Epidemiology and Prediction of Injuries in the National Football League (NFL) Open Access

Binney, Zachary (Spring 2018)

Permanent URL: https://etd.library.emory.edu/concern/etds/zw12z530q?locale=en


American football is the most popular sport in the U.S.; over a million high school students play it every year. It also has high injury rates. In the National Football League (NFL) the reported in-game injury rate is at least three times that of any other major U.S. professional sport. Yet the literature on NFL injuries has major gaps, including a lack of descriptive epidemiology, few assessments of rule changes designed to reduce injuries, and an inability to predict which players are most likely to get injured. This dissertation used a prospectively-collected database of reported NFL injuries from 2007-2016 to address these gaps.


In Aim 1 we calculated NFL injury rates and investigated their associations with an array of potential risk factors. Over two-thirds of player-seasons resulted in an injury. Injury rates were higher among older players, those with longer injury histories, heavier players, and on certain types of artificial turf. While injury rates did rise over time, it appeared a majority of the rise was due to better reporting.


In Aim 2 we assessed whether the NFL’s 2011 collective bargaining agreement (CBA), which restricted practice times, affected the number of NFL injuries. Some had expressed concerns that the restrictions would worsen player conditioning and increase injury risk. To investigate this we split injuries by those likely and unlikely to be affected by poor conditioning and observed their changes pre- and post-CBA. After accounting for a pre-CBA increase in injuries, we found little evidence for an increase in injuries following the 2011 CBA.


In Aim 3 we used publicly-available data on established NFL injury risk factors to develop prediction models for the 1-seaon risk of missing one or more games due to any injury or a lower extremity (LE) muscle injury. In all models discrimination and calibration were poor, and they were unable to predict NFL injuries in a way that would be useful for NFL players or teams.


This work identified some new possible risk factors (specific brands of artificial turf) and suggested some common explanations for increased injuries such as short rest and practice restrictions may be overstated.

Table of Contents

Table of Contents

Chapter 1: Introduction and Literature Review 1

I. Introduction to Football 1

II. Who Plays Football 3

III. Football Injuries – Descriptive Epidemiology and Injury Surveillance 4

A. General Football Injury Studies – High School and Younger 4

B. General Football Injury Studies – College 7

1. NCAA ISS Data 7

2. Big 10 Conference ISS (B10 ISS) Data 11

3. Other College Data 13

4. Synthesis 14

C. General Football Injury Studies – NFL 14

1. NFL Injury Surveillance System (NFL ISS) 14

2. NFL Injury Report Data 16

3. Single Team Data 19

4. Review of NFL Injury Studies 22

5. Synthesis 22

D. Specific Football Injury Studies – NFL 23

E. Football Injury Surveillance Systems 35

IV. Football Injuries – Analytic Epidemiology and Factors Influencing/Predicting Injuries 37

A. Age 37

B. Prior Injuries 38

C. Anthropometrics 42

1. Athlete Size (BMI) 42

2. Other Anthropometric Measures 47

D. Games vs. Practices 48

1. Games vs. Practices Overall 48

2. Practice Intensity 49

3. Seasonal Variation (Spring vs. Fall Practices) 51

E. Play Type 53

F. Position 54

G. Rest 59

H. Game-Day Conditions 60

1. Altitude 60

2. Precipitation 61

3. Temperature 62

4. Turf 64

I. Calendar Time (Years/Seasons) 66

J. Within-Season Time (Weeks) 67

K. Coach Experience 69

L. Travel 71

M. Pre-season Conditioning 71

N. Synthesis 72

V. Injury Risks and Rates in Football vs. Other Sports 73

Chapter 2: Specific Aims, Data Sources, and Methods Overview 76

I. Overview and Specific Aims 76

II. Cross-Aim Methods 78

A. Data Sources 78

1. Study Design and Data Collection Methods 78

2. Study Population 79

B. Limitations and Methodologic Challenges 80

1. Limitations Inherent to Injury Report Data 80

2. Limitations with the FO Injury Data 84

3. Methodologic Challenge: Categorizing Injuries 85

4. Methodologic Challenge: Calculating Injury Risks and Rates 88

5. Methodologic Challenge: Multiple Injuries 90

6. Methodologic Challenge: Recurrent Injuries 91

7. Methodologic Challenge: Estimating Injury Severity and Return-to-Play 91

8. Methodologic Challenge: Survivor Bias in Aging and Career Snapcount Analyses 92

9. Methodologic Challenge: Categorizing Special Teams Players 92

Chapter 3: Specific Aim 1. Descriptive Epidemiology of NFL Injuries, 2007-2015. 93

I. Abstract 93

II. Introduction 94

III. Methods 95

A. Data Sources 95

B. Injury Counts and Measures of Injury Frequency 96

C. Injury Risk Factors 98

D. Statistical Analysis 98

IV. Results 100

V. Discussion 125

VI. Conclusions 128

VII. Appendix 130

Chapter 4: Specific Aim 2. 2011 CBA’s Effects on Injuries. 154

I. Abstract 154

II. Introduction 155

III. Methods 156

IV. Results 159

V. Discussion 168

VI. Conclusions 170

VII. Appendix 171

Chapter 5: Specific Aim 3. Predicting 1-Year Injury Risk in the NFL. 178

I. Abstract 178

II. Introduction 179

III. Methods 180

IV. Results 183

V. Discussion 192

VI. Conclusions 195

VII. Appendix 196

Chapter 6: Public Health Impact/Significant Contribution to the Field 203

References 207

About this Dissertation

Rights statement
  • Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.
  • English
Research Field
Committee Chair / Thesis Advisor
Committee Members
Last modified

Primary PDF

Supplemental Files