Printer Friendly

A short walk a day shortens the hospital stay: physical activity and the demand for hospital services for older adults.

The proportion of people aged 65 and above is growing faster than other age groups, and inactivity is more prevalent among them. (1) Insufficient physical activity negatively affects their health and quality of life due to its association with several chronic diseases. (2,3) Studies show a positive association between inactivity and additional health care utilizations, and health care costs, (4-10) however, these studies do not specifically study older adults. Another research stream focuses on clinical trials of physical fitness and wellness programs and associated health and health care utilization. Studies for various fitness programs have used samples from Texaco employees, (11) Bank of America retirees, (12) employees of insurance companies, (13-16) and employees of the University of Kentucky. (17) However, the samples in these studies are small and mostly specific to the younger or working age groups. Therefore, their results cannot be generalized to older adults.

In order to fill this research gap, this study aims to estimate potential saving from an increase in physical activity among older adults in Canada. Using count data regression models, I estimate the reduction in hospital stays associated with an increase in physical activity. I then examine total reduction in hospital stays that can be achieved from an overall increase in physical activity among older adults.

METHODS

Data source and variables

I used the Canadian Community Health Survey (CCHS), Cycle 2.1, which includes individual level information on health determinants, health status, health care utilization, and health behaviours as well as health conditions, socio-economic and demographic factors. (1) The study sample included 18,196 individuals aged 65 and above.

Physical Activity Variable

The CCHS includes information on individuals' leisure-time physical activities (LTPAs). In the survey, total daily energy expenditure (TEE) from all LTPAs is measured as kilocalories (kcal) per kilogram (kg) of body weight, and calculated using the duration and type of the exercise and the corresponding Metabolic Energy Expenditure (MET). These MET values correspond to low intensity value for each activity since survey participants were not asked to specify intensity of activities. This conservative approach has been adopted in the survey since individuals tend to overestimate the intensity, frequency and duration of activities. (18,19)

It is, however, important to note that this instrument may underestimate older adults' LTPAs since their daily living and activity preferences can be different from the rest of the population, (20) and this instrument may not appropriately reflect more prevalent activities among older adults.

Other Independent Variables

Other than physical activity, there are other factors influencing demand for hospital services. For instance, individuals' health status, social and economic environments, and other health behaviours affect the demand. There are systematic health inequalities among individuals from different socio-economic backgrounds, which cannot be solely explained by the differences in individual health behaviours. People with low income or education suffer more disease and premature mortality and their low socioeconomic background also affects their health behaviour. (21,22) To control for socio-economic differences and variations in health status, economic and social factors, household composition and size, smoke-free environment, work-related factors, mental and general health status, and number of chronic conditions are included in the analysis.

Individuals' behaviour related to LTPAs and health care use can also be affected by other health behaviours. To account for them, alcohol and tobacco consumption and diet are included. Active individuals may utilize more health care services due to sports-related injuries and they may stay inactive during the injuries. (23) To control for these factors, dummy variables for injuries related to exercise are included.

In this study, physical activity is measured using LTPAs. However, individuals can be active during work or other daily activities. To account for these factors, dummy variables for walking or biking to work or while doing errands, and physical effort required at work or daily activities are also included. Tables 1 and 2 present a complete list of all variables.

Count data regressions

Following the literature, I used count data models to estimate the demand for hospital services, measured by the number of hospital stays. (24,25) A number of different count data models are developed to deal with potential problems that exist in count data sets. For instance, non-zero counts are typically observed for a small share of the population that increases skewness in the data, and the variance of the dependent variable can be higher than the mean (overdispersion problem). As opposed to other models, a zero inflated negative binomial (ZINB) model provides a solution to these problems. (26)

The ZINB explicitly models the number of predicted zeros and allows for the variance to differ from the mean. This model implies that conditional mean of the dependent variable,, for an individual i with characteristics X is:

E([y.sub.i]|X) = [[lambda].sub.i](1 - [p.sub.i]) (1)

where [[lambda].sub.i] stands for average utilization among users, and (1 - [p.sub.i]) denotes the probability of being a user. Using the ZINB, one can jointly estimate the probability of being a user with a logit, and average utilization among users with a negative binomial (NB) model.

