1. Physical Activity Assessment Techniques
Validity, Reliability, Tools

2. Introduction
walking-exercise-obesity-weight

3. Where are we?
You should be familiar with the PA guidelines and the ways PA can be quantified
The purpose of this lecture is to help you understand and identify multiple methods to quantify PA so that you can determine if your client meets the guidelines (and prescribe PA)
We must determine what makes one assessment technique “better” than another

4. Measurement Issues How can one monitor be “better” than another?
Consider:
Sensitivity to change (precision)
Reactivity
Comfort to respondent (feasibility)
Cost of administration (feasibility)
The above can be considered your preference
Two considerations are far more important – validity and reliability

5. Validity + Reliability
Valid + Reliable +
Reliable w/
poor Validity
Poor

6. Statistical Note
Reliability and Validity are generally expressed as correlation coefficients
Range from -1 to +1
Stronger (and better) relationships are seen closer to 1
A correlation of 0.00 basically means there is no relationship (poor reliability, poor validity)

7. Recap of Measurement Issues
Evaluate each measurement tool based on these concepts
Sensitivity to change (precision)
Reactivity
Comfort to respondent (feasibility)
Cost of administration (feasibility)
Validity (accuracy)
Reliability (consistency)

8. Physical Activity Assessment Tools: Figuring Out Energy Expenditure and Minutes of Activity

9. Activity Monitoring Tools
Currently, there is a strong relationship between the feasibility of a PA assessment tool and it’s validity – and that relationship is in the wrong direction
We’ll tackle these different methods in order of weakest validity to strongest
Keep in mind that the tool you select is based on what you want to know and how you are going to use it!!

10. Physical Activity Assessments (Feasibility-Validity)
Questionnaire
Diary
Pedometers
HR monitors
Accelerometers
Feasibility
Direct observation
There is an inverse association between feasibility and validity of PA instruments
I. calorimetery
DLW
Validity

11. Self Report Tools
Subjective assessment tools are at the mercy of those providing the information
Self-Report Instruments
Recall
Diaries/Logs
Questionnaire
These tools require the client identify their own PA levels (to some degree)
Use time and intensity of PA to further extrapolate to EE

12. Recall Example: 7-Day PAR (Physical Activity Recall)
One week period, interview based, participant recalls:
Sleep
Moderate (minutes)
Hard (minutes)
Very Hard (minutes)
Minutes working on strength/flexibility
Note: Interviewer script, memory cues

13. Diary/Log
Instead of having the subject recall activity, perhaps validity/reliability can be improved by having them keep constant measurement?
Bouchard 3-Day Physical Activity Diary
Minutes of PA and EE
Note:
Short intervals, MET-reference guide

14. Questionnaire
Questionnaires do not ask clients to recall a specific time, nor do they use a continuous log
Simply ask the client how active they are normally
Some are more detailed than others
Example: Godin Leisure-Time PA Questionnaire

15. Other Self-Report Instruments
Numerous examples:
ACLS Questionnaire
Paffenbarger Physical Activity Recall
CARDIA Physical Activity History
Modifiable Physical Activity Questionnaire
Yale Physical Activity Survey
Historical Physical Activity Questionnaire
NHANES
Etc…

16. Other Self-Report Instruments
Numerous examples:
ACLS Questionnaire
Paffenbarger Physical Activity Recall
CARDIA Physical Activity History
Modifiable Physical Activity Questionnaire
Yale Physical Activity Survey
Historical Physical Activity Questionnaire
NHANES
Etc…

17.                                                                                                                                                                         
No Caption Available.

18. Other Self-Report Instruments
Numerous examples:
ACLS Questionnaire
Paffenbarger Physical Activity Recall
CARDIA Physical Activity History
Modifiable Physical Activity Questionnaire
Yale Physical Activity Survey
Historical Physical Activity Questionnaire
NHANES
Etc…

19. NHANES
For NHANES, participants are asked very specific questions on about 10 physical activities (like walking, cycling, aerobics, playing sports, etc…)
“On how many of the previous 30 days did you participate in ________?”
“On average, when you participated in _______, how long did you do this activity?”

