Page 99 - JSOM Fall 2018

diagnosed radiographically, although some reviews used only exercise. Reviews that pooled results from six or more studies
radiographic evidence. Exercise was generally considered to be and examined outcomes at the end of training programs re­
any regular, systematic physical activity (e.g., walking, aquatic ported SMDs ranging from 0.16 to 0.81 for pain and 0.21
51
59
57
activity, resistance training, flexibility, calisthenics). to 0.86 for physical functioning. Two more recent reviews 59,60
56
containing 35 to 47 studies found SMDs of 0.50 and 0.49
The outcome measures examined in the reviews were self­re­ for pain and 0.49 and 0.52 for physical functioning. These
ported pain and/or physical function or disability as quantified findings indicate exercise had a moderate effect on reducing
on self­reporting instruments. Examples of instruments used pain and improving physical functioning among individuals
were the Western Ontario and McMaster University Osteoar­ with OA.
thritis Index (WOMAC), the Knee Injury and Osteoarthritis
Outcome Score (KOOS), and the Visual Analogue Scale (VAS). Two reviews that looked at longer term follow­ups (after the
The WOMAC has 24 questions that are summarized in 3 sub­ end of formal physical training) reported lower SMDs of 0.08
60
scales including pain (5 questions), stiffness (2 questions), and 0.16 for pain, and 0.20 and 0.21 for physical func­
60
51
51
physical functioning (17 questions), and a total score (24 ques­ tioning. This was likely because many patients did not con­
tions by adding the three subscales). Different versions of the tinue to exercise after the formal training concluded. There
51
questionnaire contain either a 5­point Likert­like scale or a were two or three investigations where the authors provided
100mm VAS (range, “none” to “extreme”). Patients answer “booster sessions,” but the definition of booster sessions was
51
the questions based on their pain, stiffness, and physical func­ not clear. An examination of the individual studies suggested
tioning over the last 48 hours. 65,66 The KOOS is specific to that booster sessions involved unsupervised exercise sessions
the knee. It includes many questions from the WOMAC but at home, some with periodic communication to assess ex­
72
has additional subscales and was developed for younger pa­ ercise compliance and efficacy. Much higher SMDs were
73
tients. The KOOS has 42 questions summarized in 5 subscales shown in the few studies that included these booster sessions
including pain (9 questions), other symptoms (7 questions), after the end of the “formal” training (see Table 1, row 3).
51
functions of daily life (17 questions), function in sport and Not surprisingly, this suggests that exercise must be continued
recreation (5 questions), and knee­related quality of life (4 to maintain the favorable effects.
questions). A 5­point Likert­like scale is used to answer each
question. The questionnaire can be used over short and longer Characteristics of Exercise Programs
time intervals, for example, to assesses changes from week to Characteristics of exercise programs that can be manipulated
week due to a treatment. 67,68 Although there are variations, the include the mode of training (e.g., aerobic, resistance, flexi­
VAS is generally a 10­point scale on which participants rate bility), frequency (sessions/week), duration (minutes/session),
the intensity of their pain (e.g., “no pain” to “worst possible intensity (e.g., speed in aerobic training or load in resistance
pain”). 69 training), and length (weeks). 74
Differences between groups (e.g., exercise versus nonexercise) The most studied characteristic is the mode of training. A low
in the reviews were determined by the standardized mean dif­ level of strength or power in the knee extensor muscles has
ference (SMD). A SMD is the difference in the average change been related to knee pain and functional disability 75–77 and
in pain or disability (before minus after treatment) between this has been hypothesized to contribute to the pain and dis­
the two groups divided by the pooled standard deviation, as ability of OA. Thus, it is possible that strength training of
78
follows: muscle groups around symptomatic areas might be more ef­
fective than other modes of exercise in reducing pain and im­
(pre­post treatment group 1 − pre­post treatment group 2 ) / proving physical functioning. Several reviews 50,57,59 separately
standard deviation. compared aerobic and resistance exercise with nonexercising
control subjects. An early review involving 14 studies found
50
This measure reflects the difference in the change in the ratings larger SMDs for walking versus resistance exercise for both
of pain or disability between the two groups while considering pain and disability (Table 1, row 2), but the latest review
59
the amount of variability in the measurement. A SMD of 0.2 (Table 1, row 11), involving 35 to 47 studies, found SMDs
to 0.4 is considered small, 0.5 to 0.7 as moderate, and 0.8 differed little between the two modes of exercise (aerobic ver­
or greater as large. 70,71 If any included review did not provide sus resistance training) for either pain or physical functioning.
pooled SMD, SMDs were calculated from data provided in the One review (Table 1, row 9) used a network meta­analysis in
57
article using Comprehensive Meta­Analysis Software, version which various exercise interventions were compared equally
3.3 (Biostat, https://www.meta­analysis.com/). by comparing treatments within the same trial (direct evi­
dence) and across different trials (indirect evidence) so that
Overall Results the most effective type(s) of exercise could be determined.
Table 1 provides the pooled SMD from the included reviews They concluded that a training program incorporating aero­
of randomized controlled trials examining the effectiveness of bic training, resistance training, and flexibility exercises would
exercise for reducing pain and disability associated with OA. likely be most effective in the management of OA.
All reviews in Table 1 considered studies that compared exer­
cising groups with nonexercising groups, but several 50,55,57,59,61 Aquatic exercise may be advantageous for individuals with
looked at more than one mode of exercise, allowing compari­ OA because (1) water buoyancy supports the body, making
sons between modes. movement less painful; (2) warm water promotes relaxation,
which may reduce muscle spasm and tightness; and (3) ex­
As shown in Table 1, virtually all reviews reported lower lev­ ercise intensity can be increased by increasing the speed of
els of pain and higher levels of physical functioning among movement in the water. However, the one review (Table 1,
79
55
groups involved in exercise compared with those that did not row 7) that examined seven studies comparing aquatic and
Osteoarthritis and Exercise | 97