Table 2

Data extraction of the included prospective and retrospective cohort studies

StudyStudy typeDuration of follow-up (weeks)Participants (total group and cases of AT)Sex (% male)Age, mean±SD (years)Location injuryRisk factors (risk ratio, OR and HR)Quality score (points)
Barge-Caballero et al 24 RCNR149 (14); heart transplant patients who were prescribed quinolones.77.9%58.8±10.6AT (not specified midportion or insertional).
  • A creatinine clearance <60 mL/min was associated with AT compared with a creatinine clearance ≥60 mL/min (OR 6.14; 95% CI 1.23 to 30.64; p=0.03).

  • Increased time (in years) between heart transplantation and initiation of quinolone treatment for infectious disease was associated with AT (OR 1.39; 95% CI 1.11 to 1.74; p=0.005).

  • No associations were found for age, sex, levofloxacin use and daily prednisone dose (mg).

Hein et al 20 PC52269 (10); recreational runners.NRNRAT (not specified midportion or insertional).
  • No statistical analyses were performed.

Kaufman et al 21 PC104449 (30); Navy Sea, Air and Land (SEAL) candidates.100%22.5±2.5AT (not specified midportion or insertional).
  • A tight ankle dorsiflexion with knee extended (<11.5°) was associated with AT compared with a normal dorsiflexion (11.5–15.0°) (RR 3.57; 95% CI 1.01 to 12.68; p<0.05).

  • No associations were found for hindfoot inversion, hindfoot eversion, static arch index of the foot, dynamic arch index of the foot and dorsiflexion of the ankle with the knee bent.

Mahieu et al 16 PC669 (10); officer cadets.100.0%18.4±1.3Midportion AT.
  • Isokinetic plantar flexion strength at 30°/s was decreased in patients who developed AT for both the right and the left leg and at 120°/s for the right leg (p=0.042, p=0.036 and p=0.029, respectively). Plantar flexion strength was measured using the Cybex Norm dynamometer, which measures strength at constant velocity.

  • No associations were found for weight, BMI, length, physical activity level, Achilles tendon stiffness, isokinetic plantar flexion strength at 120°/s for the left leg, explosive gastrocnemius-soleus muscle strength (standing broad jump test) and passive and active ankle joint range of motion outcomes.

Milgrom et al 22 PC141405 (95); infantry recruits.100.0%18.7±7Midportion AT.
  • An increase in AT was seen when training in the winter season compared with summer training (p=0.001).

  • No differences were found in height, weight, BMI, external rotation of the hip, tibial intercondylar distance, arch type, physical fitness performance (2 km run and maximum number of chin-ups and sit-ups done) and shoe type.

Owens et al 17 PC5280 106 (450); military service members.70.3%NRAT (not specified midportion or insertional).
  • Being overweight and obesity were associated with AT compared with underweight or normal weight (AOR 1.29, 95% CI 1.04 to 1.59 and AOR 1.59, 95% CI 1.16 to 2.17, respectively)

  • A prior lower limb tendinopathy or fracture was associated with AT (AOR 3.87, 95% CI 3.16 to 4.75).

  • Moderate alcohol use (7–13 units per week for men, 4–6 units per week for women) was associated with AT compared with no alcohol use (AOR 1.33, 95% CI 1.00 to 1.76).

  • A birth year of 1980 and later was associated with a decreased risk for AT compared with a birth year before 1960 (AOR 0.62, 95% CI 0.38 to 1.00).

  • No associations were found for sex, ethnicity, smoking status and heavy alcohol use (14+ units per week for men, 7+ units per week for women).

Rabin et al 18 PC2670 (5); military recruits.100.0%19.6±1.0Midportion AT.
  • Every 1° increase in ankle dorsiflexion with the knee bent was associated with a decreased risk for AT (OR 0.77; 95% CI 0.59 to 0.94).

  • No associations were found for BMI and lower extremity quality of movement.

Van Ginckel et al 19 PC10129 (10); novice runners.14.7%39±10Midportion AT.
  • An increased total anterior displacement of the Y-component of the centre of force was associated with a decreased risk for AT (OR 0.919; 95% CI 0.859 to 0.984; p=0.015).

  • A more medial directed force distribution underneath the forefoot at forefoot flat was associated with a decreased risk for AT (OR 0.000; 95% CI 0.000 to 0.158; p=0.016).

  • No associations were found for age, height, weight, BMI or physical activity score.

Van der Linden et al 25 RCNR10 800 (8); patients using fluoroquinolones (index group) or amoxicillin, trimethoprim, cotrimoxazole or nitrofurantoin (reference group).29.8%46.3 (SD NR)AT (not specified midportion or insertional).
  • The use of ofloxacin was associated with AT compared with the reference group (AOR 10.1; 95% CI 2.20 to 46.04).

  • No associations were found for fluoroquinolones as a group, ciprofloxacin use and norfloxacin use compared with the reference group.

Wezenbeek et al 23 PC104300 (27); first-year students.47%18.0±0.8Midportion AT.
  • Female sex was associated with AT (HR 2.82, 95% CI 1.16 to 6.87).

  • Height and body weight were increased in patients with AT (p=0.028 and p=0.015).

  • No association was found for a pronated foot posture.

  • No differences were found for BMI, rating of perceived exertion, hours of sports participation and leg dominance.

  • AOR, adjusted OR; AT, Achilles tendinopathy; BMI, body mass index; NR, not reported; PC, prospective cohort study; RC, retrospective cohort study; RR, risk ratio.