Abstract
Background: pain may reduce stability and increase falls and subsequent fractures in older men.
Objectives: to examine the association between joint pain and any pain with falls, hip and non-spine fractures in older community-dwelling men.
Design: a cohort study.
Setting and Participants: analyses included 5,993 community-dwelling men aged ≥65 years from the MrOS cohort.
Measurements: pain at hip, knee and elsewhere (any) was assessed by self-report. Men reported falls via questionnaires mailed 3× per year during the year following the baseline visit. Fractures were verified centrally. Mean follow-up time for fractures was 9.7 (SD 3.1) years. Logistic regression models estimated likelihood of falls and proportional hazards models estimated risk of fractures. Models were adjusted for age, BMI, race, smoking, alcohol use, medications use, co-morbidities and arthritis; fracture models additionally adjusted for bone mineral density.
Results: one quarter (25%, n = 1,519) reported ≥1 fall; 710 reported ≥2 falls in the year after baseline. In multivariate models, baseline pain at hip, knee or any pain increased likelihood of ≥1 fall and ≥2 falls over the following year. For example, knee pain increased likelihood of ≥1 fall (odds ratio, OR 1.44; 95% confidence interval, CI 1.25–1.65) and ≥2 falls (OR 1.75; 95% CI 1.46–2.10). During follow-up, 936 (15.6%) men suffered a non-spine fracture (n = 217, 3.6% hip). In multivariate models, baseline pain was not associated with incident hip or non-spine fractures.
Conclusions: any pain, knee pain and hip pain were each strong independent risk factors for falls in older men. Increased risk of falls did not translate into an increased risk of fractures.
Keywords: pain, fall, fracture, cohort, MrOS, older people
Introduction
Studies have shown that almost 30% of the American population over 65 years of age will fall each year [1]. Of these, 20–30% sustain moderate to severe injuries, straining costs of the healthcare system [1, 2]. Fractures are a leading cause of morbidity in the old [3]; in 2002, treatment of osteoporotic fractures suffered by the male population in the USA cost $3.2 billion, or 18% of total costs associated with osteoporosis [4]. The increased incidence of fractures with age has no single causative factor but is due to a combination of weakening bone structure, increased falls and altered nature of falls [5].
An association between pain and future falls has been well established for both chronic and acute pain [6–14]. Location of pain has in multiple studies been shown to be of importance; however, there have been conflicting reports concerning pain at the knee, especially with regards to recurrent falls [11]. The underlying mechanisms are controversial but postural instability and cognitive distraction or impairment caused by either the pain itself or the associated depressive symptoms have been suggested.
Despite the well-established association between pain and future falls, it remains unclear whether this translates into an increased risk of fractures. Pain has in multiple studies shown to both increase and decrease bone mineral density (BMD) [15–18]. It is unclear whether increased BMD associated with pain is protective of fractures as the remodelling might lead to lower shock resistance [18–22]. A single study has examined the association between pain and future vertebral fractures in post-menopausal women, finding sufferers of low back pain to be at higher risk [23]. Patients suffering from knee osteoarthritis have in a single study been found to have increased risk of fractures even after adjusting for falls [6].Given the availability of self-reported pain in a clinical setting, pain could be an important readily available independent risk factor for falls and fractures. The aim of the present study was to look for a possible association between self-reported hip, knee or any pain with future falls or hip or non-spine fractures in older community-dwelling men.
Methods
MrOS study
MrOS is a multicentre, prospective study of healthy ageing focusing on osteoporosis and fractures in older men. Design, measures and recruitment methods have been described previously [24, 25]. Briefly, 5,994 community-dwelling men aged 65 years or older were recruited by targeted mailing, advertisements and media outreach to local communities from March 2000 to April 2002 at six clinical sites (Birmingham, AL; Minneapolis, MN; Palo Alto, CA; Monongahela Valley near Pittsburgh, PA; Portland, OR; and San Diego, CA, USA). Local institutional review boards approved study protocol, and all participants provided written informed consent. Primary inclusion criteria were ability to walk unassisted, the absence of bilateral hip replacements, ability to provide self-reported clinical experiences and willingness to visit a clinical site for physical measurements.
Measurements
Pain
We assessed pain using self-administered questionnaires. Any pain was determined by a SF-12 questionnaire question that has shown to be both valid and reliable in similar settings [26, 27]: ‘During the past 4 weeks, how much did pain interfere with your normal work (including both work outside the home and housework)?’. Pain was rated on a 5-step scale (not at all, a little bit, moderately, quite a bit and extremely), with ‘any pain’ defined as pain that was moderate or greater. This cut-off point was used as pain of lesser intensity is unlikely to impact a participant's daily life. Hip pain was queried with the question: ‘In the past 12 months, have you had pain in or around either hip joint, including the buttock, groin or either side of the upper thigh, on most days for at least 1 month? Do not include pain from the lower back’. Participants were dichotomously classified. A similar question was asked for knee pain substituting ‘knee joint’ for ‘hip joint’.
