Review

Fitness tests and occupational tasks of military

interest: a systematic review of correlations
Veronique D Hauschild, David W DeGroot, Shane M Hall, Tyson L Grier,
Karen D Deaver, Keith G Hauret, Bruce H Jones

Injury Prevention Program, ABSTRACT

Army Public Health Center Physically demanding occupations (ie, military, firefighter, What this paper adds
(Provisional), Aberdeen Proving
Ground, Maryland, USA law enforcement) often use fitness tests for job selection
or retention. Despite numerous individual studies, the
▸ Military, firefighting and law enforcement
Correspondence to relationship of these tests to job performance is not
Veronique D Hauschild, Injury employees are often required to perform
always clear.
Prevention Program, Army physical fitness tests for job selection,
This review examined the relationship by aggregating
Public Health Center placement and/or retention.
(Provisional), 5158 Blackhawk previously reported correlations between different fitness
▸ Studies regarding the relationship between
Rd, Aberdeen Proving Ground, tests and common occupational tasks.
common fitness tests and physical job
MD 21010-5400, USA; Search criteria were applied to PUBMED, EBSCO,
veronique.d.hauschild.civ@ performance have been inconsistent.
EMBASE and military sources; scoring yielded 27 original
mail.mil ▸ Aggregated correlation data from applicable
studies providing 533 Pearson correlation coefficients (r)
studies can be used to describe a relationship
Received 8 March 2016 between fitness tests and 12 common physical job task
between core fitness components and 12
Revised 16 August 2016 categories. Fitness tests were grouped into predominant
Accepted 11 October 2016 common task categories relevant to military
health-related fitness components and body regions:
Published Online First and other physically demanding occupations.
cardiorespiratory endurance (CRe); upper body, lower
3 November 2016 ▸ Fitness tests that measure cardiorespiratory
body and trunk muscular strength and muscular
endurance, lower body strength and upper
endurance (UBs, LBs, TRs, UBe, LBe, TRe) and flexibility
body muscular endurance are particularly
(FLX). Meta-analyses provided pooled r’s between each
relevant when assessing employees’ health
fitness component and task category.
status for these jobs.
The CRe tests had the strongest pooled correlations with
most tasks (eight pooled r values 0.80–0.52). Next were
LBs (six pooled r values >0.50) and UBe (four pooled r
values >0.50). UBs and LBe correlated strongly to three can provide stronger associations ( predictive value)
tasks. TRs, TRe and FLX did not strongly correlate to with performance outcomes than component-based
tasks. fitness tests,9 they can be logistically complex and
Employers can maximise the relevancy of assessing have limited applications since they are designed
workforce health by using fitness tests with strong and tested to reflect specific scenarios, tasks and
correlations between fitness components and job equipment. Component fitness tests are generally
performance, especially those that are also indicators for cheaper and more broadly applicable, but they may
injury risk. Potentially useful field-expedient tests include be challenged if the relationship between them and
timed-runs (CRe), jump tests (LBs) and push-ups (UBe). job performance has not been adequately
Impacts of gender and physiological characteristics (eg, demonstrated.11 12
lean body mass) should be considered in future study As an example, the US Army Physical Fitness
and when implementing tests. Test (APFT) is a three-test battery consisting of a
2 min push-up test, a 2 min sit-up test and a two
mile run for time.2 A passing score on the APFT is
used as a key determinant for US Army service
INTRODUCTION entry and retention to ensure a base level of phys-
Occupations such as the military, firefighting and ical fitness for every soldier.2 Instituted in 1980,
law enforcement require employees to perform vig- the APFT has been criticised because of insufficient
orous, physically demanding tasks such as dragging evidence validating its association with military job
victims to safety, moving quickly and carrying performance and combat.2 12–14 New batteries of
heavy loads. Personnel are often required to dem- fitness tests proposed in 2002 and 2010 were also
onstrate specified levels of physical capability for considered not validated, so were not implemen-
job selection, placement and/or retention.1–8 There ted.2 13 As military services redefine combat readi-
are two primary types of physical capability tests. ness requirements to support full gender
One type, known as, a ‘criterion task’ or ‘content- integration in combat positions,9 15 16 the APFT
based’ performance test uses standardised job-task has again been part of a reevaluation effort.13 17
simulations, such as lifting and carrying a manne- Fitness tests used as job selection, promotion or
quin a specific distance.6 9 The second type of test retention criteria should be linked to capabilities
To cite: Hauschild VD, measures general physical ‘constructs’ or ‘compo- critical to the nature of the job in order to address
DeGroot DW, Hall SM, et al. nents’ that are important to successful job perform- legal and antidiscrimination requirements.9 11 18
Occup Environ Med ance, such as cardiovascular endurance and Towards this goal, numerous studies have evaluated
2017;74:144–153. muscular strength.2 9 10 Though criterion task tests the relationship between individual fitness tests and
144 Hauschild VD, et al. Occup Environ Med 2017;74:144–153. doi:10.1136/oemed-2016-103684
 Review

Table 1 Physical fitness component groups and occupational task categories used for correlation meta-analyses
Four health-related physical fitness component groups* Twelve common occupational task categories†

Cardiorespiratory Ability to sustain low-intensity muscle contractions for extended ▸ Lift and lower (single)‡
endurance (CRe) period of time. Gold-standard physiological measure is the body’s One time maximum lift of equipment
maximum rate oxygen (VO2max). Also known as ‘aerobic fitness’, ▸ Lift and lower (repeated)‡
‘aerobic capacity’ and ‘stamina’. Repeated lifting equipment on/off ground or vehicles
Example tests: ▸ Lift and carry‡
▸ Machine tests (treadmill, step) that include measuring or Carry various equipment items various distances
estimating VO2max ▸ Stretcher carry‡
▸ Surrogate measures: distance run tests for time (eg, 1–3 miles), A specific type of lift and carry task (two person)
fixed-distance runs (12 min) ▸ Push–pull‡
Muscular strength Ability to exert maximal force against a fairly immovable object for Manual movement of equipment along a surface (not lifting)
(UBs, LBs, TRs) very brief period (seconds). Measurements reflect force; no ▸ Casualty drag‡
physiological gold standard exists.Example tests: Life-saving task may include extrication and/or different carrying techniques
▸ One-repetition maximum (1RM) lifts using various free weights ▸ Dig‡
or isometric machines (UBs and LBs, depending on machine) Establish fighting position, structural support, fill sandbags
As a surrogate: Explosive strength is a form of power that refers to ▸ March/walk (with a load)‡
ability to use maximum energy to rapidly project object or body in Move body long distances wearing some form of gear
a single maximum effort in a very brief amount of time (seconds) ▸ Move fast§
Example tests: With or without change of direction for short distances
▸ Jumps (eg, vertical, broad jump, squat) (LBs) ▸ Climb§
▸ Shot put, ball throws (UBs) Includes scale, jump, descend stairs, walls, vehicles, obstacles
▸ Crawl§
Muscular endurance Ability to conduct high-intensity muscle contractions repeatedly for
High and low techniques
(UBs, LBs, TRs) relatively short periods (30 s to 2 min). Measurements reflect force/
▸ Multiactivity‡,§
time but no physiological gold standard exists.
Combination of three or more tasks; ‘obstacle course’; ‘circuit’
Example tests:
▸ Push-ups and pull-ups (UBe)
▸ Weight lifting maximum repetitions (UBe)
▸ Endurance squats (LBe)
▸ Sprints and shuttles (LBe)
▸ Sit-ups (TRe)
Flexibility (FLX) Ability to flex or lengthen various parts of the body. Includes static
(steady hold) or ballistic (repeated, rapid) forms.
Example test:
▸ Sit-and-reach (static, back and hamstring stretch)

