Table 1 Findings from 12 studies of body weight and non-fatal traumatic injury
ReferenceStudy aim and research designStudy populationDefinition of body weight*Definition of occupational injuryAssociation between body weight and obesity†
Heineman et al12To identify risk factors for fireground injuries among professional firefighters; case–control study1200 fire fighters in the Boston areaSelf-reported height and weight from a telephone interview; Quetlet’s index: weight (kg)/height (cm2)Lost work time injury (due to burns and falls) between 1/1/86 and 31/12/86OR for BMI and fireground fall-related injury = 3.3 (0.5 to 22.8) when matched on position, size, and type of fire‡
Gauchard et al13To determine individual employee characteristics associated with workplace injuries due to imbalance; case­–control study427 male employees employed for 3 or more years by the French National Society of RailwayWeight and height measured at physical exams; BMI (only assessed risk for BMI ⩾30 kg/m2)At least one non-fatal occupational accident with sick leave between 1/3/99 and 29/2/00. Three categories of injuries: slip, trip, collision with a moving vehicle, or bad landing on the floor (when getting out of a vehicle)OR for BMI ⩾30 and: slips (1.13 (0.53 to 2.44)); trips (1.33 (0.76 to 2.37)); bad landing (1.10 (0.42 to 2.89)). Adjusted OR for BMI ⩾30 and sick leave of ⩾8 days −2.07 (1.03 to 4.16), p<0.05
Chau et al14To assess the relationships of job, age, and life conditions with the causes and severity of occupational injuries; case–control study880 male construction workersPhysician measured height and weight; BMI dichotomized into ⩽25 and ⩾26At least one occupational injury with subsequent sick leave between 1/1/95 and 31/12/96, and seen by an occupational physician; in workers with more than one injury, only the last injury was retainedBivariate association between BMI and injury, p<0.10, and between BMI and hospitalization, p<0.10; difference in distribution of BMI for falls to the same level and falls to the lower level both had p = 0.06; OR for BMI ⩾26 and falls on same level = 1.85 (1.15 to 3.00), p<0.01, and falls to a lower level = 1.68 (1.19 to 2.37), p<0.01; stepwise forward procedure OR for BMI ⩾26 and falls on same level = 2.04 (1.30 to 3.21), p<0.05, and falls to a lower level = 1.66 (1.20 to 2.29), p<0.05; OR for BMI ⩾26 and sick leave >60 days = 1.32 (0.96 to 1.82)‡
Myers et al15To examine the effects of anthropometric, ergonomic, and psychosocial factors on the risk of low back injury; case–control study200 injured case patients and 400 controls of Baltimore City Municipal workers: education, public works, recreation and parks, transportationHeight and weight measured by interviewers within 10 days of an injury; BMIWorkers with incident reported back injury 1/3/90 – 1/4/91 who had been assigned restricted activity or lost work timeMean BMI (27.9 and 27.0) different between cases and controls (p<0.05); OR for case and 1st control: 1.47 (0.99 to 2.19); OR for case and 2nd control: 1.67 (1.11 to 2.52); OR for case and both controls: 1.54 (1.09 to 2.16)
Bigos et al16To evaluate the impact of low back injuries in manufacturing employees; cross-sectional study31 200 hourly employees at the Boeing CompanyHeight and weight from claims4645 injury claims (including 900 back injury) filed between 1/7/79 and 28/2/81No significant differences in the height and weight, by sex, of back versus non-back injury claimants and of high- versus low-cost back injury claimants; data not shown
Kraus et al17To determine the effect of back belt use on the incidence of back injury in home attendants; cluster-randomized trial9 home attendant agencies employing 12 772 home attendantsHeight and weight obtained at baseline (May–August 1997); BMINumber of injury claims filed by each person over a 28-month period, from 1/6/97 to 30/9/99Crude rate ratio for risk of injury and: underweight/normal (1.03 (0.57 to 1.86)); marginally overweight (1.39 (0.55 to 3.52)); overweight (1.29 (0.74 to 2.25)); severely overweight (1.41 (0.74 to 2.66)). Adjusted rate ratio for increasing BMI and injury (1.21 (0.98 to 1.50)). Adjusted rate ratio for risk of injury and: underweight/normal (1.22 (0.66 to 2.26)); marginally overweight/overweight (1.36 (0.55 to 1.06)); severely overweight (1.22 (0.73 to 2.02))