Bone mineral density
BMD at the total hip and femoral neck was obtained using DXA scans (model QDR 4500; Hologic, Waltham, MA, USA). Extensive quality assurance protocols were used throughout, including central training and certification of technicians and regular phantom DXA scans within and across centres [25].
Falls and fractures
Participants were queried about incident falls and fractures quarterly by mail, with response rates exceeding 99%. All fractures were centrally verified by a physician adjudicator using medical records. Hip and non-spine fractures were analysed. Men were followed for fractures from baseline until 25 January 2013; mean follow-up time was 9.7 (SD 3.06) years. We defined a participant with ‘any fall’ as someone who reported at least one fall in the year after baseline, and ‘2+ falls’ as someone who reported 2+ falls in the year after baseline.
Other measurements
Demographic characteristics included age, race/ethnicity, smoking, alcohol consumption and pre-existing diseases (questions were designed as ‘have you ever been told you have X disease by a physician?’). Participants were asked to bring all prescription medications they had taken daily or almost daily for at least 1 month to the clinic visit. If a participant forgot to bring one or more medications, clinic staff obtained this information by telephone or at a return visit. All medications were verified by pill bottle examination and matched to its ingredient(s) based on the Iowa Drug Information Service (IDIS) Drug Vocabulary (College of Pharmacy, University of Iowa, Iowa City, IA, USA) [28].
Body weight (balance beam or digital scale) and height (stadiometer) as well as physical performance measurements (timed 6 m walk, chair rise and grip strength) were collected at clinic visit. The frailty definition was derived from the one used in the Cardiovascular Health Study and has been used in prior analysis of the MrOS cohort. A man was considered frail if three or more of the following five criteria were present: shrinking/sarcopenia (lowest 20% of appendicular lean mass, adjusted for height and total body fat), low activity (lowest 20% of Physical Activity for the Elderly score), weakness (lowest 20% in grip strength by BMI), slowness (slowest 20% 6 m walking speed by height) or low energy (self-reported). Men with one or two criteria were considered intermediate.
Statistical analysis
Difference in baseline characteristics between participants with or without pain was assessed using t-tests for normally distributed continuous variables, Mann–Whitney U test for skewed continuous variables and χ2 tests for categorical variables (P < 0.05). Time to event risk of hip and non-spine fractures was estimated using proportional hazard models. Likelihood of falls was estimated from logistic regression models.
Models used in the analysis were unadjusted; adjusted for age, BMI and clinical site (base model); and multivariate models adjusted for relevant covariates with two different models for physical performance status. All final multivariate models include age, BMI, clinical site, race/ethnicity, smoking, alcohol consumption, number of pain medications, number of other medications, number of co-morbidities and site-specific arthritis (knee, hip or any). Fracture models additionally included femoral neck BMD. Participants missing covariates were excluded from the specific multivariate models.
Of the 5,994 men in MrOS, 1 was excluded prior to data analysis due to missing pain data, 2 for missing BMI, 97 for missing falls data and 111 men for missing data for 2+ falls. This left 5,894 in the ‘any falls’ base model, 5,880 in the ‘2+ falls’ base model and 5,991 in the fracture base model.
Final multivariate adjusted models included 5,550 for any falls, 5,536 for 2+ falls and 5,628 for both hip and non-spine fractures.
Statistical analysis was performed using SAS v 9.x (SAS Institute Inc, Cary, NC, USA).
Results
Participants
Of 5,993 subjects with pain data, 1,218 (20.3%) reported any pain, 1,453 (24.2%) reported hip pain and 1,964 (32.8%) reported knee pain at baseline. Of 1,218 who reported any pain, 338 (27.8%) did not report either hip or knee pain. When stratifying by pain categories, there was a significant difference between pain and no pain groups for most covariates (Table 1). Those who reported pain (either any or at the hip or knee) were older (except those reporting knee pain), more likely to smoke, consumed less alcohol and had worse physical performance as well as were more likely to be frail. Those reporting pain used more medication and were also more likely to suffer from co-morbidities, especially arthritis. Mean femoral neck BMD was significantly higher for those reporting pain at knee or hip, but not for those reporting any pain. Those reporting any pain or knee pain were more likely to take bisphosphonates.
Table 1.