*The four primary activity-based health-related physical fitness components in conjunction with definition and example tests used to provide measurements.7 19
†Categories from review of tasks common to US Army and other military ground forces13 17 24 25 and found to be similar to tasks in occupational studies.
‡Includes manual movement of equipment, supplies, people; variables include amount of weights, duration, distances, heights, terrain and environmental conditions and other
preceding/concurrent activities. Since stretcher carry has been historically studied as a unique task, it is considered separate from the lift and carry task.
§Movement of body activities include variable loads, distances, duration, heights, terrain and environmental conditions and other preceding/concurrent activities.

various job tasks.17 Many studies use regression models, but data and synthesise the outcomes using a meta-analytical tech-
the selection of different tasks and test variables has made it dif- nique to describe the collective evidence.20–22 This type of
ficult to compare study outcomes. Several studies have also review did not require human use consent. The comprehensive
examined the relationship by calculating Pearson correlation investigative team included a military physiologist, a public
coefficients (r values) between measures of individual fitness test health scientist, a kinesiologist, a physical therapist, a
and occupational task performance. However, findings have not physician-epidemiologist and two statisticians. Because the
been consistent or robust and have been constrained to the project was conducted in response to a directed military initia-
existing studies’ specific settings, tasks and fitness tests. tive, the data collected and analysed were archived in a military
As a result, employers have continued to be confronted technical report.17
with the dilemma of how to select the best and most practical In order to evaluate the data amassed from selected studies,
means to promote, monitor and test the physical fitness of individual fitness tests were organised into the four components
employees.1 4 5 9 13 This systematic review aggregated data from of health-related physical fitness activity: cardiovascular endur-
applicable studies to describe the relationship between core ance (CRe), muscular strength, muscular endurance and FLX
components of health-related physical fitness activity (cardio- (table 1). Since fitness tests for muscular strength and muscular
respiratory endurance (CRe), muscular strength, muscular endurance primarily test one region of the body, these two com-
endurance and flexibility (FLX))7 19 and the performance of ponents were also grouped into predominant body regions.
common physically demanding job tasks. The physical fitness These groups included strength and muscular endurance of the
components and associated tests that most strongly correlate upper body (UBs and UBe), lower body (LBs and LBe) and trunk
with common physical job tasks provide a basis for the selection (TRs and TRe)). Skill-related fitness components (eg, agility,
of relevant health-related occupational fitness tests. coordination, balance, power and speed) were not a focus of this
review. However, some common tests of power or speed (eg,
METHODS jump tests, sprints, shuttle runs) encompass elements of muscular
Study design strength and muscular endurance so were included.9 23
Systematic review methodology was used to identify and select In addition to grouping the fitness tests, 12 task categories
applicable quality original studies, extract the desired correlation (table 1) were identified as representative of the common
Hauschild VD, et al. Occup Environ Med 2017;74:144–153. doi:10.1136/oemed-2016-103684 145
 Review

physically demanding occupational tasks of interest. These task characteristics, health-related component fitness tests, applicable
categories were derived from a review of common military tasks job tasks with metrics and measurements and correlation values.
to identify those of greatest relevance to all Army person- Individual selection study bias was minimised by including sig-
nel.13 17 24 25 However, these task categories are also relevant nificant and non-significant correlations from all selected
to civilian public safety and emergency response-related occupa- studies. When applicable, reported correlations were standar-
tions such as firefighting and law enforcement.26–28 While the dised using absolute r values.
exact task details can vary among different military services or
civilian occupations, the general functional movements and
Statistical analyses and interpretation
types of physical demands represented by each task category
To synthesise the extracted data, investigators grouped the r
were deemed to be similar. For example, victims or casualties
values into fitness component and task category combinations
can be different sizes and weights, and they can be carried over
and mathematically combined them into single representative
varying distances and surfaces. Though individual studies used
effect estimates.21 29 Meta-analysis techniques were applied
unique quantitative metrics (eg, specific weights or distances),
using the number of studies, original study r values and sample
the variation captured by the aggregate data more appropriately
sizes to generate pooled r values between each fitness compo-
replicates the real-world variability.
nent and job task combination.29 For fitness component and
task combinations with only one study, a pooled r was not cal-
Search criteria for studies
culated. Original r values were converted to a common test
Eligible literature included original studies of military and civil-
metric (z-values and corresponding variance) using Fisher’s r to
ian healthy adults, ages 18–65 years, who performed one or
z transformation.29
more of the identified tasks. The population of interest was
intended to reflect the diversity of personnel serving in the US ! "
military as well as other occupational sectors (ie, firefighters and 1 1 þ ri 1
Zri ¼ ln s2z ¼ ð1Þ
law enforcement). Included study populations spanned a wide 2 1 # ri n#3
range of demographic and motivational variables and included
Fisher’s z values from the original studies were combined using
extremely fit, athletic personnel and less fit personnel. Selected
fixed or random effect models depending on their homogen-
studies had to provide Pearson correlation coefficients (r values)
eity.21 The presence of heterogeneity was tested using the
between physical fitness tests and performance measurements
Cochran Q statistic, with p<0.05 as the level of significance. An
for one or more tasks.
I2 test was also performed to quantify heterogeneity. The fixed
effect model was used if there was no evidence of heterogeneity.
Search strategy
In cases of statistically significant heterogeneity, the pooled
The systematic literature search included English language
effect estimate was determined using the random effect model.
studies published in peer-reviewed journals and scientific tech-
This was performed using formulas within Excel file of
nical reports between 1970 and February 2013. Sources
extracted data. CIs 95% and p values were calculated for each
included PubMed (MEDLINE), Biomedical Reference
pooled correlation coefficient. Subgroup analyses of separate
Collection, Academic Search Premier, Nursing & Allied Health
gender data and specific fitness tests were conducted for
Collection: Comprehensive, Cochrane Methodology Register,
task-test combinations with correlations from at least two
CINAHL & CINAHL Full Text and EMBASE. In addition, the
studies for one or more task for each physical component. The
Defense Technical Information Center (DTIC) was searched for
mean weighted correlations were also calculated for each fitness
relevant military technical reports and subject matter experts
component for all task categories combined.
contacted for additional references. Search terms were adapted
Since empirical guidelines for interpreting the magnitude of
to address different database systems but included variations of
correlation coefficients are lacking, a scale to interpret the
the following: ‘test, requirement or standard’, ‘performance’ or
pooled r values was created a priori. The five-tiered scale was
‘capability’, ‘functional ability’ and ‘work’, ‘job’ or ‘occupation’,
based on a review of guidelines for similar types of relationships
or ‘task,’ as well as ‘physical fitness’, ‘mobility’ and a variation
found in human performance and social sciences.17 30 31 For
of each of our selected key physical fitness component terms:
example, r≥0.70 has been described as a very strong or excellent
‘cardiorespiratory’, ‘aerobic fitness’, ‘muscle strength’, ‘muscle
linear relationship, r≥0.50 as a demarcation for large or strong
endurance’ and ‘flexibility’.
relationships and r≤0.30 or<0.10 as a weak
relationship.17 30 31
Study selection and data extraction
Two team investigators (VDH, DWD) conducted separate data-
base searches and merged results into a single Endnote file. RESULTS
Investigators reviewed titles and abstracts, excluding those that The literature search and study selection are summarised in
did not meet study criteria or provide Pearson correlation coeffi- figure 1. Of 273 studies identified for full text review, most
cients. Eligible full-text studies were reviewed and scored by were excluded due to the lack of requisite statistical analyses. Of
each investigator to ensure study relevance and quality. The the 33 studies selected for scoring, investigators’ scores were
quality scoring tool included 10 appraisal criteria adapted from generally consistent and brief discussions reconciled differences.
similar reviews:21 23 (1) study objective, (2) design and Five studies were eliminated because study procedures and stat-
methods, (3) sample size, (4) population and setting character- istical methods were not suitable for this analysis. Two studies
istics, (5) control for confounders in study design to isolate described data from the same study, so one was eliminated.
effect of interest, (6) repeatability, (7) data presentation that While the quality scores of the remaining studies varied, each
demonstrated controls for confounding, (8) analyses techniques, study provided adequate methodological validity to minimise
(9) adequacy of results and (10) variability. Scoring differences individual study biases for the purpose of this review.
between investigators were discussed to achieve consensus. Data The 27 selected studies (table 2) represented a variety of US and
extracted from selected studies included study and population international military and non-military healthy adult populations.
146 Hauschild VD, et al. Occup Environ Med 2017;74:144–153. doi:10.1136/oemed-2016-103684
 Review