Specification tests for model selection and sensitivity analysis

To compare various count data approaches, specification tests are performed (see Table 3). First, using an overdispersion test, I test whether the mean and variance of the dependent variable is equal. The test shows that they are not equal, implying that the ZINB or NB are preferable to others. Next, a Vuong test (27) is used to see if the overdispersion is due to excess zeros. The Vuong test suggests that the ZINB is preferable to the NB, implying that the appropriate approach is the ZINB.

In order to check the robustness of the results, the model is also estimated using alternative measures. For instance, income adequacy and ownership of dwellings are used as alternative income and wealth measures. Since the sample consists predominantly of retired seniors, it is also estimated without including active commuting terms. These analyses show that the results presented here are robust.

Partial effect estimations from the ZINB regression

While estimated coefficients show the direction of the association between physical activity and hospital stays, determining the size effect for an improvement in physical activity requires the estimation of partial effects. Partial effects, which are an impact of a one-unit change in TEE on hospital stays, are calculated using a partial derivative of E(y|X) with respect to TEE for age group j as:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2)

where j stands for age groups 65-69 or 70-79. [[??].sub.k] and [[??].sub.k] denote the estimated coefficients for variable k from the NB and logit results in Appendix 1. [??] and [??] show the predicted numbers of hospital stays among users and the probability of being a non-user. The partial effect for age group 80+ can be calculated using Equation (2) after deleting [[??].sub.TEE*j] and [[??].sub.TEE*j].

TEE is expressed as kcal burned per kg of body weight. One can also express it as total daily kcal burned by an individual with average body weight of 72.5 kg in this sample. In this case, a one-unit increase in TEE is equivalent to burning 72.5 kcal/day. This corresponds to walking the dog for 20 minutes, or a 20-minute walk for pleasure, or a 20-minute walk at 2.5 miles/hour on a firm surface. (28)

Given the coefficient estimates and Equation (2), one can estimate the impact of a one-unit increase in TEE (20-minute daily walk) on numbers of hospital stays. However, the partial effects can only be calculated for known values of individual characteristics since [??] and [??] are a function of independent variables. To do this, I divide each age group into three subgroups (inactive, moderately active, active) based on their current physical activity.

To calculate corresponding effects for all subgroups in each age group, I use all relevant values for TEE in each activity group (i.e., 0 [less than or equal to] TEE <1.5 for inactive) and subgroup mean values for all other variables. These results are presented in Table 4 for an individual in each subgroup at two different current LTPA levels.

RESULTS

Table 3 presents estimated coefficients for LTPAs. The NB model estimates the number of hospital stays among users, and the logit model predicts the probability of being a user. They imply that physical activity is negatively associated with any hospital stay and with the numbers of stays among users.

Table 4 shows that the impact of an additional 20-minute walk is higher for inactive people than for others. These results suggest that if an inactive 65-69 year old with a current activity level of 20 minutes per day walked an additional 20 minutes per day, his/her hospital stays would be decreased by 0.115 days/year (a 16% decrease). The effects for a similar individual aged 70-79 or 80+ would be 0.193 and 0.179 days per year, respectively (a 19% decrease). For an individual with higher current activity level (i.e., a 40-minute/day walk), the impact of an additional 20-minute/day walk on hospital stays is lower for all age groups. These results show that the gain from additional physical activity decreases as physical activity increases.

The partial effects show per capita gain from additional physical activity. One can estimate total number of fewer hospital stays using total number of people in each sub-group, and the per capita gain. To show the breakdown by age groups, I estimate the total gain as a result of an additional 20-minute/day walk for all age groups. To do this, I use conservative estimates (i.e., -0.115 rather than -0.134 for the 65-69 year inactive group) for each subgroup.

For inactive 65-69 year olds, these findings suggest that total hospital stays decrease by 65,614 days/year. Total impact is even higher for other age groups. For inactive older adults aged 70-79 and 80+, total gain is more than 201,000 and 125,000 days/year, respectively. This implies that an additional 20-minute/day walk by all inactive people aged 65+ decreases hospital stays by about 392,000 days/year. When compared to the 20.8 million inpatient days, or the total bed capacity in all hospitals, (29) potential gain from an additional 20-minute walk by inactive older Canadians accounts for about 2% of total annual inpatient days, or 1.2% of total hospital bed capacity.