Characteristics of MrOS participants by any pain, hip pain or knee pain
| No any pain (n = 4,775) | Any pain (n = 1,218) | P values | No hip pain (n = 4,540) | Hip pain (n = 1,453) | P values | No knee pain (n = 4,029) | Knee pain (n = 1,964) | P values | |
|---|---|---|---|---|---|---|---|---|---|
| Age [years], mean (SD) | 73.5 (5.8) | 74.4 (6.1) | <0.001 | 73.6 (5.8) | 74.0 (6.0) | 0.02 | 73.6 (5.8) | 73.8 (5.9) | 0.25 |
| White, n (%) | 4,257 (89.2) | 1,104 (90.6) | 0.13 | 4,055 (89.3) | 1,306 (89.9) | 0.54 | 3,605 (89.5) | 1,757 (89.5) | 0.99 |
| BMIa cat, n (%) | |||||||||
| ≤25 | 1,371 (28.7) | 272 (22.3) | <0.001 | 1,299 (28.6) | 344 (23.7) | <0.001 | 1,255 (31.2) | 388 (19.8) | <0.001 |
| >25–30 | 2,497 (52.3) | 572 (47.0) | 2,338 (51.5) | 731 (50.3) | 2,080 (51.6) | 989 (50.4) | |||
| >30–35 | 747 (15.6) | 291 (23.9) | 747 (16.5) | 291 (20.0) | 589 (14.6) | 449 (22.9) | |||
| >35–40 | 142 (3.0) | 67 (5.5) | 134 (3.0) | 75 (5.2) | 97 (2.4) | 112 (5.7) | |||
| >40 | 18 (0.4) | 16 (1.3) | 22 (0.5) | 12 (0.8) | 8 (0.2) | 26 (1.3) | |||
| Smoking, n (%) | 2,898 (60.7) | 846 (69.5) | <0.001 | 2,773 (61.1) | 971 (66.8) | <0.001 | 2,444 (60.7) | 1,300 (66.2) | <0.001 |
| Alcohol use, n (%) | |||||||||
| 0 drinks per week | 1,569 (32.9) | 551 (45.4) | <0.001 | 1,541 (34.0) | 579 (39.9) | <0.001 | 1,374 (34.1) | 746 (38.1) | 0.01 |
| 1–13 drinks per week | 2,622 (55.0) | 546 (44.9) | 2,463 (54.3) | 705 (48.6) | 2,176 (54.1) | 992 (50.6) | |||
| 14+ drinks per week | 579 (12.1) | 118 (9.7) | 531 (11.7) | 166 (11.5) | 476 (11.8) | 221 (11.3) | |||
| Physical performance | |||||||||
| History of falls, n (%) | 895 (18.7) | 373 (30.6) | <0.001 | 905 (19.9) | 363 (25.0) | <0.001 | 779 (19.3) | 489 (24.9) | <0.001 |
| 6 m walk [m/s], mean (SD) | 1.23 (0.21) | 1.09 (0.24) | <0.001 | 1.21 (0.22) | 1.15 (0.24) | <0.001 | 1.22 (0.22) | 1.15 (0.23) | <0.001 |
| Average grip strength [kg] (SD) | 39.0 (8.1) | 36.9 (8.3) | <0.001 | 38.8 (8.1) | 37.9 (8.3) | <0.001 | 38.8 (8.0) | 38.1 (8.5) | 0.002 |
| Ability to rise from chair w/out arms, n(%) | 4,713 (98.8) | 1,151 (94.7) | <0.001 | 4,455 (98.2) | 1,409 (97.1) | <0.001 | 3,967 (98.6) | 1,897 (96.6) | <0.001 |
| Frailty status, n (%) | |||||||||
| Robust | 2,269 (47.5) | 355 (29.2) | <0.001 | 2,071 (45.6) | 553 (38.1) | <0.001 | 1,872 (46.5) | 752 (38.3) | <0.001 |
| Intermediate | 2,189 (45.9) | 612 (50.3) | 2,090 (46.1) | 711 (49.0) | 1,827 (45.4) | 974 (49.6) | |||
| Frail | 315 (6.6) | 251 (20.6) | 378 (8.3) | 188 (13.0) | 329 (8.2) | 237 (12.1) | |||
| Analgesics use | |||||||||
| Opiods, n (%) | 56 (1.2) | 105 (8.9) | <0.001 | 77 (1.8) | 84 (6.0) | <0.001 | 70 (1.8) | 91 (4.8) | <0.001 |
| COX2, n (%) | 196 (4.3) | 176 (14.9) | <0.001 | 224 (5.2) | 148 (10.6) | <0.001 | 176 (4.6) | 196 (10.4) | <0.001 |
| NSAIDsb, n (%) | 529 (11.6) | 360 (30.5) | <0.001 | 552 (12.7) | 337 (24.1) | <0.001 | 420 (10.9) | 469 (24.9) | <0.001 |
| Oral corticosteroids, n (%) | 75 (1.6) | 51 (4.3) | <0.001 | 95 (2.2) | 31 (2.2) | 0.95 | 80 (2.1) | 46 (2.4) | 0.36 |
| No. of prescription drugs used, n (%) | |||||||||