Ryden et al18To examine the relationship between low back injury and individual and workplace risk factors. Case–control study84 cases of low back injuries and 168 controls of hospital and health center employeesMedical records. Used the 1983 Metropolitan Height and Weight Table to classify employees as normal, overweight, or obese. Up to 30 lb = overweight for height (overweight); more than 30 lb = obeseLow back injury occurring during the work day, regardless of time lost or workers’ compensation claimTwo-sample t test found no difference between cases and controls in terms of mean height and weight (p<0.4). Risk of injury and: underweight (1.47 (0.70 to 3.10)); normal weight (0.93 (0.30 to 2.82)); overweight (0.90 (0.25 to 3.19)); overweight/obese (0.91 (0.27 to 3.05))
Brown and Thomas19To evaluate a number of risk factors for back injury among medical center employees. Descriptive study233 medical centre employeesBMI, obtained from medical records. Height and weight taken when a claim was reportedWorkers compensation claims that resulted in lost work time71% of injured workers had a BMI>25; of these, 31% were obese and 11% were extremely obese‡. 67% of workers with a strain injury had a BMI>25‡. Association between increasing BMI and loss of productivity cost (Pearson χ2 = 24.13; df = 9, p = 0.004). Association between increasing BMI and injury site (Pearson χ2 = 36.82; df = 9, p = 0.098)
Wohl et al20To assess the influence of family-related factors outside of work on a woman’s risk of injury. Case–control study1400 female manufacturing workers at a large aircraft manufacturing company, 1/1/89–31/12/89. 1400/4200 (33%) female manufacturing workersHeight and weight from company health records. BMIAny woman who reported an acute traumatic work injury during calendar year 1989Crude odds ratio for injury and: underweight (0.9 (0.3 to 1.7)); normal (referent); overweight (1.6 (1.0 to 2.6)); obese (1.5 (0.9 to 2.6)).χ2 test for trend, p = 0.03. Adjusted odds ratio for injury and:‡ underweight (0.8 (0.4 to 2.03)); normal (referent); overweight (1.5 (0.9 to 2.6)); obese (1.7 (1.0 to 3.0))
Low et al21To assess the incidence, diversity, and personal risk factors for farm injuries. Cross-sectional study919 surveyed farms in New South WalesSelf-reported height and weight from a telephone interview. BMIInjury that resulted in suspension of usual farm activities for 1 day, restriction for 5 days, or professional medical careBMI and injury parameter estimate (on the logit scale) = 0.189, SE = 0.0615.‡ BMI by sleep interaction significant with injury‡
Bhattacherjee et al22To assess the relationship of individual level risk factors on occupational injuries. Population based cross- sectional study2562 employed individuals randomly selected from the populationSelf-reported height and weight from a survey administered in 1996. BMI, three groups: ⩽19, 20–24, ⩾25 kg/m2Presence of at least one occupational injury in the 2-year period before the survey (1994–96). Occupational injury had to result in sick leave from work and lead to compensationDistribution of incidence of injuries by BMI: ⩽19 = 2.6%; 20–24 = 4.3%; ⩾25 = 5.4%; p<0.01. Interaction of BMI and injury: partial likelihood ratio χ2 = – 2.7, df = 2, not statistically significant‡
Froom et al23To evaluate body weight as an independent risk factor for injuries in industrial workers. Case–control study4306 men, 21 industrial plants in Israel: electronics, textiles, and furniture, food, and tire factories, and iron production plantsWeight and height measured at physical exams that occurred 1985–1987. BMI, Categorized as: <22.8, 22.8–25.3, 25.4–27.9, >27.9On-site injuries with at least one day’s loss of work 1986–1987. Injury categories: falls, struck by moving objects, caught in machinery, road accidents, and other. 870 injuries (707/3801 male workers)OR for BMI and single injury versus no injury = 1.01 (0.92 to 1.09), p = 0.89. OR for BMI and two or more injuries versus no injuries = 1.25 (1.03 to 1.5), p = 0.02
  • *BMI, body mass index.

  • †OR, odds ratio; estimates reflect OR for BMI and injury (lower bound CI, upper bound CI).

  • ‡No p value presented.