However, these results imply that the potential gain from an additional 20-minute walk decreases as current physical activity increases. For instance, the total decrease in annual hospital stays for those who are at least moderately active would be less than 1% of total annual inpatient days.

DISCUSSION

This paper suggests that a small change in sedentary lifestyles for inactive older adults could translate into a substantial decrease in hospitals stays: adding some activity for inactive older adults decreases hospital stays by 16% to 19%. In terms of total hospital days, the results suggest that an additional 20-minute/day walk by inactive older adults decreases total inpatient stays by about 2%. If all older adults increase their activity level by the equivalent of a 20-minute/day walk, total inpatient days decrease by about 2.7%, equivalent to 1.7% of annual bed capacity in Canadian hospitals.

These results should be evaluated in the light of findings from related literature on health and physical activity. Epidemiological research firmly established positive effects of physical activity on people's health due to its strong association in preventing chronic diseases. (3,30) As suggested in the literature, these benefits can be derived in a short period of time. For instance, the studies on physical fitness, diabetes and obesity show that changes in fitness level and their influence on risk factors may occur within about six months after individuals reach the appropriate activity level. (10)

This paper may have some limitations due to its cross-sectional design. It is likely that unobservable factors can create bias in cross-sectional studies. For instance, if exercisers are inherently different from sedentary people, these unmeasured differences can partly explain the differences in utilization. To account for the bias, I included a comprehensive set of control variables. However, any unobserved factors not controlled by the included variables would still create bias. Therefore, further research dealing with these methodological issues may extend our understanding and provide additional insights on this issue.
Appendix 1. ZINB Regression Results

                                   NB                    Logit

                         Coefficient   t-value   Coefficient   t-value

TEE                          -0.098     -2.01         0.105      1.94
Age65-69                     -0.074     -0.60         0.348      2.46
Age70-79                     -0.084     -0.89         0.170      1.51
TEE*Age65-69                  0.008      0.13        -0.030     -0.44
TEE*Age70-79                  0.004      0.07         0.005      0.07
Alcohol                       0.0002     0.0003       0.333      3.24
Alcohol2                      0.001      0.04        -0.017     -0.81
Smoker                       -0.152     -1.14         0.041      0.27
Moderately Smoker            -0.112     -0.59        -0.087     -0.36
Former Smoker                 0.125      1.55        -0.138     -1.37
Former Mod Smoker            -0.106     -1.01        -0.273     -2.15
Fruit & Veggie               -0.017     -1.15        -0.019     -1.12
Handsmoker_home               0.016      0.17        -0.192     -1.99
Handsmoker_car               -0.096     -0.55        -0.088     -0.44
Handsmoker_bldg               0.051      0.39         0.075      0.45
Minority                      0.227      1.15         0.612      3.09
Immigrant                    -0.185     -2.16        -0.178     -1.55
Recent Immigrant             -1.348     -1.04         1.323      1.16
Poor Health                   0.210      3.01        -0.584     -7.11
Poor Mental Health           -0.102     -0.84         0.605      3.37
Stress                        0.164      1.62        -0.096     -0.80
Stress at Work                0.006      0.01         0.158      0.35
Life Dissatisfaction         -0.007     -0.06        -0.414     -2.38
Chronic Conditions            0.008      0.56        -0.148     -7.36
Live Together                -0.164     -1.75        -0.027     -0.27
No Income                     0.134      0.69         0.091      0.42
Low Income                    0.210      1.18         0.149      0.75
Medium Income                 0.207      1.16         0.176      0.90
Med-high Income               0.106      0.58         0.070      0.32
Household Size                0.035      0.54        -0.009     -0.10
Bedrooms                     -0.062     -1.59         0.017      0.42
Education                     0.046      1.76        -0.004     -0.12
Has MD                        0.056      0.30        -0.449     -2.60
BMI                          -0.007     -0.92         0.011      1.24
Male                          0.013      0.17        -0.555     -6.13
Repetitive Injury            -0.365     -2.04        -0.181     -0.90
Repetitive (Sports)           0.083      0.22         0.257      0.33
Other Injury                  0.185      2.07        -0.597     -5.39
Other Injury (Sports)        -0.454     -1.45         0.248      0.43
Occupation1                  -0.073     -0.37         0.027      0.08
Occupation2                   0.652      1.43         0.535      1.59
Occupation3                  -0.209     -0.82         0.133      0.56
Walk_<1                      -0.307     -3.01         0.083      0.74
Walk_1-5                     -0.103     -1.33         0.199      2.13
Walk_6-10                    -0.201     -1.57         0.022      0.17
Walk_11-20                   -0.152     -0.80        -0.163     -0.74
Walk_>20                     -0.216     -1.29         0.052      0.29
Bike_<1                      -0.701     -1.57         0.175      0.34
Bike_1-5                     -0.129     -0.37         0.392      1.22
Bike_6-10                    -0.046     -0.12         1.388      2.13
Bike_11-20                   -2.407     -4.43        -2.949     -1.55
Bike_>20                     -0.303     -0.88        -1.427     -1.47
Work(stand walk)             -0.346     -4.60         0.287      3.36
Work(light load)             -0.440     -3.75         0.425      3.80
Work(heavy load)             -0.227     -0.52         0.960      2.74
Constant                      2.579      6.15         2.047      4.15