| 0 | 614 (12.9) | 60 (4.9) | <0.001 | 563 (12.4) | 111 (7.6) | <0.001 | 503 (12.5) | 171 (8.7) | <0.001 |
| 1–3 | 1,846 (38.7) | 322 (26.4) | 1,691 (37.3) | 477 (32.8) | 1,566 (38.9) | 602 (30.7) | |||
| 4+ | 2,315 (48.5) | 836 (68.6) | 2,286 (50.4) | 865 (59.5) | 1,960 (48.7) | 1,191 (60.6) | |||
| Medical conditions, n (%) | |||||||||
| Arthritis | 1,958 (41.0) | 888 (72.9) | <0.001 | 1,892 (41.7) | 954 (65.7) | <0.001 | 1,518 (37.7) | 1,328 (67.6) | <0.001 |
| Knee osteoarthritis | 358 (7.5) | 231 (19.0) | <0.001 | 386 (8.5) | 203 (14.0) | <0.001 | 113 (2.8) | 476 (24.2) | <0.001 |
| Hip osteoarthritis | 151 (3.2) | 163 (13.4) | <0.001 | 79 (1.7) | 235 (16.2) | <0.001 | 157 (3.9) | 157 (8.0) | <0.001 |
| Cancer | 1,354 (28.4) | 391 (32.1) | 0.01 | 1,290 (28.4) | 455 (31.3) | 0.03 | 1,149 (28.5) | 596 (30.4) | 0.14 |
| COPDc | 433 (9.1) | 207 (17.0) | <0.001 | 454 (10.0) | 186 (12.8) | 0.003 | 389 (9.7) | 251 (12.8) | <0.001 |
| Angina | 564 (11.8) | 292 (24.0) | <0.001 | 578 (12.7) | 278 (19.1) | <0.001 | 519 (12.9) | 337 (17.2) | <0.001 |
| Hypertension | 1,967 (41.2) | 613 (50.3) | <0.001 | 1,867 (41.1) | 713 (49.1) | <0.001 | 1,639 (40.7) | 941 (47.9) | <0.001 |
| Cong. heart failure | 221 (4.6) | 96 (7.9) | <0.001 | 211 (4.7) | 106 (7.3) | <0.001 | 185 (4.6) | 132 (6.7) | <0.001 |
| Heart attack | 588 (12.3) | 246 (20.2) | <0.001 | 587 (12.9) | 247 (17.0) | <0.001 | 512 (12.7) | 322 (16.4) | <0.001 |
| Diabetes | 478 (10.0) | 174 (14.3) | <0.001 | 458 (10.1) | 194 (13.4) | 0.001 | 394 (9.8) | 258 (13.1) | <0.001 |
| Low thyroid | 306 (6.4) | 113 (9.3) | 0.001 | 313 (6.9) | 106 (7.3) | 0.60 | 278 (6.9) | 141 (7.2) | 0.69 |
| High thyroid | 68 (1.4) | 29 (2.4) | 0.02 | 68 (1.5) | 29 (2.0) | 0.19 | 56 (1.4) | 41 (2.1) | 0.04 |
| Parkinson | 33 (0.7) | 19 (1.6) | 0.004 | 37 (0.8) | 15 (1.0) | 0.44 | 37 (0.9) | 15 (0.8) | 0.54 |
| Stroke | 240 (5.0) | 104 (8.5) | <0.001 | 234 (5.2) | 110 (7.6) | 0.001 | 230 (5.7) | 114 (5.8) | 0.88 |
| Bone health | |||||||||
| Femoral neck BMDd [g/cm2], mean (SD) | 0.78 (0.13) | 0.79 (0.14) | 0.60 | 0.78 (0.13) | 0.79 (0.13) | 0.01 | 0.78 (0.13) | 0.80 (0.13) | <0.001 |
| Total hip BMDd [g/cm2], mean (SD) | 0.96 (0.14) | 0.96 (0.15) | 0.71 | 0.96 (0.14) | 0.97 (0.15) | 0.02 | 0.95 (0.14) | 0.97 (0.15) | <0.001 |
| Bisphosphonates, n (%) | 69 (1.5) | 32 (2.7) | 0.01 | 78 (1.8) | 23 (1.7) | 0.70 | 79 (2.1) | 22 (1.2) | 0.02 |
| Previous fractures, n (%) | 2,552 (53.5) | 765 (62.8) | <0.001 | 2,488 (54.8) | 829 (57.1) | 0.13 | 2,165 (53.8) | 1,151 (58.7) | <0.001 |
aBody mass index.
bNon-steroidal anti-inflammatory drug.
cChronic obstructive pulmonary disease.
dBone mineral density.
Falls
During first year of follow-up, 1,519 (25.3%) men reported any falls and 710 (11.8%) men reported 2+ falls. In unadjusted logistic regression models, men reporting any pain, knee pain and hip pain were more likely to have any falls or 2+ falls in the year after baseline than men without pain (Table 2). Adjustment for age, clinic and BMI as well as full multivariate adjustment attenuated associations, but results remained significant. The largest odds ratio for falls was seen for any pain and lowest for hip pain.
Table 2.