20. Remember this?
Data show that nearly four in ten (38.3 percent) adults reported no participation in leisure-time physical activity.
Does knowing how PA was assessed change your perspective on inactivity in the US?

21. Self-Report: Validity and Reliability
In general, reliability of self-report measures are low-moderate
Low to non-existent reliability for light activity
High reliability for vigorous activity
In general, validity of self-report measures are moderate for energy expenditure and low for minutes of physical activity
Godin ~ compared with kcal/day

22. Self-Report Strengths: Low time, cost, work commitment (feasible)
When used with MET-compendium, both energy expenditure and minutes of activity can be identified
Can identify time-frame easily (past month, year, lifetime, etc…)
Allows determination of the “dimensions” of physical activity with both qualitative and quantitative information

23. Self-Report Limitations:
Biased towards structured, high intensity (exercise)
Low-moderate validity and reliability
Not good for estimating minutes of PA (alright at total EE)
Participant ignorance (what’s moderate again?)
Certain groups may have poor recall ability (children)
Potential content validity problems (participants may misinterpret questions)
Social desirability

24. Physical Activity Assessment Techniques
Gadgets

25. Physical Activity Assessments (Feasibility-Validity)
Questionnaire
Diary
Pedometers
HR monitors
Accelerometers
Feasibility
Direct observation
There is an inverse association between feasibility and validity of PA instruments
I. calorimetery
DLW
Validity

26. Pedometers
Simple pedometers are the first objective monitoring tool we will discuss
Provide information on ambulatory physical activity, priced between $20-$300
Walking:
Is the #1 ranked physical activity in the US
Accounts for a major portion of total daily activity
Confers substantial health benefits

27. Simple Question How valid do you think a pedometer is?
High validity in measuring steps, but…
Moderate validity in measuring physical activity ( )
Pedometers assess steps/day, not minutes of activity, or intensity, or energy expenditure
Construct Validity with minutes/EE
Raw Volume of PA

28. Pedometers
Although this is a limitation, pedometers are good at what they are supposed to do
Due to their cost and simplicity, they are widely studied and used
We even have separate recommendations for pedometers – and you all probably know it!

29. Target Steps/day
Hanato (1997), in a Japanese sample, found that subjects getting 10,000 steps/day were more likely to have lower blood pressure and %BF
This “10,000” was pulled from thin air
Nice round number, easily remembered, easily interpretable to clients (not like METs or “moderate” activity
In Japan, the term for pedometer is manpo-kei, which is literally translated to “10,000 step meter”

30. Versus the 30 minutes (old rec)
Welk et al. found that 73% of subjects that averaged >10,000 steps/day performed >30 minutes of moderate activity
This is due to a moderate construct validity between steps/day and minutes of activity
Overall, it appears that getting about 10,000 is somewhere close to the recommended amount of PA

31. Pedometer-Specific Groups
Pedometers alone can be used to categorize clients into physical activity groups:
Sedentary: <5000 steps/day
Low Active: steps/day
Somewhat Active: ,999 steps/day
Active: 10,000-12,499 steps/day
Highly Active: 12,500 steps/day
These targets can be used to motivate clients outside of the gym

32. Old-Order Amish
In 2004, Bassett completed a study analyzing the steps/day (physical activity) of an old-order Amish community
Besides shaming typical US Adults, it demonstrates what physical activity levels may have looked like 150 years ago
Keep in mind that ~ 10,000 steps indicates they meet US PA guidelines and 12,500 steps is Highly Active

35. Amish Men: 0% Obesity, Women: 9% Obesity
Self-Report
US Adult Guidelines: Minimum 150 min. of moderate activity or 75 min. vigorous
Amish Men: 0% Obesity, Women: 9% Obesity

36. Pedometer Strengths: Limitations: Objective
Inexpensive and easy to use
Moderate Validity
Interpretable
Specific targets exist for prescription
Limitations:
Doesn’t capture all PA
Not direct minutes of PA or EE
Reactivity?