Note: Sample size in all models is 18,196 individuals, aged 65 and
above. NB denotes the negative binomial regression model. The logit
coefficients predict the probability of being in the non-user group.
All models include provincial dummies.


Acknowledgement: The author thanks the conference participants at the 7th European Conference on Health Economics and the seminar participants at the University of Manitoba and TOBB University of Economics and Technology for their insightful comments and suggestions.

Conflict of Interest: None to declare.

REFERENCES

(1.) Statistics Canada. Canadian Community Health Survey 2.1. Available at: http://www.statcan.ca/english/concepts/hs/index.htm (Accessed January 10, 2009).

(2.) Katzmarzyk P, Janssen J. The economic costs associated with physical inactivity and obesity in Canada: An update. Can J Appl Physiol 2004;29(1):90-115.

(3.) Warburton DE, Nicol CW, Bredin SS. Health benefits of physical activity: The evidence. CMAJ 2006;174(6):801-8.

(4.) Haapanen-Niemi N, Miilunpalo S, Vuoir I, Pasanen M, Oja P. The impact of smoking, alcohol consumption and physical activity on use of hospital services. Am J Public Health 1999;89:691-98.

(5.) Keeler EB, Manning WG, Newhouse JP, Sloss ES, Wasserman J. The external costs of a sedentary life-style. Am J Public Health 1989;79:975-81.

(6.) Manning WG, Keeler E, Newhouse JP, Sloss E, Wasserman J. The Costs of Poor Health Habits. Cambridge, MA: Harvard University Press, 1991.

(7.) Garrett NA, Brasure M, Schmitz KH, Schultz MM, Huber MR. Physical inactivity: Direct cost to a health plan. Am J Prev Med 2004;27(4):304-9.

(8.) Sari N. Physical inactivity and its impact on healthcare utilization. Health Econ 2009;18(8):885-901.

(9.) Colditz GA. Economic costs of obesity and inactivity. Med Sci Sports Exerc 1999;31(11):S663-S667.

(10.) Mitchell TL, Gibbons LW, Devers SM, Earnest CP. Effects of cardiorespiratory fitness on healthcare utilization. Med Sci Sports Exerc 2004;36(12):2088-92.

(11.) Baun WB, Bernacki EJ, Tsai SP. A preliminary investigation: Effect of a corporate fitness program on absenteeism and healthcare costs. J Occup Med 1986;28:19-22.

(12.) Leigh JP, Fries JF. Health habits, healthcare use and costs in a sample of retirees. Inquiry 1992;29:44-54.

(13.) Bowne DW, Russell ML, Morgan JL, Optenberg SA, Clarke AE. Reduced disability and healthcare costs in an industrial fitness program. J Occup Med 1984;26:809-16.