Figure 1 Study selection flow chart.

Studies included 13 military population studies,32–36 specific pooled r values with relatively robust data included
38 39 42 44–46 53 54
10 firefighter, law enforcement, or peace officer hand grip tests for UBs, the standing broad jump (SBJ) and ver-
studies,26–28 37 40 43 47 49 50 52 and 4 other relevant studies of tical jump (VJ) for LBs, push-ups for UBe and sprint tests for
healthy civilian populations.41 48 49 51 The selected studies pro- LBe. Though data were too limited to assess differences among
vided 533 distinct Pearson r correlation coefficient values between males and females for all tasks, separate correlation data for
physical fitness tests and job tasks. The most frequently studied male and female sample populations were adequate for calculat-
fitness components were UBs (18 studies, 122 correlations) and ing pooled r values between each fitness component and the
UBe (20 studies, 117 correlations). The least studied fitness com- crawl and stretcher carry tasks. The CRe component had the
ponent was flexibility (FLX, addressed in only three studies yield- highest pooled correlations for both genders for the stretcher
ing 15 correlations). Of the task categories, multiactivity was the carry (male pooled r=0.63, female pooled r=0.60) and the
most frequently studied (11 studies, 76 correlations), followed by crawl task (male pooled r=0.63, female pooled r=0.74).17
the lift and carry (6 studies, 64 correlations) and stretcher carry (6 Table 4 summarises the overall strengths of the pooled
studies, 59 correlations). No correlations were found between the r values for each fitness component. The CRe component had
crawl task and LBe, and only one study evaluated the correlation the greatest number of very strong (≥0.70) to strong (≥0.50)
between the loaded march task and CRe fitness; therefore, pooled pooled correlations (8 of 12 task categories). The next most
r values were not calculated for these combinations. notable fitness component was LBs, followed by UBe and UBs,
Table 3 presents the pooled correlation values between each respectively. When all tasks were combined, CRe was the only
physical fitness component and occupational task category. fitness component with a strong weighted average (mean) correl-
Though the number of correlations for each meta-analysis ation; weighted mean correlations were moderate for UBs, LBs,
varied and fixed models were primarily used to address hetero- UBe, LBe and TRs, fair for TRe and weak for FLX (data not
geneity among studies, the vast majority of the pooled r values shown).17
had narrow CIs ( p<0.05). Pooled values also did not appear
sensitive to outlier correlation values (data not shown).17 DISCUSSION
Separate pooled r values were calculated for some specific types Summary of evidence
of physical fitness tests (footnotes in table 3). For example, from This review combined correlation data from existing studies to
the CRe fitness component group, separate pooled r values were examine the evidence of association between core components of
calculated for timed-distance runs (ie, 1.5, 2, or 3.1 miles), physical fitness and common physically demanding occupational
fixed-time distance runs (ie, distance in 12 min) and tests that tasks. Numerous individual studies regarding these relationships
provided maximum oxygen uptake or VO2max. Individual test- exist, but findings have been inconsistent and no prior review has
Hauschild VD, et al. Occup Environ Med 2017;74:144–153. doi:10.1136/oemed-2016-103684 147
 Review

Table 2 Summary of selected studies based on year of publication*

Study Sample Tests groups(# Study quality
Source* Type† country Population type size Gender correlations) Task categories rank‡