Risk of falls by pain category in older men
| Any falls | 2+ falls | |
|---|---|---|
| Any pain | ||
| Unadjusted, OR (95% CI) | 2.02 (1.76–2.31) | 2.57 (2.17–3.04) |
| Base model adjusted, OR (95% CI)a | 1.92 (1.67–2.20) | 2.39 (2.01–2.84) |
| Multivariate adjusted, OR (95% CI)b | 1.65 (1.42–1.92) | 2.00 (1.65–2.41) |
| Multivariate and physical performance, OR (95% CI)c | 1.55 (1.32–1.81) | 1.79 (1.47–2.19) |
| Multivariate and frailty, OR (95% CI)d | 1.53 (1.31–1.78) | 1.82 (1.50–2.21) |
| Hip pain | ||
| Unadjusted, OR (95% CI) | 1.48 (1.30–1.69) | 1.72 (1.45–2.03) |
| Base model adjusted, OR (95% CI)a | 1.43 (1.25–1.64) | 1.64 (1.38–1.95) |
| Multivariate adjusted, OR (95% CI)e | 1.30 (1.12–1.50) | 1.46 (1.21–1.76) |
| Multivariate and physical performance, OR (95% CI)f | 1.24 (1.06–1.43) | 1.37 (1.13–1.67) |
| Multivariate and frailty, OR (95% CI)g | 1.27 (1.10–1.47) | 1.43 (1.19–1.73) |
| Knee pain | ||
| Unadjusted, OR (95% CI) | 1.62 (1.43–1.83) | 2.00 (1.71–2.35) |
| Base model adjusted, OR (95% CI)a | 1.58 (1.39–1.79) | 1.93 (1.63–2.27) |
| Multivariate adjusted, OR (95% CI)h | 1.44 (1.25–1.65) | 1.75 (1.46–2.10) |
| Multivariate and physical performance, OR (95% CI)i | 1.36 (1.18–1.57) | 1.65 (1.37–1.99) |
| Multivariate and frailty, OR (95% CI)j | 1.41 (1.23–1.62) | 1.72 (1.43–2.06) |
aAge, BMI and clinical site of recruitment.
bAge, BMI, clinical site of recruitment, race, smoking, alcohol, analgesics, meds, co-morbidities and arthritis.
cAge, BMI, clinical site of recruitment, race, smoking, alcohol, analgesics, meds, co-morbidities, arthritis, walking speed, grip strength and ability to rise from chair.
dAge, BMI, clinical site of recruitment, race, smoking, alcohol, analgesics, meds, co-morbidities, arthritis and frailty.
eAge, BMI, clinical site of recruitment, race, smoking, alcohol, analgesics, meds, co-morbidities and hip arthritis.
fAge, BMI, clinical site of recruitment, race, smoking, alcohol, analgesics, meds, co-morbidities, hip arthritis, walking speed, grip strength and ability to rise from chair.
gAge, BMI, clinical site of recruitment, race, smoking, alcohol, analgesics, meds, co-morbidities, hip arthritis and frailty.
hAge, BMI, clinical site of recruitment, race, smoking, alcohol, analgesics, meds, co-morbidities and knee arthritis.
iAge, BMI, clinical site of recruitment, race, smoking, alcohol, analgesics, meds, co-morbidities, knee arthritis, walking speed, grip strength and ability to rise from chair.
jAge, BMI, clinical site of recruitment, race, smoking, alcohol, analgesics, meds, co-morbidities, knee arthritis and frailty.
Fractures
During follow-up over a mean of 9.7 years, 936 (15.6%) men experienced a non-spine fracture and 217 (3.6%) a hip fracture. The most common locations of first non-spine fractures were ribs (n = 187, 19.98% of non-spine fractures), hip fractures (any site) (n = 186, 19.87%) and ankle (n = 85, 9.08%). In unadjusted models, men reporting any pain were more likely to experience hip or non-spine fracture during follow-up (Table 3). After adjustment for age, BMI and clinic site, this association was attenuated, but remained significant. In fully adjusted models, this was further attenuated and no longer significant. Neither hip nor knee pain was associated with hip or non-spine fractures in any models.
Table 3.