(14.) Shephard RJ, Corey P, Renzland P, Cox M. The influence of an employee fitness and lifestyle modification program upon medical care costs. Can J Public Health 1982;73:259-63.

(15.) Shephard RJ, Corey P, Renzland P, Cox M. The impact of changes in fitness and lifestyle upon health care utilization. Can J Public Health 1983;74:51-54.

(16.) Shephard RJ. Twelve years experience of a fitness program for the salaried employees of a Toronto life assurance company. Am J Health Promot 1992;6:293-301.

(17.) Dunnagan T, Haynes G, Noland M. Healthcare costs and participation in fitness programming. Am J Health Behav 1999;23:43-51.

(18.) Canadian Fitness and Lifestyle Research Institute. Physical Activity Levels among Canadian Adults. Available at: http://www.cflri.ca/ (Accessed April 23, 2009).

(19.) Irwin ML, Ainsworth BE, Conway JM. Estimation of energy expenditure from physical activity measures: Determinants of accuracy. Obesity Res 2001;9:517-25.

(20.) Taylor AH, Cable NT, Faulkner G, Hillsdon M, Narici M, Van Der Bij AK. Physical activity and older adults: A review of health benefits and the effectiveness of intervention. J Sports Sci 2004;22:703-25.

(21.) Eisenberg D, Okeke E. Too cold for a jog? Weather, exercise, and socioeconomic status. B.E. J Economic Analysis Pol 2009;9 (1--Contributions): Article 25. Available at: http://www.bepress.com/bejeap/vol9/iss1/art25 (Accessed September 15, 2009).

(22.) Marmot M, Ryff CD, Bumpass LL, Shipley M, Marks NF. Social inequalities in health: Next questions and converging evidence. Soc Sci Med 1997;44(6):901 10.

(23.) Jones BH, Cowan DN, Knapik JJ. Exercise, training and injuries. Sports Med 1994;18(3):202-14.

(24.) Jones AM. Health econometrics. In: Culyer AJ, Newhouse JP (Eds.), The Handbook of Health Economics. Amsterdam, Netherlands: Elsevier, 2000;1A:265-346.

(25.) Winkelmann R. Econometric Analysis of Count Data. Berlin, Germany: Springer Verlag, 1997.

(26.) Lambert D. Zero-inflated Poisson regression, with an application to defects in manufacturing. Technometrics 1992;34:1-14.

(27.) Vuong QH. Likelihood ratio tests for model selection and non-nested hypotheses. Econometrica 1989;57:307-33.

(28.) Ainsworth BE, Haskell WL, Whitt MC, Irwin ML, Swartz AM, Strath SJ, et al. Compendium of physical activities: An update of activity codes and MET intensities. Measurement of moderate physical activity. Med Sci Sports Exerc 2000;32(9)Supplement:S498-S516.

(29.) Canadian Institute for Health Information. Quick Stats A-Z. Available at: http://secure.cihi.ca/ (Accessed March 5, 2009).

(30.) Myers J, Prakash M, Froelicher V, Do D, Partington S, Atwood JE. Exercise capacity and mortality among men referred for exercise testing. N Engl J Med 2002;346(11):793-801.

Acknowledgement:

Nazmi Sari, PhD

Author's Affiliation

Department of Economics & SPHERU, University of Saskatchewan, 9 Campus Drive, Saskatoon, SK S7N 5A5, Tel: 306-966-5216, Fax: 306-966-5232, E-mail: Nazmi.Sari@usask.ca
Table 1. Summary Statistics and Variable Description for Main
Variables

Variable Name     Variable Description                  Mean     S.D.