Wright et al32 1984 TR USA Military (Army) 272 M, F CRE (3) Lift and lower ++
(221, 51) UB-S (3) (single)
LB-S (3) Lift and lower
TR-S (3) (repeated)
UB-E (3)
TR-E (3)
Robertson and Trent33 TR USA Military (Navy) 45 M, F UB-E (4) Lift and carry ++
1985 (24, 21) TR-E (6) Stretcher carry
UB-S (6)
Beckett and Hodgdon34 TR USA Military (Navy) 102 M, F CRE (4) Lift and carry +++
1988 (64, 38) UB-E (16) Lift and lower (S)
TR-E (5) Move fast
UB-S (6)
LB-S (10)
FLX (4)
Mello et al35 1988 TR USA Military (Army) 28 M LB-E (16) Loaded march ++++
LB-S (16)
Stevenson et al36 1989 J Canada Military 16 M UB-E (2) Lift and lower (S) ++++
UB-S (4) Lift and lower (R)
Schonfeld et al37 1990 J USA Firefighters 20 M CRE (6) Climb +++
Casualty drag
Multiactivity
Singh et al38 1991 TR Canada Military 116 M TR_S (10) Casualty drag ++
UB_S (10) Lift and lower (R)
Dig
Stretcher carry crawl
Arvey et al28 1992 J USA Police 276 M, F CRE (6) Casualty drag +++
UB_S (6) Move fast
LB_S (3) Multiactivity
UB_E (6)
LB_E (3)
TR_E (6)
Stevenson et al39 1992 J Canada Military Personnel 132 M, F UB_S (8) Crawl ++
(99, 33) UB_E (8) Dig
TR_E (8) Lift and carry
Stretcher carry
Myhre et al40 1997 TR USA Firefighters 279 M, F UB_S (3) Multiactivity ++
UB_E (1)
Kraemer et al41 1998 J USA Civilian volunteers 123 F AER (3) Lift and lower (R) +
(military) UB_S (6) Lift and lower (S)
LB_S (6) Loaded march
UB_E (3)
LB_E (3)
Knapik et al42 1999 J USA Military (Army) 11 M, F CRE-tr (1) Stretcher carry +++
(7, 4) UB-S (4)
LB-S (1)
UB-E (1)
TR-E (1)
Williford et al43 1999 J USA Firefighters 91 M CRE (5) Casualty drag ++
UB_S (5) Climb
TR_E (5) Push–pull
UB_E (10) Lift and lower (S)
FLX (5) Multiactivity
Deakin et al44 2000 TR Canada Military 623 M, F CRE (5) Lift and carry ++
(416, UB_S (15) Lift and lower (R)
207) LB_S (5) Dig
TR_S (5) Stretcher carry
UB_E (10) Crawl
LB_E (10)
TR_E (5)
Pandorf et al45 2001 J USA Military (Army) 12 F UB-E (4) Crawl +++
TR-E (3) Multiactivity
Bilzon et al46 2002 J UK Military 93 M, F CRE (4) Stretcher carry +++
(52, 41) UB_S (4)
LB_S (2)
UB_E (4)
Continued

148 Hauschild VD, et al. Occup Environ Med 2017;74:144–153. doi:10.1136/oemed-2016-103684

 Review

Table 2 Continued
Study Sample Tests groups(# Study quality
Source* Type† country Population type size Gender correlations) Task categories rank‡

LB_E (2)
TR_E (2)
Rhea et al47 2004 J USA Firefighters 20 M, F CRE (5) Casualty drag +++
(17, 3) UB_S (10) Climb
LB_S (10) Lift and carry
UB_E (25) Multiactivity
LB_E (5) Push–pull
TR_E (5)
Barnes et al48 2007 J USA Volleyball players 29 F LB_S (1) Move fast ++++
Harman et al49 2008 J USA Civilian volunteers 32 M CRE (4) Casualty drag ++
(military) UB_E (4) Move fast
LB_E (4) Multiactivity
TR_E (4)
LB_S (8)
Williams-Bell et al50 J Canada Firefighters recruits 41 M, F UB_S (2) Multiactivity +
2008 (32, 14) LB_S (1)
UB_E (1)
LB_E (1)
Michaelides et al49 J USA Firefighters 38 M UB_S (1) Multiactivity ++
2008 UB_E (1)
TR_E (1)
LB_S (1)
FLX (1)
Hoffman29 2009 TR USA Peace officers 128 M, F AER (4) Move fast ++
UB_S (4) Multiactivity
LB_S (10)
UB_E (4)
TR_E (4)
FLX (4)
McBride 200951 J USA Football players 17 M LB_S (3) Move fast ++
Phillips 201052 J Australia Firefighters 38 M UB_S (1) Loaded March ++
LB_S (1)
UB_E (3)
LB_E (1)
Aandstad 201153 J USA Military cadets 42 M CRE (1) Move fast ++
Michaelides et al 26 J USA Firefighters 67 M UB_S (21) Casualty drag ++
2011 LB_S (7) Climb
TR_S (7) Lift and carry
UB_E (7) Dig
TR_E (7) Push–pull
FLX (1) Multiactivity
Thebault 201154 J France Military 19 M LB_S (2) Move fast ++
*Study publications order based on year of publication then alphabetically.
†Publication type: J refers to an article in a peer-review journal; TR is a publically available government/military technical report.
‡Quality score descriptor based on consensus of two independent investigators evaluation of maximum of 20 criteria: >15=++++, 15–13=+++, 12–10=++, 9=+; <9 eliminated.
CRe, cardiorespiratory endurance; E, muscle endurance; FLX, flexibility; LB, lower body; S, muscular strength; TR, trunk/core; UB, upper body.

systematically compiled data in a construct evaluation. Though ori- importance of these fitness components. Since this review did not
ginal studies varied in size, population and design, considerable identify notable correlations between job tasks and TRs, TRe or
commonalities warranted this rationalised grouping of the data. FLX, these fitness components appear to be less relevant. This
The resulting pooled correlations between the fitness components finding may be affected by data gaps (such as between FLX and
and common job task categories provide quantitative evidence to various tasks, or TRe and TRs and the load carriage walking task)
support the selection of occupational fitness tests. The results dem- or limited to the singular tests used. For example, though data for
onstrate the important contributions of CRe, LBs and UBe and LBe TRe was fairly abundant for most task categories, timed sit-ups
to overall performance of 12 task categories. Not surprisingly, the have essentially been the only test used. There were no strong cor-
evaluation demonstrated that some individual tasks are more relations between the sit-up test and any of the occupational task
strongly correlated to certain fitness components than others (eg, categories. This finding supports a prior review that questioned the
UBs is more correlated with the single lift and lower task than any value of the sit-up test and its reliability as an occupational test.23
other fitness component). The results also support the concept that In addition to the relationship with task performance, other
no single fitness test represents the overall fitness needed for strenu- factors should be considered to maximise the value of selected
ous jobs. Though much of the research has focused on UBs and tests, ensure safety and minimise costs. For example, fitness tests
UBe, this evaluation indicates that CRe, LBs and LBe deserve par- can also be used as metrics for other constructs of job success, such
ticular attention. Data gaps between CRe and LBe for some tasks as injury risk, attrition, absenteeism and even mortality.4 7 9 23 55
and are not expected to have led to an overestimation of the These constructs can be particularly important for public safety
Hauschild VD, et al. Occup Environ Med 2017;74:144–153. doi:10.1136/oemed-2016-103684 149
 Review

Table 3 Pooled correlation values (r) between health-based component fitness test groups and occupational task categories
Muscular strength Muscular endurance

Task categories Cardio respiratory* Upper body‡ Lower body§ Trunk¶ Upper body** Lower body†† Trunk‡‡ Flexibility†