Risk of fractures by pain category in older men
| Hip fractures | Non-spine fractures | |
|---|---|---|
| Any pain | ||
| Unadjusted, HR (95% CI) | 1.35 (0.99–1.84) | 1.16 (0.98–1.37) |
| Base model, HR (95% CI)a | 1.42 (1.04–1.94) | 1.31 (1.12–1.53) |
| Multivariate adjusted, HR (95% CI)b | 1.16 (0.82–1.63) | 1.18 (0.99–1.40) |
| Multivariate and physical performance, OR (95% CI)c | 1.00 (0.69–1.45) | 1.11 (0.92–1.32) |
| Multivariate and frailty, OR (95% CI)d | 1.01 (0.71–1.44) | 1.09 (0.91–1.30) |
| Hip pain | ||
| Unadjusted, HR (95% CI) | 1.09 (0.80–1.49) | 1.11 (0.95–1.31) |
| Base model, HR (95% CI)a | 1.08 (0.80–1.47) | 1.15 (1.00–1.34) |
| Multivariate adjusted, HR (95% CI)e | 1.04 (0.75–1.46) | 1.16 (0.99–1.36) |
| Multivariate and physical performance, OR (95% CI)f | 0.98 (0.69–1.39) | 1.18 (1.00–1.39) |
| Multivariate and frailty, OR (95% CI)g | 0.95 (0.68–1.33) | 1.13 (0.96–1.33) |
| Knee pain | ||
| Unadjusted, HR (95% CI) | 1.21 (0.91–1.60) | 1.05 (0.90–1.22) |
| Base model, HR (95% CI)a | 1.19 (0.90–1.58) | 1.09 (0.95–1.25) |
| Multivariate adjusted, HR (95% CI)h | 1.17 (0.86–1.59) | 1.10 (0.95–1.28) |
| Multivariate and physical performance, OR (95% CI)i | 1.10 (0.80–1.52) | 1.06 (0.91–1.24) |
| Multivariate and frailty, OR (95% CI)j | 1.11 (0.81–1.52) | 1.07 (0.92–1.25) |
aAge, BMI and clinical site of recruitment.
bAge, BMI, clinical site of recruitment, race, smoking, alcohol, analgesics, meds, co-morbidities, arthritis and femoral neck BMD.
cAge, BMI, clinical site of recruitment, race, smoking, alcohol, analgesics, meds, co-morbidities, arthritis, femoral neck BMD, walking speed, grip strength and ability to rise from chair.
dAge, BMI, clinical site of recruitment, race, smoking, alcohol, analgesics, meds, co-morbidities, arthritis, femoral neck BMD and frailty.
eAge, BMI, clinical site of recruitment, race, smoking, alcohol, analgesics, meds, co-morbidities, hip arthritis and femoral neck BMD.
fAge, BMI, clinical site of recruitment, race, smoking, alcohol, analgesics, meds, co-morbidities, hip arthritis, femoral neck BMD, walking speed, grip strength and ability to rise from chair.
gAge, BMI, clinical site of recruitment, race, smoking, alcohol, analgesics, meds, co-morbidities, hip arthritis, femoral neck BMD and frailty.
hAge, BMI, clinical site of recruitment, race, smoking, alcohol, analgesics, meds, co-morbidities, knee arthritis and femoral neck BMD.
iAge, BMI, clinical site of recruitment, race, smoking, alcohol, analgesics, meds, co-morbidities, knee arthritis, femoral neck BMD, walking speed, grip strength and ability to rise from chair.
jAge, BMI, clinical site of recruitment, race, smoking, alcohol, analgesics, meds, co-morbidities, knee arthritis, femoral neck BMD and frailty.
Discussion
Any pain, i.e. moderate or greater, was the strongest risk factor for future falls among the three pain variables analysed, with a modest association for hip pain. The increased odds ratios were not explained by age, BMI or clinical study site. Associations were attenuated when controlling for covariates, including physical performance, but significance was retained for all categories of pain. Thus, association between pain and falls was independent of physical performance, medical conditions, analgesic and other medication use.
Any pain being more strongly related to falls than hip pain or knee pain may be due to differences in how these categories of pain were assessed. For any pain, participants rated how pain interfered with their work over the past for weeks on a scale (which was dichotomised for analysis), while for the hip and knee pain questions, participants were asked about the presence of pain, aching or stiffness over the past 12 months without rating severity. This difference could have resulted in differing severities of pain being included in the any pain versus the hip and knee pain definitions. Both Arden et al. [6] and Eggermont et al. [ 29] reported that fall risk increased as pain intensity increased, emphasizing the importance of pain intensity. Additionally, those with higher pain intensity might be more likely to remember and subsequently report a fall. While all three pain locations are likely to compromise gait and consequent balance, it is likely that any pain, as the only category including back pain, effected gait the most.
We found no association between hip or knee pain and future hip or non-spine fractures in unadjusted models. We did find a moderate association between any pain and hip and non-spine fractures after adjusting for age, BMI and clinical site of recruitment, but there was no significant association in fully adjusted multivariate models. This finding contradicts reports by Arden et al. [6] who reported an increased risk of both hip and non-spine fractures for those reporting knee pain.
There are several plausible explanations why an increased risk of falls did not translate into an increased risk of fractures. (i) Even though we found an increased risk of falls, these falls were not of a ‘character’ that would result in fractures. It is possible that those who fell managed to brace themselves or the resultant force would be in a direction in which a fracture was unlikely. As no standard data were collected concerning nature of the falls or activity at time of falling, it is not possible to assess this. (ii) The pain-induced inflammation that we hypothesised may lower strength of bones through bone remodelling might not have translated into an altered BMD. The level of inflammation associated with lower intensity pain might not be sufficient to cause the altered BMD reported in other studies [18–20]. No information regarding duration of pain was collected; it is possible that pain and subsequent inflammation had not been present in participants long enough to translate into lowered bone strength. (iii) It is possible that pain status changed during the follow-up, that is, some of those reporting no pain at baseline may have developed pain in subsequent years or vice versa. However, analyses that truncated follow-up at 5 years were similar to analyses with full follow-up time, suggesting changing pain status did not substantially influence results. (iv) Non-spine fractures are a heterogeneous group and risk factors for fracture in different skeletal locations likely differ [30]. We found no association between pain and non-spine fractures, nor with hip fractures—the most clinically important type of fracture.
