both men and women from varied ethnic backgrounds. Currently, the level of evidence would be considered to be at a Level 3A. The quality of the investigations was generally low with a mean (and median) Downs and Black score of 11.
Recommendation #7 For a reduced risk for osteoporosis, it is recommended that individuals should participate in load bearing activities for 30 min or more on most days of the week. [Level 3, Grade A]

Other Considerations Musculoskeletal Fitness and Health
In the present analyses, all indices of physical activity/ fitness were incorporated into our systematic reviews. Although the majority of the data is related to aerobic activities, it should be noted that many of these activities also had a significant musculoskeletal component. Moreover, direct measurements of musculoskeletal fitness were included in various studies included in our review. Although there is limited information available (in comparison to aerobic activities) there is compelling evidence that musculoskeletal fitness is also positively associated with health status [9,10,16].
Warburton and colleagues [9,10] in two reviews of the literature reported that enhanced musculoskeletal fitness is associated positively with glucose homeostasis, bone health, functional independence, mobility, psychological well-being, and overall quality of life and negatively associated with fall risk, morbidity and premature mortality. They also reported that interventions that increase musculoskeletal fitness also have a significant positive effect on the health status of the individuals with a low musculoskeletal reserve (e.g., the frail elderly).
In an evaluation of the current literature some key findings emerge. Grip strength has particularly been shown to be inversely related to premature mortality and/or morbidity (e.g., functional limitations) [153][154][155][156]. Rantanen et al. (1998) reported that those individuals with the lowest grip strength had a higher rate of mortality at a younger age (over a 27-year period) than their counterparts with higher muscular strength. Furthermore, they revealed that those with a faster rate of decline in muscular strength (>1.5% per year) or a very low grip strength (<21 kg) had a greater incidence of chronic diseases, such as type 2 diabetes, stroke, arthritis, coronary heart disease, and pulmonary disorders. It was shown that those in the lowest grip strength tertile had an 8-fold increased risk for disability. Individuals with high muscular strength have also been shown to develop less functional limitations in comparison to their counterparts with lower strength over a 5-year period [157]. Katzmarzyk and colleagues [126,154,158] in Canada have also demonstrated a positive relationship between musculoskeletal fitness and health status. For instance, Katzmarzyk and Craig (2002) revealed that there was a significantly higher risk of premature mortality in the lower quartile of sit-ups in both men (RR = 2.72) and women (RR = 2.26). Grip strength was also predictive of mortality in men (RR = 1.49), but not women. In a recent study, Mason et al. revealed that musculoskeletal fitness was a significant predictor of weight gain over a 20-year period [158]. Importantly, they also reported that individuals with low musculoskeletal fitness had 78% greater odds of significant weight gain (≥ 10 kg) compared to those with high musculoskeletal fitness. These studies provide direct support for the inclusion of resistance and flexibility training in Canada's physical activity guidelines for adults [3,159].
Recommendation #8 For improved health status and reduced risk for chronic disease and disability, it is recommended that individuals should include daily activities that tax the musculoskeletal system [Level 2, Grade A]

Limitations
It is important to note that for each chronic condition, the methods used to determine the relationship between physical activity and the specific clinical outcome were often quite varied. For instance, early work in the field generally controlled for few confounding variables (such as age). In comparison, current literature often controls for a myriad of potential confounding variables. These discrepancies make the comparison of the relative risk reductions between studies and across clinical conditions more difficult. Moreover, the multivariate analyses (controlling for various potential confounding factors) may inappropriately decrease the level of risk reduction associated with physical activity and the clinical endpoint [31]. This is owing to the fact that some of the health benefits associated with physical activity may be mediated through these variables [31].
There was often considerable variability in the findings and major conclusions of the studies examined. Often the available literature was limited by the lack of a clear standard for assessing physical activity. In many instances, it was extremely difficult to determine the actual dosage of physical activity used to group the participants. This lack of clarity makes it very difficult to clearly define the dose-response relationship between physical activity and various chronic conditions.

Conclusions
There is incontrovertible evidence that regular exercise is an effective preventative strategy against premature mortality, cardiovascular disease, stroke, hypertension, colon cancer, breast cancer, and type 2 diabetes. There is also compelling indirect evidence to support the protective effects of physical activity with respect to osteoporosis. In many instances the dose-response relationship is linear with further health benefits with increasing levels of activity. The current Canadian physical activity guidelines for adults are sufficient to reduce the risk for multiple chronic diseases simultaneously. The acknowledgement that every bit of exercise counts towards health benefits (with greater benefits at higher energy expenditures) is consistent with the literature and a reasonable message to promote to the general population. However, further investigation is likely required to evaluate the relationship between physical activity and health status in non-Caucasian populations.