Lift and lower (single) 0.30 0.75 0.60 {0.57} 0.42 0.56 0.16 0.16
[5] [10] [7] [1] [11] [3] [4] [3]
(0.15, 0.44) (0.66, 0.81) (0.52, 0.67) (0.31, 0.53) (0.48, 0.63) (0.80, 0.24) (0.05, 0.27)
Lift and lower (repeated) 0.60 0.61 0.57 0.56 0.62 {0.55} 0.29 –
[5] [11] [6] [5] [6] [1] [3]
(0.48, 0.70) (0.47, 0.73) (0.37, 0.72) (0.32, 0.73) (0.46, 0.74) (0.05, 0.51)
Lift and carry 0.72 0.43 0.41 0.41 0.50 0.47 0.25 {0.01}
[4] [19] [7] [4] [17] [4] [8] [1]
(0.51, 0.85) (0.34, 0.52) (0.20, 0.59) (0.18, 0.60) (0.37, 0.61) (0.35, 0.57) (−0.08, 0.52)
Casualty drag 0.32 0.38 0.27 0.27 0.33 0.46 0.16 {0.06}
[7] [9] [5] [3] [11] [5] [6] [1]
(0.23, 0.40) (0.24, 0.51) (0.14, 0.39) (0.16, 0.37) (0.19, 0.45) (0.20, 0.66) (0.08, 0.25)
Stretcher carry 0.66 0.65 0.73 {0.67} 0.58 – 0.31 –
[7] [22] [5] [1] [15] [9]
(0.53, 0.76) (0.60, 0.69) (0.62, 0.81) (0.48, 0.66) (0.12, 0.48)
Push–pull 0.09 0.46 0.21 0.42 0.46 0.35 0.20 {0.06}
[2] [7] [5] [2] [9] [4] [4] [1]
(−0.10, 0.28) (0.28, 0.61) (0.10, 0.32) (0.27, 0.55) (0.29, 0.60) (0.21, 0.48) (0.08, 0.32)
Loaded march/walk {0.60} 0.28 0.32 0.01 0.48 0.38 – –
[1] [5] [19] [2] [4] [18]
(0.04, 0.49) (0.25, 0.39) (−0.12, 0.13) (0.25, 0.66) (0.31, 0.45)
Move fast 0.59 0.35 0.58 – 0.47 0.69 0.39 0.08
[8] [5] [13] [9] [2] [7] [3]
(0.51, 0.66) (0.20, 0.49) (0.52, 0.63) (0.35, 0.57) (0.62, 0.75) (0.33, 0.45) (−0.03, 0.18)
Climb 0.55 0.22 −0.09 {0.38} 0.46 0.44 0.43 {0.25}
[4] [5] [3] [1] [8] [3] [3] [1]
(0.42, 0.66) (−0.04, 0.45) (−0.24, 0.08) (0.37, 0.54) (0.26, 0.58) (0.30, 0.54)
Crawl 0.80 0.49 0.65 {0.64} 0.66 – 0.48 –
[2] [5] [2] [1] [5] [5]
(0.72, 0.86) (0.38, 0.59) (0.39, 0.82) (0.50, 0.77) (0.22, 0.68)
Dig 0.62 0.44 0.53 0.47 0.38 {0.15} 0.21 –
[2] [9] [3] [4] [5] [1] [4]
(0.51, 0.71) (0.31, 0.56) (0.37, 0.65) (0.23, 0.65) (0.12, 0.59) (−0.04, 0.44)
Multiactivity 0.52 0.42 0.47 {0.53} 0.46 0.64 0.38 0.08
[9] [15] [9] [1] [17] [10] [10] [5]
(0.47, 0.58) (0.33, 0.51) (0.36, 0.58) (0.38, 0.54) (0.62, 0.70) (0.32, 0.44) (−0.02, 0.18)
[ ]=n correlations from original studies; ()=95% CIs; bold numbers are pooled correlation values; { } only a single study found.
*Results reflect pooled correlations for all fitness tests that provided a measurement for CRe including timed runs, fixed-distance runs and tests that measured or estimated VO2max.
Though data became increasingly limited when attempting to evaluate these three separate types of CRe tests, they were also evaluated separately when data were adequate for tasks
(two or more studies). From these separate analyses, timed runs (1, 1.5, 2 miles) yielded several task for which pooled r>0.50: lift and lower (repeated) r=0.51 (n=3, CI 0.45 to 0.56);
move fast r=0.58 (n=7, CI 0.49 to 0.66); multiactivity r=0.52 (n=6, CI 0.46 to 0.59). Tests providing an estimated or measured VO2max included: lift and lower (repeated) r=0.70 (n=2,
CI 0.59 to 0.79); stretcher carry r=0.71 (n=4, CI 0.57 to 0.81) and crawl r=0.80 (n=2, CI 0.72 to 0.86).
†The sit-and-reach was the only test used.
‡Results reflect pooled correlations for all fitness tests that provided a measurement for UBs. A common test was the dynamometer grip test. Pooled correlations for grip tests with
adequate data yielding pooled r>0.50 included: lift and lower (single) r=0.67 (n=2, CI 0.43 to 0.82); lift and lower (repeated) r=0.59 (n=4, CI 0.27 to 0.80); stretcher carry r=0.61
(n=10, CI 0.52 to 0.70).
§Results reflect pooled correlations for all fitness tests that provided a measurement for LBs including measures of power (single maximum bursts). Common tests included the SBJ and
the VJ. Pooled correlations for SBJ and VJ with adequate data yielding pooled r>0.50 included: lift and lower (single) SBJ r=0.71 (n=2, CI 0.63 to 0.77) and VJ r=0.52 (n=2, CI 0.41 to
0.61); stretcher carry SBJ r=0.83 (n=2, CI 0.77 to 0.87); move fast SBJ r=0.52 (n=2, CI 0.31 to 0.68) and VJ r=0.60 (n=6, CI 0.54 to 0.66); multiactivity VJ r=0.52 (n=4, CI 0.44 to
0.60).
¶Results reflect pooled correlations for all fitness tests that provided a measurement for TRs, no single test type had adequate data to pool correlation data.
**Results reflect pooled correlations for all fitness tests that provided a measurement for UBe including machines and weights, pull-ups and push-up tests. Pooled correlations for
push-up tests with adequate data yielding pooled r>0.50 included: lift and lower (repeated) r=0.57 (n=4, CI 0.31 to 0.75); move fast r=0.52 (n=5, CI 0.45 to 0.59); crawl r=0.58 (n=4,
CI 0.21).
††Results reflect pooled correlations for all fitness tests that provided a measurement for LBe including sprint tests 100–400 m and shuttle tests, though recognised that the shorter
distances completed in <30 s could be considered measures of LB power. Pooled correlations for sprint tests with adequate data yielding pooled r>0.50 included: lift and lower (single)
r=0.63 (n=2, CI 0.54 to 0.71); lift and carry r=0.55 (n=2, CI 0.41 to 0.66); casualty drag r=0.53 (n=3, CI 0.44 to 0.61) and multiactivity r=0.71 (n=5, CI 0.66 to 0.75).
‡‡The sit-up test was the test used in all studies except one.
CRe, cardiorespiratory endurance; SBJ, standing broad jump; TR, trunk/core; UB, upper body; VJ, vertical jump.