The strengths of our study are the large, well-characterised cohort with a high rate of follow-up and centralised verification of fractures. A limitation of our study is time between assessment of pain and incidence of fracture; however, truncating follow-up did not significantly alter our findings. Other limitations of our study are lack of pain intensity measurements, lack of pain duration measurements, possible under-reporting of falls and limited data concerning use of non-prescription medications, specifically analgesics. The MrOS cohort consists mainly of white, community-dwelling, relatively healthy men, which might reduce generalisability of these findings.
In summary, we confirmed that complaints of pain, whether at knee, hip or any site, were associated with significantly increased risk of future falls in older men. We did not find an association between pain and fractures. Future studies should account for intensity of pain at all locations, pain duration and use of non-prescription analgesics. Datasets with repeated measures of pain over time such that pain assessments are available in closer temporal proximity to subsequent fractures may shed more light on the association between pain and fractures.
Key points.
Self-reported pain is readily available and has the potential to increase risk of falls as well as fractures.
In several studies, self-reported pain at hip, knee or elsewhere was an independent risk factor for falls in older men.
Self-reported pain was not associated with an increased risk of fractures in community-dwelling men.
Conflicts of interest
None declared.
Funding
The Osteoporotic Fractures in Men (MrOS) Study is supported by National Institutes of Health funding. The following institutes provide support: the National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS); the National Institute on Aging (NIA); the National Center for Research Resources (NCRR); and NIH Roadmap for Medical Research with following grant numbers: (U01 AR45580, U01 AR45614, U01 AR45632, U01 AR45647, U01 AR45654, U01 AR45583, U01 AG18197, U01-AG027810 and UL1 RR024140). National Institutes of Health grants (K24-AR048841, P50 AR063043, AR052000 and P50AR060752 to N.E.L.). PMC has received grants from GSK, IMS Health and Merck as well as consulted for KineMed and Eli Lilly. J.T.S. has received a grant from Bone Ultrasound Finland, Ltd. The remaining authors declare no conflict of interest. Financial sponsors played no role in the design, analysis or interpretation of data.
References
- 1.Roudsari BS, Ebel BE, Corso PS et al. The acute medical care costs of fall-related injuries among the U.S. older adults. Injury 2005; 36: 1316–22. [DOI] [PubMed] [Google Scholar]
- 2.Rizzo JA, Friedkin R, Williams CS et al. Health care utilization and costs in a Medicare population by fall status. Med Care 1998; 36: 1174–88. [DOI] [PubMed] [Google Scholar]
- 3.Hausdorff JM, Rios DA, Edelberg HK. Gait variability and fall risk in community-living older adults: a 1-year prospective study. Arch Phys Med Rehabil 2001; 82: 1050–6. [DOI] [PubMed] [Google Scholar]
- 4.U.S. Department of Health and Human Services. Bone health and osteoporosis: A Report of the Surgeon General. Rockville, MD: U.S. Department of Health and Human Services, Office of the Surgeon General, 2004. [Google Scholar]
- 5.Stone KL, Seeley DG, Lui LY et al. BMD at multiple sites and risk of fracture of multiple types: long-term results from the Study of Osteoporotic Fractures. J Bone Miner Res 2003; 18: 1947–54. [DOI] [PubMed] [Google Scholar]
- 6.Arden NK, Crozier S, Smith H et al. Knee pain, knee osteoarthritis, and the risk of fracture. Arthr Rheumat 2006; 55: 610–5. [DOI] [PubMed] [Google Scholar]
- 7.Lee SH, Dargent-Molina P, Breart G et al. Risk factors for fractures of the proximal humerus: results from the EPIDOS prospective study. J Bone Miner Res 2002; 17: 817–25. [DOI] [PubMed] [Google Scholar]
- 8.Nahit ES, Silman AJ, Macfarlane GJ. The occurrence of falls among patients with a new episode of hip pain. Ann Rheum Dis 1998; 57: 166–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Sturnieks DL, Tiedemann A, Chapman K et al. Physiological risk factors for falls in older people with lower limb arthritis. J Rheumatol 2004; 31: 2272–9. [PubMed] [Google Scholar]
- 10.Stubbs B, Binnekade T, Eggermont L et al. Pain and the risk for falls in community-dwelling older adults: systematic review and meta-analysis. Arch Phys Med Rehabil 2014; 95: 175–87.e9. [DOI] [PubMed] [Google Scholar]