37. Physical Activity Assessments (Feasibility-Validity)
Questionnaire
Diary
Pedometers
HR monitors
Accelerometers
Feasibility
Direct observation
There is an inverse association between feasibility and validity of PA instruments
I. calorimetery
DLW
Validity

38. Accelerometer
Due to the lack of “other” activities being captured by the pedometer, accelerometers have been created
Walking and running 20 steps are equally weighted in pedometry
Accelerometers can tell the difference

39. Instrumentation / Principles
Detection of acceleration:
Piezo-electric bender element is sensitive to acceleration - transducer bends and a proportionate electrical charge created
Amount of movement is summed over time
Movement is reported in counts/time
Real time recording allow temporal patterns to be observed

40. Data Output Whole-Day Accelerometry Adult
Walk
Walk
Lunch TV
Sleep
Off

41. Outcome Measures Counts/min Kcal/min or METS
Average movement count (total activity)
Counts/min
Equations to predict EE from movement counts
Kcal/min or METS
Time spent being active
Number of minutes above threshold (cut point derived from regression equation)
Number of bouts of activity
Number of times with continuous movement greater than threshold
Ie – were they active for 10 continuous minutes?

42. Measurement Issues Accelerometry
Validity
Indirect calorimetry lab r = 0.80 to 0.90 lab
Indirect calorimetry field r = 0.40 to 0.60
Reliability
Inter-instrument reliability consistently high (r > 0.90)
Sensitive to change
Non-reactive
Acceptable to subject
Acceptable cost? (units $200 to 500, interface $500)

43. Development of Prediction Equations and Activity “Cutpoints”
Activity counts
VO2 or energy expenditure
Light
Moderate
Vigorous
Characterize relationship between movement counts oxygen consumption
To avoid errors associated with estimating EE, many individuals have begun using activity cutpoints based on VO2 or EE values. Here, the goal is to determine the number of counts associated with different levels of EE and then estimate the amount of time spent in light, moderate and vigorous activity.
This procedure has good general application when working with adults but there is still some problems associated with individual variability in scores. There are also some major problems associated with using this technique with children.
The reason is that these cutpoints are typically developed using continuous activity which children do not typically perform.
I’ll provide an example on the next slide

44. Application of Count Thresholds for Processing Accelerometer Data
If a threshold accurately defines a particular level of activity, the time spent in activity can be determined.

45. NHANES
Recall that about 40% of US adults report no leisure-time activity
Based on what assessment of PA…?
NHANES recently introduced accelerometry into their survey
Objective assessment of PA
How many people “meet” guidelines?

46. Males, Moderate and Vigorous Activity x Age Group
Remember: This is the MEAN or average minutes

47. Females, Moderate and Vigorous Activity x Age Group
Remember: This is the MEAN or average minutes

49. How does that compare?
Approximately 5-10% of Americans meet or exceed the recommendation of 30 minutes of at least moderate activity on 5-7 days of the week
Objective vs Subjective methods of assessment

50. Limitations:
Accelerometers, while very good at what they do, are not without limitations:
Not sensitive some movements
Not sensitive to grade
Not sensitive to increased weight
Restricted to mainly ambulatory activity

51. Physical Activity Assessments (Feasibility-Validity)
Questionnaire
Diary
Pedometers
HR monitors
Accelerometers
Feasibility
Direct observation
There is an inverse association between feasibility and validity of PA instruments
I. calorimetery
DLW
Validity

52. Gold Standard Methods
If you really need to know energy expenditure – then doubly labeled water is THE best way to do it
Metabolism boils down to O2 consumption and CO2 production
If I know how much oxygen you use and carbon dioxide you produce, then I know how many calories you burn (not an estimate)
How can we keep track of O2 and CO2?