and life-saving occupations. Employers should also consider the early 1900s.2 This review supports the inclusion of CRe tests in
feasibility and reliability of fitness tests. Factors that pertain to spe- occupational monitoring since this fitness component was
cific test selection for each fitness component are described below. strongly correlated with the greatest number of the task categor-
ies. This makes physiologic sense given that aerobic metabolism
Cardiorespiratory endurance tests increasingly becomes the dominant source of energy for tasks
Though included in less than half of the selected studies, run lasting more than a few minutes.4 7 23 In fact, because real-life
tests are routinely used by armed forces to test CRe and have occupational tasks often occur over extended periods of time,
been a component of US Army fitness test batteries since the the significance of the CRe component is likely underestimated

150 Hauschild VD, et al. Occup Environ Med 2017;74:144–153. doi:10.1136/oemed-2016-103684

 Review

Table 4 Number of pooled correlations between fitness components and task categories* by correlation strength†
Muscular strength (231 total r values) Muscular endurance (231 total r values)
Strength of pooled Upper body Upper body
correlations (pooled Cardio respiratory (122 r Lower body Trunk (25 (117 r Lower body Trunk (63 Flexibility (15
r value)† (56 total r values) values) (84 r values) r values) values) (51 r values) r values) total r values)

Very strong (r>0.70) 2 1 1 0 0 0 0 0

Strong (0.50≤ r<0.70) 6 2 5 1 4 3 0 0
Moderate (0.40≤r<0.50) 0 5 2 3 6 3 2 0
Fair (0.30≤r<0.40 2 2 1 1 2 2 2 0
Weak (r<0.30) 1 2 3 2 0 0 7 3
Single study or no study; 1 0 0 5 0 4 1 9
data inadequate to pool for
task(s)
No studies/data found for – – – 1 – 3 1 5
task(s)
*For the 12 occupational task categories evaluated in this review, described in table 1.
†Strength of correlation qualified in terms of a priori scale described in methods.17 30 31

by this review of short duration tasks. Studies of military and common UBs test was the dynamometer hand grip test, presum-
firefighter activities provide evidence for this high demand for ably because it is simple, safe and easily administered. A com-
CRe during these occupations.1 3 4 56 An additional value of parison of hand grip test correlations with same-study UBs tests
measuring CRe is that it has consistently been associated with that required activation of larger muscle groups (eg, bench
risk of injury and cardiovascular disease in studies of personnel press) suggests some comparative values.46 47 Though hand grip
in physically demanding jobs.1 4 7 17 55 57 Poor CRe has also tests strongly correlated with one-quarter of the task categories,
been linked to higher military attrition or drop-out.7 23 Though test–retest reliability can be affected by several factors (ie, cali-
CRe appears to be a critical occupational fitness component for bration, use of one or both hands, number of attempts and
both genders, application of CRe tests should consider potential pre-post maximal exertion).59 Of the various tests used to
sex differences in occupational CRe physical ability.9 13 In add- measure LBs, field-expedient surrogate tests of power (eg, VJ
ition, body composition may need to be factored into test appli- and SBJ) were especially common. These jump tests were
cations since anthropomorphic measures (ie, lean body mass) strongly correlated with one-third of the task categories and
took precedence in some of the regression models used in ori- have had good test–retest reliability coefficients ranging from
ginal studies.27 28 58 0.76 to 0.96.23 60 So regression analyses also suggest the poten-
The gold-standard means to monitor CRe is to directly tial relevance of jump tests. For example, the SBJ was included
measure VO2max using calibrated calorimeter equipment and in the predictive equation identified by Bilzon et al46 and
trained test personnel.4 7 23 Since this is impractical for mass Harman et al49 found that only the VJ (aside from anthropo-
routine screening required in many occupational settings, morphic measures such as lean body mass) consistently entered
VO2max is often estimated from calculations of more expedient several predictive regression equations.
tests or represented by simple surrogate measures such as run
time. Run times from timed-distance runs (1.5–26 miles) and
fixed-time (12 min) run tests have been validated against Muscular endurance tests
VO2max testing with correlation coefficients ranging from 0.70 Upper body muscular endurance tests are frequently studied and
to 0.90.23 Their test–retest reliability has also been reported as more commonly included in military and occupational physical
high, with reliability coefficients ranging from 0.82 to 0.98.23 fitness tests than LBe tests.2 6 28 However, LBe strongly corre-
Though the strongest correlations between CRe and tasks identi- lates to different task categories compared to UBe. Since there is
fied in this review were from tests that provided direct and esti- no physiological gold standard, fitness tests used to measure
mated VO2max measurements, the timed-distance (ie, 1–3 miles) muscular endurance have encompassed a variety of repeated lift
and fixed-time (12 min) runs also yielded strong correlations. tests (free weights or machine) and maximum repetition tests.
Push-ups, the most common UBe test, strongly correlated with a
quarter of the task categories. A review of the test–retest reliabil-
Muscular strength tests ity of push-up tests indicates good reliability coefficients ranging
Despite their emphasis in scientific studies, muscular strength tests from 0.76 to 0.83.23 This field-expedient test has particular
have not been commonly used in occupational test batteries.2 12 13 appeal because it does not rely on equipment (eg, calibrated
Practical considerations, such as safety and the need for equipment weights or machines or bars) and is not associated with safety
that must be calibrated or standardised, may limit the use of some risks (eg, dropped weights, falling from bars). As an added
strength tests in routine mass testing. However, the inclusion of value, low scores on push-up tests have also been found to be a
the VJ and hand grip strength tests in the newest Canadian military risk indicator for injury in military and law enforcement training
basic fitness test battery demonstrates potential feasibility.16 populations.6 55 Measuring LBe, such as with sprint tests (100–
Though UBs has been more frequently studied, this review shows 400 m), may be of particular value for occupations that require
that LBs is strongly correlated with more occupational tasks and the casualty drag task, since these were the only fitness tests
therefore may be a more important test consideration. with a strong pooled correlation to this task. Sprint tests require
No physiological gold-standard measurement exists for meas- minimal equipment and have reported good test–retest reliabil-
uring muscular strength, so tests were quite varied. The most ity (reliability coefficients ranging from 0.87 to 0.98).23
Hauschild VD, et al. Occup Environ Med 2017;74:144–153. doi:10.1136/oemed-2016-103684 151
 Review