- 11.Stubbs B, Schofield P, Binnekade T et al. Pain is associated with recurrent falls in community-dwelling older adults: evidence from a systematic review and meta-analysis. Pain Med 2014; 15: 1115–28. [DOI] [PubMed] [Google Scholar]
- 12.Muraki S, Akune T, Oka H et al. Prevalence of falls and the association with knee osteoarthritis and lumbar spondylosis as well as knee and lower back pain in Japanese men and women. Arthritis Care Res 2011; 63: 1425–31. [DOI] [PubMed] [Google Scholar]
- 13.Leveille SG, Jones RN, Kiely DK et al. Chronic musculoskeletal pain and the occurrence of falls in an older population. JAMA 2009; 302: 2214–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Patel KV, Phelan EA, Leveille SG et al. High prevalence of falls, fear of falling, and impaired balance in older adults with pain in the United States: findings from the 2011 National Health and Aging Trends Study. J Am Geriatr Soc 2014; 62: 1844–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Briggs AM, Straker LM, Wark JD. Bone health and back pain: what do we know and where should we go? Osteoporos Int 2009; 20: 209–19. [DOI] [PubMed] [Google Scholar]
- 16.Briggs AM, Straker LM, Burnett AF et al. Chronic low back pain is associated with reduced vertebral bone mineral measures in community-dwelling adults. BMC Musculoskelet Disord 2012; 13: 49. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Lee S, Nam CM, Yoon DH et al. Association between low-back pain and lumbar spine bone density: a population-based cross-sectional study. J Neurosurg Spine 2013; 19: 307–13. [DOI] [PubMed] [Google Scholar]
- 18.Snider KT, Johnson JC, Degenhardt BF et al. Low back pain, somatic dysfunction, and segmental bone mineral density T-score variation in the lumbar spine. J Am Osteopath Assoc 2011; 111: 89–96. [PubMed] [Google Scholar]
- 19.Leppala J, Kannus P, Natri A et al. Bone mineral density in the chronic patellofemoral pain syndrome. Calcif Tissue Int 1998; 62: 548–53. [DOI] [PubMed] [Google Scholar]
- 20.Gaber TA, McGlashan KA, Love S et al. Bone density in chronic low back pain: a pilot study. Clin Rehabil 2002; 16: 867–70. [DOI] [PubMed] [Google Scholar]
- 21.Jones G, Nguyen T, Sambrook PN et al. Osteoarthritis, bone density, postural stability, and osteoporotic fractures: a population based study. J Rheumatol 1995; 22: 921–5. [PubMed] [Google Scholar]
- 22.Bergink AP, van der Klift M, Hofman A et al. Osteoarthritis of the knee is associated with vertebral and nonvertebral fractures in the elderly: the Rotterdam Study. Arthritis Rheum 2003; 49: 648–57. [DOI] [PubMed] [Google Scholar]
- 23.Kuroda T, Shiraki M, Tanaka S et al. The relationship between back pain and future vertebral fracture in postmenopausal women. Spine 2009; 34: 1984–9. [DOI] [PubMed] [Google Scholar]
- 24.Blank JB, Cawthon PM, Carrion-Petersen ML et al. Overview of recruitment for the osteoporotic fractures in men study (MrOS). Contemp Clin Trials 2005; 26: 557–68. [DOI] [PubMed] [Google Scholar]
- 25.Orwoll E, Blank JB, Barrett-Connor E et al. Design and baseline characteristics of the osteoporotic fractures in men (MrOS) study—a large observational study of the determinants of fracture in older men. Contemp Clin Trials 2005; 26: 569–85. [DOI] [PubMed] [Google Scholar]
- 26.Ware J Jr, Kosinski M, Keller SD. A 12-Item Short-Form Health Survey: construction of scales and preliminary tests of reliability and validity. Med Care 1996; 34: 220–33. [DOI] [PubMed] [Google Scholar]
- 27.Haywood KL, Garratt AM, Fitzpatrick R. Quality of life in older people: a structured review of generic self-assessed health instruments. Qual Life Res 2005; 14: 1651–68. [DOI] [PubMed] [Google Scholar]
- 28.Pahor M, Chrischilles EA, Guralnik JM et al. Drug data coding and analysis in epidemiologic studies. Eur J Epidemiol 1994; 10: 405–11. [DOI] [PubMed] [Google Scholar]
- 29.Eggermont LH, Bean JF, Guralnik JM et al. Comparing pain severity versus pain location in the MOBILIZE Boston study: chronic pain and lower extremity function. J Gerontol Ser A Biol Sci Med Sci 2009; 64: 763–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Armstrong ME, Cairns BJ, Banks E et al. Different effects of age, adiposity and physical activity on the risk of ankle, wrist and hip fractures in postmenopausal women. Bone 2012; 50: 1394–400. [DOI] [PMC free article] [PubMed] [Google Scholar]