53. Capable of TEF + PAEE + RMR
Few options
1) Metabolic Analyzers (Indirect calorimetery)
Lab-based
Portable units
Capable of TEF + PAEE + RMR

54. Few options
2) Metabolic Chamber
Pennington Biomedical Research Unit:

55. Few Options
Even portable metabolic analyzers are restrictive – not applicable for long-term energy expenditure
Doubly labeled water combines with metabolic analyzer estimates of RMR and TEF to determine PAEE over the course of multiple days with no restrictions

56. Few Options 3) Doubly Labeled Water Free living estimates
Drink known amount of two stable isotopes (deuterium and oxygen-18)
Collect urine on “drink day” and “end day” to determine the quantity of the isotopes used in metabolism (CO2 production)
Subtract RMR and TEF = PAEE

57. Doubly Labeled Water Precise kcal estimates
Can be used from 3 days to 2 weeks
Lacks estimates of minutes of PA or any other qualitative information
Cost ~ $800 per person, per measurement period (depends on size)
Difficult to analyze (and that costs more money)

58. New Technologies Feasibility Validity Pattern Recognition Monitors
Diaries
Self-reports
Pedometers
HR monitors
Accelerometers
Feasibility
Direct observation
I. Calorimetry
DLW
Validity

59. Gadgets Pattern Recognition Monitors
Combine multiple PA assessment methods – plus a few new ones
High Validity
Cost $80-$5000
Combine the strengths of the previous methods and advanced equations to accurately assess:
What you do
When you do it
How intense
Energy Expenditure and Minutes of Activity

60. Gadgets Kam Monitor Cost ~ $80 Pedometer + Accelerometer
Output ~ METs (Kam Points)
It’s like Facebook, but for physical activity
Corporate Wellness, School-Health

61. Gadgets
Actiheart
A single-unit instrument that uses accelerometry in combination with electrocardiogram (ECG) signals to determine respiration, movement, and intensity.
HR monitor + Accelerometer
Respironics/Mini Mitter, Bend, OR –

62. Actiheart monitor Combines activity monitor and heart rate monitor.
Branch chain algorithms used to estimate EE.
Branched-equation modeling used to detect conditions most accurately estimated by: heart rate, accelerometry or both measures combined.
The main sensor and left lead are connected by wire and attached directly to the chest via standard ECG electrodes
$750 per/unit.
60s epochs for up to 11 continuous days.

63. Actiheart Output / Report

64. Gadgets
IDEEA Monitor
Intelligent Device for Energy Expenditure Activity
Places 5 accelerometers in 5 locations on body
Wires connect sensors and feed into a small computer

65. IDEEA- Characteristics
A light (59-grams) data
collection microcomputer.
Five sets of sensors:
Chest
Thighs
Soles of the feet (2).
It can measure angles of body segments and movement (acceleration) in 2 directions.

66. IDEEA- Output Data

67. IDEEA Reports- Detailed gait profiles

69. IDEEA
Extremely invasive
Cost ~ $4000-$5000
Extremely valid/reliable

70. Gadgets SenseWear Pro Armband (research grade) www.bodymedia.com
GoWearFit
BodyBugg
Non-invasive armband that combines temperature readings with pedometer and accelerometer type monitoring

71. SenseWear Mini Armband (Mini)

72. SenseWear Pro Armband High Heat Flux High Motion Low Motion
ALGORITHM
High Motion
Low Motion
ALGORITHM
ALGORITHM
Low Heat Flux

73. SWA output

74. SWA output
Biking
Heat Flux
Energy Expenditure

75. SWA output
Energy Expenditure
Heat Flux
Acceleration

76. Sensewear Reports

77. Gadgets Sensewear ~ $250 $1900 software
BodyBugg or GoWearFit ~ $
$85 yearly payment (online software)

78. Calabro and Welk 30 people wore two armband monitors for 2 weeks.
No significant differences in TEE.
More recent versions within 20 kcal of DLW

79. Summary
Multiple methods exist to determine the various quantities of physical activity
As most clients are concerned with diet, PA, and weight management, these are certainly useful tools
Keep in mind strengths and limitations of each method – what are they good at? Not so good?