Hospital stays    Annual number of nights at a         1.175    4.576
                  hospital as a patient

TEE               Total daily energy expenditure       1.628    1.654
                  from LTPAs

Age65-69          Age group: 65-69 years               0.334    0.461

Age70-79          Age group: 70-79 years               0.494    0.500

Interaction terms for age and physical activity

TEE*Age65-69      Total daily energy expenditure for   0.633    1.330
                  those who are in age group 65-69

TEE*Age70-79      Total daily energy expenditure for   0.800    1.375
                  those who are in age group 70-79

Note: S.D. = standard deviation

LTPA = leisure-time physical activity

Table 2. Summary Statistics and Variable Description for Other
Variables

Variable Name            Variable Description

Alcohol                  Average daily alcohol consumption
Alcohol2                 Square of average daily alcohol consumption
Smoker                   Daily smoker
Moderately Smoker        Occasional smoker
Former Smoker            Former daily smoker
Former Mod Smoker        Former occasional smoker
Fruit & Veggie           Daily consumption of fruit and vegetables
                           (servings)
Handsmoker_home          Number of people who smoke inside home every
                           day
Handsmoker_car           Exposed to second-hand smoke in private
                           vehicle
Handsmoker_bldg          Exposed to second-hand smoke in public places
Minority                 Visible minority
Immigrant                Immigrant
Recent Immigrant         Lived in Canada less than 10 years
Poor Health              Self-rated poor or fair health
Poor Mental Health       Self-rated poor or fair mental health
Stress                   Self-perceived stress: extremely or quite a
                           bit
Stress at Work           Self-perceived stress at work: extremely or
                           quite a bit
Life Dissatisfaction     Dissatisfied from life
Chronic Conditions       Number of chronic conditions
Live Together            Married or common law partner
No Income                No household income
Low Income               Household income ($15,000-$29,999)
Medium Income            Household income ($30,000-$49,999)
Med-high Income          Household income ($50,000-$79,999)
Household Size           Household size
Bedrooms                 Number of bedrooms at home
Education                Highest level of education (4 categories)
Has MD                   Has regular medical doctor
BMI                      Body Mass Index
Male                     Sex of the respondents
Repetitive Injury        Had repetitive injury in the last 12 months
Repetitive (Sports)      Sports-related repetitive injury in the last
                           12 months
Other Injury             Had other injuries in the last 12 months
Other Injury (Sports)    Sports-related other injuries in the last 12
                           months
Occupation1              Occupations related to management, business,
                           finance, natural, social and health
                           sciences, education, religion, culture
Occupation2              Occupations related to sales and service
Occupation3              Occupations related to trades and
                           transportation, processing, manufacturing
                           and utilities
Walk_<1                  Less than 1 hour/week walking to work or
                           while doing errands
Walk_1-5                 1-5 hours/week walking to work or while
                           doing errands
Walk_6-10                6-10 hours/week walking to work or while
                           doing errands
Walk_11-20               11-20 hours/week walking to work or while
                           doing errands
Walk_>20                 More than 20 hours/week walking to work or
                           while doing errands
Bike_<1                  Less than 1 hour/week biking to work or
                           while doing errands
Bike_1-5                 1-5 hours/week biking to work or while doing
                           errands
Bike_6-10                6-10 hours/week biking to work or while
                           doing errands
Bike_11-20               11-20 hours/week biking to work or while
                           doing errands
Bike_>20                 More than 20 hours/week biking to work or
                           while doing errands
Work(stand walk)         Work habits (if relevant) and/or daily
                           physical activities: stand or walk
Work(light load)         Work habits (if relevant) and/or daily
                           physical activities: lift light loads
Work(heavy load)         Work habits (if relevant) and/or daily
                           physical activities: lift heavy loads

Variable Name             Mean     S.D.