Limitations 2 Knapik JJ, East WB. History of the United States army physical fitness and physical
There are various limiting factors inherent to the selection and readiness testing. US Army Med Dep J 2014:5–19.
3 Carlson MJ, Jaenen SP. The development of a pre-selection physical fitness training
statistical analyses of original studies in this review. Project time- program for Canadian special operations regiment applicants. J Strength Cond Res
lines constrained the study selection to English language docu- 2012;26(Suppl 2):S2–S14.
ments published between January 1970 and February 2013. 4 Smith DL. Firefighter fitness: improving performance and preventing injuries and
Publication bias may have further limited the acquisition of rele- fatalities. Curr Sports Med Rep 2011;10:167–72.
5 Plat MJ, Frings-Dresen MH, Sluiter JK. A systematic review of job-specific workers’
vant data. Original studies were of varied design, quality, limited
health surveillance activities for fire-fighting, ambulance, police and military
sample sizes and did not provide adequate participant data. personnel. Int Arch Occup Environ Health 2011;84:839–57.
Consequently, gender, age, body size and composition and other 6 Payne W, Harvey J. A framework for the design and development of physical
characteristics such as health status and motivation levels could employment tests and standards. Ergonomics 2010;53:858–71.
not be quantified. The correlation values described the linear 7 National Research Council (NRC). Assessing fitness for military enlistment.
Washington (DC): National Academy Press, 2006:211.
relationship as a metric of association, so potential non-linear or 8 Harbin G, Olson J. Post-offer, pre-placement testing in industry. Am J Ind Med
multivariable relationships are not described. Although studies 2005;47:296–307.
were reviewed and scored to ensure relevance and quality, ori- 9 Courtright SH, McCormick BW, Postlethwaite BE, et al. A meta-analysis of sex
ginal studies describe different confounding variables and differences in physical ability: revised estimates and strategies for reducing
differences in selection contexts. J Appl Psychol 2013;98:623–41.
addressed them in different ways. Therefore, the impact of
10 Sothmann MS, Gebhardt DL, Baker TA, et al. Performance requirements of
potentially significant confounders like body composition (lean physically strenuous occupations: validating minimum standards for muscular
body mass and lean body mass to dead mass ratio)28 58 or strength and endurance. Ergonomics 2004;47:864–75.
gender9 on the calculated pooled r values could not be evalu- 11 Jamnik V, Gumienak R, Gledhill N. Developing legally defensible physiological
ated. It has also been shown that although meta-analyses with employment standards for prominent physically demanding public safety
occupations: a Canadian perspective. Eur J Appl Physiol 2012;113:2447–57.
Fisher’s z value and the Q statistic is a scientifically accepted 12 Department of Defense Joint Technical Coordinating Group. Summary Report:
technique to aggregate comparable data, a limited number of Research Workshop on Physical Fitness Standards and Measurements within Military
studies can underestimate heterogeneity.22 Services; August 31 to 2 September. Herndon (VA): Department of Defense,
1999:3–9.
13 U.S. Army Public Health Command (USAPHC). Public Health Report No.
CONCLUSIONS 12-01-0614. Prepared by Jones BH, Nindl B, Hauret K, et al. Development of a
Occupational health practitioners charged with ensuring a phys- New Army Standardized Physical Readiness Test: 2012 through 2013. Aberdeen
ically capable workforce can use physical fitness tests as indica- Proving Ground (MD): USAPHC, 2015:1–8.
14 General Accounting Office. Gender issues: improved guidance and oversight are
tors of employees’ physical health status. The most beneficial needed to ensure validity and equity of fitness standards. Washington (DC):
tests are those that are strongly related to the common tasks crit- Subcommittee on Readiness, Committee on Armed Services, 1998:4–6, 27, 29.
ical to overall job performance. For example, this evaluation of 15 Epstein Y, Yanovich R, Moran DS, et al. Physiological employment standards IV:
12 task categories collectively relevant to military and other integration of women in combat units physiological and medical considerations.
Eur J Appl Physiol 2013;113:2673–90.
physical occupations demonstrated physically demanding jobs 16 Reilly T. Canada’s physical fitness standard for the land force: a global comparison.
should consider CRe, LBs and UBe tests. To maximise the value Can Army 2010;13:59–69.
of testing, selection of specific tests should also consider the 17 US Army Public Health Command. Public Health Report No. 12-02-0614.
association of tests to other indicators of job success, such as Correlations Between Physical Fitness Tests and Performance of Military Tasks:
injury risk and attrition, as well as test reliability, feasibility and A Systematic Review and Meta-Analyses. Prepared by Hauschild VD, DeGroot D,
Hall S, et al. Aberdeen Proving Ground (MD): USAPHC, 2014. Available from DTIC:
cost. Valid and reliable field-expedient tests to consider include http://www.dtic.mil/docs/citations/ADA607688
timed runs (1.5–3 miles) for CRe, jump tests (VJ or SBJ) for 18 Equal Employment Opportunity Commission. Uniform guidelines on employee
LBs, push-ups for UBe and sprints for LBe. Grip tests for meas- selection procedures (1978 as revised 2003). Fed Reg 2003;43:38289–315.
uring UBs may be useful if reliability can be established. Though 19 Caspersen CJ, Powell KE, Christenson GM. Physical activity and physical fitness:
definitions and distinctions for health-research. Public Health Rep
frequently used, sit-up and sit-and-reach tests may not provide 1985;100:126–31.
relevant occupational health metrics. Future studies should 20 Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for systematic reviews
evaluate the impacts of physiological characteristics such as lean and meta-analyses: the PRISMA statement. PLoS Med 2009;6:e1000097.
body mass and gender on the application of these tests and the 21 Garg AX, Hackman D, Tonelli M. Systematic reviews and meta-analysis: when one
establishment of appropriately scaled fitness test standards. study is just not enough. Clin J Am Soc Nephrol 2008;3:253–60.
22 Von Elm E, Altman DG, Egger M, et al., Strobe Initiative. The Strengthening the
Acknowledgements The authors acknowledge editorial contributions of Ms Mimi Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines
Eng, statistical consult of Angela Mound and Bonnie Taylor, and reviews by Jessica for reporting observational studies. Ann Intern Med 2007;147:573–7.
Sharkey, MPH, and Anna Schuh, PhD. 23 US Army Center for Health Promotion and Preventive Medicine (USACHPPM).
Technical Report No 12-HF-01Q9D-04. The case for pre-enlistment physical fitness
Contributors VDH, DWD, TLG and KGH conceived the study. VDH and DWD testing: research and recommendations. Prepared by Knapik JJ, Jones BH, Sharp
conducted the literature search, assessed quality and selected studies and extracted MA, et al. Aberdeen Proving Ground (MD): USACHPPM, 2004: 100 p. Available
data. TLG, KGH and BHJ reviewed data and advised on meta-analytical grouping. from DTIC. http://www.dtic.mil/get-tr-doc/pdf?AD=ADA426848
SMH and KDD gave methodological advice and conducted statistical analyses. VDH 24 Department of the Army. Soldier’s manual of common tasks warrior skills level 1.
wrote the draft of the paper. All authors contributed to writing, reviewing and Soldier Training Publication (STP) 21-1. Chapter 3. Washington (DC): Headquarters
revising paper and approved final manuscript. All authors are employed by the Army Department of the Army, 2011:3–233.
Public Health Center (Provisional), Aberdeen Proving Ground, MD. 25 North Atlantic Treaty Organization. Optimizing operational physical fitness. North
Disclaimer The views expressed are those of the authors and do not necessarily Atlantic Treaty Organization Research and Technology Organization, 2009. http://
reflect official policy of the Department of Defense, Department of the Army, US www.dtic.mil/dtic/tr/fulltext/u2/a502544.pdf (accessed online October 2016).
Army Medical Department or the USA. 26 Michaelides MA, Parpa KM, Henry LJ, et al. Assessment of physical fitness aspects
and their relationship to firefighters’ job abilities. J Strength Cond Res
Competing interests None declared. 2011;25:956–65.
Provenance and peer review Not commissioned; externally peer reviewed. 27 Hoffman and Associates. Physical Readiness Validation for the Nevada POST
Category III, Final Report. 6 March 2009. P A1–17; p E4.
28 Arvey RD, Landon TE, Nutting SM. Development of physical ability tests for police
REFERENCES officers: a construct validation approach. J Appl Psychol 1992;77:996–1009.
1 Storer TW, Dolezal BA, Abrazado ML, et al. Firefighter health and fitness 29 Cooper H. Research synthesis and meta-analysis: a step by step approach. 4th edn.
assessment: a call to action. J Strength Cond Res 2014;28:661–71. Washington (DC): Duke University, SAGE Publications, 2009:1–22, 145–197.

152 Hauschild VD, et al. Occup Environ Med 2017;74:144–153. doi:10.1136/oemed-2016-103684

 Review
30 Mood DR, Morrow JR. Introduction to statistics in human performance. Scottsdale 44 Deakin JM, Pelot R, Smith JT, et al. Development and validation of Canadian forces
(AZ): Holcomb Hathaway Publishers, 2015:92–9. Minimum Physical Fitness Standard (MPFS 2000). Kingston (ON): Ergonomics
31 Hemphill JF. Interpreting magnitudes of correlation coefficients. Am Psychol Research Group, Queen’s University, 2000. Chapters 3 and 4, 9–36; Chapter 5,
2003;28:78–80. 45–9.
32 Wright JE, Sharp DS, Vogel JA, et al. Assessment of muscle strength and prediction 45 Pandorf CE, Harman EA, Frykman PN, et al. Correlates of load carriage and
of lifting capacity in US army personnel. Natick (MA): US Army Research Institute of obstacle course performance among women. Work 2002;18:179–89.
Environmental Medicine (available from DTIC), 1984:46. 46 Bilzon JL, Scarpello EG, Bilzon E, et al. Generic task-related occupational
33 Robertson DW, Trent T. Documentation of muscularly demanding job tasks and requirements for Royal Naval personnel. Occup Med (Lond) 2002;52:503–10.
validation of an occupational strength test battery (STB). Natick (MA): US Army 47 Rhea MR, Alvar BA, Gray R. Physical fitness and job performance of firefighters.
Research Institute of Environmental Medicine (available in DTIC), 1985:101. J Strength Cond Res 2004;18:348–52.
34 Beckett MB, Hodgdon JA. Lifting and carrying capacities relative to physical fitness 48 Barnes JL, Schilling BK, Falvo MJ, et al. Relationship of jumping and agility
measures. San Diego (CA): Naval Health Research Center, 1987:49. performance in female volleyball athletes. J Strength Cond Res 2007;21:1192–6.
35 Mello RP, Damokosj AI, Reynolds KL, et al. The physiological determinants of load 49 Harman EA, Gutekunst DJ, Frykman PN, et al. Prediction of simulated battlefield
bearing performance at different march distances. Technical report T15–88. DTIC. physical performance from field-expedient tests. Mil Med 2008;173:36–41.
Natick (MA): Army Research Institute of Environmental Medicine, April 1988:1–25. 50 Williams-Bell FM, Villar R, Sharratt MT, et al. Physiological demands of the
http://www.dtic.mil/docs/citations/ADA197733 firefighter Candidate Physical Ability Test. Med Sci Sports Exerc 2009;41:653–62.
36 Stevenson JM, Andrew GM, Bryant JT, et al. Isoinertial tests to predict lifting 51 McBride JM, Blow D, Kirby TJ, et al. Relationship between maximal squat strength
performance. Ergonomics 1989;32:157–66. and five, ten, and forty yard sprint times. J Strength Cond Res 2009;23:1633–6.
37 Schonfeld BR, Doerr DF, Convertino VA. An occupational performance test 52 Phillips M, Petersen A, Abbiss CR, et al. Pack hike test finishing time for Australian
validation program for fire fighters at the Kennedy Space Center. J Occup Med firefighters: pass rates and correlates of performance. Appl Ergon 2011;42:411–8.
1990;32:638–43. 53 Aandstad A, Holme I, Berntsen S, et al. Validity and reliability of the 20 meter
38 Singh M, Lee SW, Wheeler G, et al. Task related physical fitness and performance shuttle run test in military personnel. Mil Med 2011;176:513–8.
standards for the Canadian army. Canada: University of Alberta, 1991:569. 54 Thebault N, Leger LA, Passelergue P. Repeated-sprint ability and aerobic fitness.
39 Stevenson JM, Bryant JT, Andrew GM, et al. Development of physical fitness J Strength Cond Res 2011;25:2857–65.
standards for Canadian Armed Forces younger personnel. Can J Sports Sci 55 Jones BH, Hauschild VD. Physical training, fitness, and injuries: lessons learned from
1992;17:214–21. military studies. J Strength Cond Res 2015;29(Suppl 11):S57–64.
40 Myhre LG, Tucker DM, Bauer DH, et al. Relationship between selected measures of 56 Tharion WJ, Lieberman HR, Montain SJ, et al. Energy requirements of military
physical fitness and performance of a simulated fire fighting emergency Task. personnel. Appetite 2005;44:47–65.
Brooks Air Force Base (TX): Armstrong Lab (available in DTIC), 1997:1–26. 57 Blacker SD, Wilkinson DM, Bilzon JL, et al. Risk factors for training injuries among
41 Kraemer WJ, Nindl BS, Gotshalk LA, et al. Prediction of military relevant British Army recruits. Mil Med 2008;173:278–86.
occupational tasks in women from physical performance components. Advances in 58 Vanderburgh PM. Occupational relevance and body mass bias in military physical
Occupational Ergonomics and Safety. Washington (DC): IOS Press, 1998:719–22. fitness tests. Med Sci Sports Exerc 2008;40:1538–45.
42 Knapik JJ, Harper W, Crowell HP. Physiological factors in stretcher carriage 59 Reuter SE, Massy-Westropp N, Evans AM. Reliability and validity of indices of
performance. Eur J Appl Physiol Occ Physiol 1999;79:409–13. hand-grip strength and endurance. Austr Occupational Ther 2010;58:82–7.
43 Williford HN, Duey WJ, Olson MS, et al. Relationship between firefighting 60 Vizcaya FJ, Viana O, Del Olmo MF, et al. Could the deep squat jump predict
suppression tasks and physical fitness. Ergonomics. 1999;42:1179–86. weightlifting performance? J Strength Cond Res 2009;23:729–34.

Hauschild VD, et al. Occup Environ Med 2017;74:144–153. doi:10.1136/oemed-2016-103684 153