Page 101 - JSOM Spring 2023

TABLE 1 Characteristics of Selected Studies
Quality Score
Height & Injury (points/% of
Study Participants Year(s) Weight Recording Recording Injury Case Definition available points a
Billings et al., 1,198 Air Force cadets 2002 Administrative Medical records Any physical damage to 12/86
2004 65 in 6-week officer basic records body
training
Taanila et al., 869 Finnish male 2006–2007 Measured Medical records MSK disorders including 12/86
2010 48 conscripts followed for overuse and acute physical
6 months damage to the body
Knapik et al., 475,746 men and 1997–2007 Measured Medical records Stress fractures 12/86
2012 57 107,906 women in US
Army BCT followed
for 10 weeks
Knapik et al., 1,627 men in 14-week 2010–2011 Self-reported Medical records Any physical damage to 12/86
2013 66 US Army Combat body
Engineer BCT
Taanila et al., 1272 Finnish male 2006–2008 Measured Medical records Acute: physical damage to 11/78
2015 58 conscripts followed for the body involving sudden
6 months onset and known trauma
Overuse: physical damage to
body with gradual onset and
without known trauma
Hruby et al., 736,608 US Army 2001–2011 Measured Medical records MSK injuries & connective 12/86
2016 67 personnel followed for tissue disorders
first incident injury
during 2001-2011
Sulsky et al., 139,020 men and 2002–2007 Measured Medical records Defined only as “injury” 10/71
2018 55 54,784 women in US
Army BCT
Hollander et al., 72,934 IDF conscripts 2003–2012 Measured Medical records Overuse injuries: “. . . caused 12/86
2020 49 training for 7 months by repetitive, submaximal
(infantry and armor) or loading of the MSK system
3 weeks (intelligence) where there is insufficient
rest for recovery and
adaptation of tissues”
Schwartz et al., 84,348 IDF infantry 2007–2018 Medical records Medical records Knee meniscus damage 12/86
2020 68 soldiers followed for diagnosed by orthopedist
10 years on MRI
a Scoring based on National Institute of Health Study Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies
Abbreviations: US=United States, BCT=Basic Combat Training, MSK=musculoskeletal, MRI=magnetic resonance imaging, IDF=Israeli Defense
Force

TABLE 2 Summary of Meta-Analyses with Summary Risk Ratios, Heterogeneity, and Publication Bias Statistics
Trim & Fill Procedure
Summary Risk Ratios Q-Statistic Egger Test Trimmed & Imputed Risk Ratio
Comparison (95%CI) p-value I (%) p-value Imputed Studies (n) (95%CI)
2
Underweight/Healthy Weight 1.17 <0.01 97 0.21 2 1.21
(1.07–1.28) (1.11–1.33)
Overweight/Healthy Weight 1.03 <0.01 85 0.33 1 1.03
(1.01–1.06) (1.01–1.06)
Obese/Healthy Weight 1.15 <0.01 87 <0.01 4 1.12
(1.11–1.20) (1.08–1.17)

number of data points on both sides and the Egger Test sug- Fill Procedure suggested four studies were missing, and imput-
gested no publication bias. The Trim and Fill Procedure sug- ing these studies slightly decreased the risk in the obese group.
gested one study was missing but imputing that data point had
no effect on the RR or 95%CI.
Discussion
To compare injury risk between healthy weight and obese This meta-analysis indicated that compared with healthy
SMs, Figure 6 shows the forest plot, Figure 7 the funnel plot, weight individuals, individuals who were underweight, over-
and fourth line of Table 2 provides summary statistics. Obese weight, or obese were at higher risk of injury. Given the wide
SMs were at higher injury risk than normal weight with con- range of RRs among individual studies, considerable hetero-
siderable heterogeneity in the estimate. Examination of the geneity was evident in the pooled estimates. All three analyses
funnel plot indicated more data points to the right of the mean had some indication of publication bias, but the Trim and Fill
and the Egger Test suggested publication bias. The Trim and Procedure had only minor effects (if any) on the RRs.

Body Mass Index and Injuries | 99