Page 121 - JSOM Summer 2018

workers, 30,31 railway workers, construction workers, 33,34 coal Cessation of Exposure
32
miners, and community-based samples of workers. 35,36 These The seventh criterion to assist in determining a causal effect is
studies involved a several countries, including France, 32,33,35 that once an individual no longer smokes, a decrease in injury
Japan, Korea, India, 27,37 Taiwan, the United Kingdom, risk should be expected. This was discussed in part 1 of this
30
36
31
25
and the United States. 26,28,29,34 Study designs have included series. Most data directly addressing this question come from
1
prospective cohort, 28,29,41 retrospective cohort, 25–27,30,31,34,35 and studies on bone health and wound healing. Smoking cessation
case-control. 32,33 No randomized prospective cohort studies reduced the loss of bone mineral density over 1–3 years. 74,75 Sev-
were found, but studies of this type (in which smoking and eral meta-analyses of cross-sectional studies have found that the
nonsmoking would be randomized among participants) can- risk of fractures in former smokers was intermediate between
not be conducted because of ethical concerns related to the that of smokers and nonsmokers, suggesting that the effects of
well-documented health effects of smoking. Nonetheless, the smoking on fracture risk are at least somewhat reversible. 42,43,48
relationship between smoking and injuries has been seen in Five to 30 years after smoking cessation, former smokers had a
many types of workers in several countries, and using a variety considerably reduced risk of fractures compared with current
of different study designs. smokers. 76–80 With regard to wound healing, resistance to in-
fection is somewhat restored after about 4 weeks of smoking
Biological Plausibility cessation, 81–83 but restoration of normal collagen metabolism
The fifth criterion is biological plausibility and this was dis- appears to take considerably longer and is likely dependent on
cussed extensively in part 1 of this series. To summarize, the amount of prior smoking. 84,85 In US Army Basic Combat
1
there is evidence that smoking increase the risk of fractures 42,43 Training, individuals are not allowed to smoke, but overall in-
and tendon ruptures, 44–46 and results in delayed or incomplete jury risk is still higher among former smokers, suggesting that
wound healing. The increased fracture risk may be related residual effects of smoking are still present during the 8- to 10-
47
to smokers’ low bone mineral density, 48,49 lower dietary intake week Basic Combat Training periods that have been studied. 15
of calcium and vitamin D, 50–53 altered calcium metabolism, 52,54
and direct effects on osteogenesis 55–58 and sex hormones. 59,60 Specificity of Association
Effects on wound healing may be due to smoking-induced al- The eighth criterion is that the effect of smoking should be
terations in the functions of neutrophils and monocytes, result- specific and effect only one type of injury. Smoking appears
ing in reduced ability to fight infections and remove damaged to have effects on several tissues and does not conform well to
tissue, reduced gene expression of cytokines important for this criterion. Because of the numerous compounds in tobacco
tissue healing, and altered fibroblast function, leading to re- smoke, it is possible that smoking could affect a number of
86
duced density and amount of new tissue formation. Active tissues and processes. Current evidence suggests that smoking
47
individuals often experience overuse injuries (i.e., repeated affects injuries in bones 5,8,21,42,43,87 and tendons 44–46 and affects
61
microtraumas as a result of repeated use of specific parts of the healing processes. Smoking may exert its influence through a
body), and alterations in tissue healing may lead to tissue that variety of mechanisms. 1
is more susceptible to future injuries, such as tendon ruptures.
Consistency With Other Knowledge
Alternative Explanations The ninth and final criterion is that the association between
The sixth criterion is consideration of alternative explana- smoking and injuries should be consistent with other knowl-
tions. The relationship between smoking and injuries has been edge. There may be several ways to address this. Part 1 of this
demonstrated in a wide variety of studies, demonstrates a series provided evidence that smoking adversely affects many
1
dose-response, and has biological plausibility, as already men- physiological processes and these effects might be expected to
tioned. Several military 6,7,10,14,18 and civilian 26–29,31,34,36 investi- result in other types of medical problems. The latest report
gations have demonstrated that smoking is an independent from the Surgeon General on smoking determined that smok-
injury risk factor when considered in multivariable statistical ing is the direct cause of certain types of cancers (e.g., lung,
models that included such factors as age, sex, years of work- oral, kidney), cardiovascular diseases (e.g., aortic aneurysm,
ing, type of work, body mass index, body fat, alcohol use, atherosclerosis), respiratory diseases (e.g., chronic obstructive
education, drug use, physical fitness, physical activity, prior pulmonary disease, pneumonia), reproductive health prob-
injury, and other factors. lems (e.g., fertility, low birth rates), and other medical prob-
lems (e.g., cataracts, peptic ulcers). In addition, secondhand
88
Nonetheless, one possible alternative explanation relates to risk smoke (i.e., smoke inhaled by those near active smokers) in-
taking. As already noted in this article and in part 1, cigarette creases the risk of lung cancer, coronary heart disease, and low
1
smokers are more likely to report risk-taking behaviors 62–66 birthweight. These data support the concept that smoking
89
and generally, 67–71 but not always, 72,73 score higher on tests of has adverse effects on multiple physiological systems.
risk-taking propensity. Furthermore, one study showed that as
the amount of smoking increased, the number of risk-taking Another way to address the consistency criterion might be to
behaviors (e.g., low physical activity, low intake of fruits and examine injury severity. That is, once an individual is injured,
vegetables, alcohol use) also increased, thus suggesting a dose- are injuries more severe among smokers? Evidence comes from
response. Studies are needed that examine injury outcomes studies on disability. In the military, if a medical board deter-
64
and include measures of risk-taking behaviors and cigarette mines that a Servicemember does not meet medical retention
smoking to determine if these two behaviors are independent standards and is no longer fit for duty, that Servicemember is
or if they interact. It is also possible that other, yet unidentified, considered disabled. Among Servicemembers who were hospi-
factors may be confounding the relationship between smoking talized for a musculoskeletal disorder, risk for disability was
and injuries so alternative explanations cannot be ruled out at 18% for heavy smokers (≥1 pack/day), 16% for light to mod-
this point. For example, smokers may be less physically active erate smokers (<1 pack/day), and 12% for nonsmokers. An-
90
or have greater use of alcohol, which could increase injury risk. other study found that hospitalized smokers were 1.36 (95%
Smoking and Injuries | 119