Alcohol                  0.354    0.751
Alcohol2                 0.724    3.468
Smoker                   0.092    0.301
Moderately Smoker        0.016    0.130
Former Smoker            0.417    0.492
Former Mod Smoker        0.143    0.346
Fruit & Veggie           5.167    2.335
Handsmoker_home          0.127    0.382
Handsmoker_car           0.037    0.182
Handsmoker_bldg          0.083    0.279
Minority                 0.061    0.188
Immigrant                0.241    0.378
Recent Immigrant         0.007    0.050
Poor Health              0.236    0.435
Poor Mental Health       0.043    0.206
Stress                   0.094    0.290
Stress at Work           0.012    0.101
Life Dissatisfaction     0.028    0.174
Chronic Conditions       2.947    2.194
Live Together            0.627    0.499
No Income                0.135    0.401
Low Income               0.345    0.485
Medium Income            0.275    0.432
Med-high Income          0.165    0.329
Household Size           1.839    0.640
Bedrooms                 2.571    0.939
Education                2.294    1.340
Has MD                   0.958    0.221
BMI                      25.95     4.34
Male                     0.456    0.491
Repetitive Injury        0.049    0.213
Repetitive (Sports)      0.007    0.083
Other Injury             0.081    0.276
Other Injury (Sports)    0.007    0.082
Occupation1              0.047    0.189
Occupation2              0.023    0.136
Occupation3              0.028    0.166
Walk_<1                  0.130    0.333
Walk_1-5                 0.343    0.470
Walk_6-10                0.094    0.293
Walk_11-20               0.042    0.205
Walk_>20                 0.048    0.218
Bike_<1                  0.010    0.097
Bike_1-5                 0.011    0.108
Bike_6-10                0.003    0.048
Bike_11-20               0.001    0.031
Bike_>20                 0.001    0.034
Work(stand walk)         0.526    0.500
Work(light load)         0.205    0.408
Work(heavy load)         0.011    0.113

Note: S.D.=standard deviation

Table 3. Summary Results From the ZINB Regression

                                 NB                       Logit

Physical Activity
  Variables          Coefficients   t-values   Coefficients   t-values

TEE                    -0.098        -2.01       -0.105        -1.94
TEE*Age65-69            0.008         0.13        0.030         0.44
TEE*Age70-79            0.004         0.07       -0.005        -0.07

Specification            Test       p-value
  Tests               Statistics

Over-dispersion
  test                  13.66        <0.01
Vuong-test              16.50        <0.01

Note: NB denotes the negative binomial regression model. The sign of
the logit coefficients are reversed to reflect that the logit models
predict the probability of being in the potential user group. Full
results are presented in Appendix 1.

Table 4. Impacts of an Additional 20-minute Daily Walk on Hospital
Stays

                            Current Activity Level

                                   Inactive *

                        (less than a 30-min daily walk)

                        0 min                  20 min

Age 65-69
  Per capita change
    Days/year           -0.134                 -0.115
    %                    -16                    -16
  Total (days/year)                -65,614

Age 70-79
  Per capita change
    Days/year           -0.230                 -0.193
    %                    -18                    -19
  Total (days/year)               -201,267

Age 80+
  Per capita change
    Days/year           -0.212                 -0.179
    %                    -18                    -19
  Total (days/year)               -125,797

                            Current Activity Level

                              Moderately active

                            (30-60 min daily walk)

                        30 min                40 min

Age 65-69
  Per capita change
    Days/year           -0.086                -0.079
    %                    -16                   -16
  Total (days/year)               -23,439

Age 70-79
  Per capita change
    Days/year           -0.133                -0.121
    %                    -19                   -19
  Total (days/year)               -56,332

Age 80+
  Per capita change
    Days/year           -0.179                -0.163
    %                    -19                   -19
  Total (days/year)               -29,042

                            Current Activity Level

                                   Active *

                          (60 min & more daily walk)

                        60 min                80 min

Age 65-69
  Per capita change
    Days/year           -0.059                -0.050
    %                    -16                   -16
  Total (days/year)               -13,693

Age 70-79
  Per capita change
    Days/year           -0.089                -0.074
    %                    -19                   -20
  Total (days/year)               -26,183

Age 80+
  Per capita change
    Days/year           -0.137                -0.144
    %                    -19                   -19
  Total (days/year)               -15,679

Note: * An active individual is one with a TEE of more than 3 kcal
per kg of body weight. This corresponds to a low intensive 1-hour/day
walk. An inactive individual, on the other hand, is one with a TEE of
less than 1.5 kcal per kg of body weight. This corresponds to less
than a 30-minute/day walk. (17)
COPYRIGHT 2010 Canadian Public Health Association
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2010 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:QUANTITATIVE RESEARCH
Author:Sari, Nazmi
Publication:Canadian Journal of Public Health
Article Type:Report
Geographic Code:1CANA
Date:Sep 1, 2010
Words:4731
Previous Article:Rabies in a Calgary puppy adopted from the Arctic.
Next Article:The additive effects of alcohol and benzodiazepines on driving.
Topics:

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters