Osteoporosis International

, Volume 16, Issue 9, pp 1086–1095

Vertebral deformity, back symptoms, and functional limitations among older women: The Framingham Study

Authors

    • University of Medicine and Dentistry New Jersey
  • Douglas P. Kiel
    • Harvard Medical School
    • Hebrew Rehabilitation Center for Aged Research and Training Institute
  • Elizabeth J. Samelson
    • Harvard Medical School
    • Hebrew Rehabilitation Center for Aged Research and Training Institute
  • Margaret Kelly-Hayes
    • Boston University School of Medicine
  • David T. Felson
    • Boston University School of Medicine
Original Article

DOI: 10.1007/s00198-004-1815-y

Cite this article as:
Edmond, S.L., Kiel, D.P., Samelson, E.J. et al. Osteoporos Int (2005) 16: 1086. doi:10.1007/s00198-004-1815-y

Abstract

Vertebral deformity is the most common manifestation of osteoporosis. The purpose of this study was to determine the relation between incident vertebral deformity and back symptoms; and limitations in nine specific functional activities. Subjects were participants in the Framingham Study, a longitudinal, population-based sample involving residents of Framingham, MA, USA. Subjects consisted of 444 surviving female members of the original cohort who had baseline (1968–1975) and follow-up (1992–1993) spine radiographs evaluated for deformity using a semiquantitative scale. Ages ranged from 72 to 96 years. At the follow-up examination, subjects were asked if they experienced pain, aching, or stiffness in their back on most days and if they had difficulty performing nine specific functional skills. We found that incident vertebral deformity was associated with limitations in several functional activities, most notably, pushing or pulling a large object (OR 2.51, 95% confidence interval 1.40, 4.52). For most activities, there was no increase in the risk of functional limitations among women with vertebral deformity who did not report back symptoms at the end of the follow-up compared with those without vertebral deformity or back symptoms. For several functional activities, individuals with both vertebral deformity and back symptoms had greater limitations than would be expected from the additive contribution of both conditions. Back symptoms were associated with limitations in most functional activities, even in the absence of vertebral deformity. Efforts to prevent and treat back pain, independent of vertebral deformity status, might help to reduce functional limitations in older women.

Keywords

Back painFunctionOlder adultsVertebral deformity

Introduction

Vertebral deformity is the most prevalent manifestation of osteoporosis [1] ranging from 10% to 45% among older women [2, 3, 4]. The full impact of vertebral deformities has not been well delineated. In this study, we investigated the association between incident vertebral deformity and functional limitations in older women. Prior research has demonstrated a positive association between back pain and discomfort, herein referred to as “back symptoms,” and functional limitations among older women [5, 6, 7]. Several investigators have questioned the extent to which functional limitations, if caused by osteoporotic vertebral deformity, are due to mechanical alterations in the spine, back symptoms that can result from the deformity [8, 9, 10], or both. We therefore investigated the role of back symptoms in the association between vertebral deformity and functional limitations.

Many vertebral deformities are asymptomatic [1, 11]. Lyles et al. postulated that osteoporotic deformities affect the alignment of the entire trunk, leading to a reduction in motion and strength [8]. The shift forward in center of gravity that occurs with progressive kyphosis from vertebral deformity could make trunk movements in a forward direction more unsteady and limited, and the loss of height could make reaching more difficult. Schlaich et al. reported that subjects with osteoporotic vertebral deformities had lower measures of vital capacity than subjects with chronic low back pain who did not have osteoporosis [10]. This reduction in lung volume could cause a reduction in exercise tolerance, which could alter functional status.

When a new osteoporotic fracture causes pain, this pain is often severe and is associated with localized tenderness and paraspinal spasm [12]. On the other hand, chronic pain secondary to osteoporotic vertebral deformity is thought to arise from muscle weakness as well as from the altered posture and spinal mechanics resulting from the deformity [9, 12]. In situations in which both vertebral deformity and back symptoms are present, functional losses could occur directly from the symptoms caused by the deformity and/or from the symptoms caused by the alteration in trunk alignment.

Numerous investigators have assessed the functional consequences associated with vertebral deformity [13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23]. Several of these investigators addressed the role of back pain by evaluating the confounding effects of pain in the association between vertebral deformity and functional limitations [14, 16, 23]. Only a few other investigators have evaluated a combination of vertebral deformity and back pain; and functional limitations [18, 22]. In one study, investigators compared the prevalence of functional limitations of subjects with vertebral deformity with subjects with back pain without vertebral deformity [18] whereas in the other study, investigators compared the prevalence of functional limitations of subjects with vertebral deformity and back pain with those of subjects with no vertebral deformity and no back pain [22]. None of those studies address the individual and the additive contributions of vertebral deformity and back symptoms to functional loss. We therefore evaluated the association of each condition and the additive contribution of these two conditions with functional limitations.

Methods

Study sample

The Framingham Study is a longitudinal cohort study of cardiovascular disease performed on a representative sample of residents of Framingham, MA, USA, a town located 18 miles west of Boston. Surviving members of the original cohort have been examined every 2 years since the study’s inception in 1948. We used data collected through examination 22, which took place between 1992 and 1993. Subjects with cognitive impairment, determined by a score of below 24 out of a total of 30 points on the Mini-Mental State Exam [24], were excluded from this study. Since the prevalence of both vertebral deformity [25, 26] and functional limitations [5, 6, 7] differs between men and women, and because there was a small number of men in our sample, we further restricted our study to female participants.

Vertebral deformity

Baseline X-rays of the lumbar spine were first taken between 1968 and 1970. At a later Framingham examination in 1975, a baseline lateral chest X-ray was taken. These two films were combined to identify baseline deformities from T4 to L4. Baseline X-rays were obtained using standard procedures for the time, although information on the precise film distance and spinal level of the X-ray beam was not recorded. Follow-up spinal radiographs were repeated between 1992 and 1993 using methods similar to those used at baseline. The radiographic technologist used the baseline lumbar spine and chest films to replicate the spinal level used in the baseline films for taking the follow-up X-rays.

A bone and joint radiologist (H.K. Genant) with full knowledge of film chronology read baseline and follow-up X-rays using a semiquantitative approach to grade vertebral deformity [27, 28]. Incident vertebral deformity was defined as grade 1 or higher (to a maximum of grade 4) at a given vertebral level on the follow-up radiograph if that deformity was not present at baseline. This classification corresponds approximately to at least a 20–25% reduction in the anterior, middle, and/or posterior height of the vertebral body [27]. Only a small percent of subjects had two or more incident vertebral deformities. Therefore, one or more deformity was compared with the absence of a deformity. Since the precise timing of the occurrence of the incident deformity between the baseline and follow-up radiographs was unknown, we calculated the cumulative incidence of vertebral deformity over 18–25 years.

Back symptoms

At the follow-up visit (1992–1993), subjects were asked, “On most days, do you have pain, aching or stiffness in any of your joints”? If subjects answered affirmatively, they were asked to identify on a homunculus where they experienced symptoms (Fig. 1). The homunculus provided a visual cue to identify symptoms in the upper back, mid back, and lower back, among other areas. This question was used in prior Framingham studies [7, 29]. We combined symptoms in the upper, mid, or lower back into one measure, representing the presence of back symptoms.
Fig. 1

The portion of the homunculus used to determine the presence and location of back symptoms using visual cues.

Functional limitations

At the follow-up visit, subjects were asked if they had no difficulty, a little difficulty, some difficulty, a lot of difficulty, or were unable to perform specific functional skills. The functional activities included standing in one place for about 15 min; walking 1/2 mile; stooping, crouching, or kneeling; lifting a 10-lb object off the floor; getting in and out of a car; pushing or pulling a large object such as a living room chair; putting on socks or stockings; reaching or extending arms above shoulder level; and writing, handling or fingering small objects. These questions are part of a generic battery of functional limitations based on work by Nagi [30] and have been used in other reports [7, 31]. This instrument remains among the most widely used instrument of functional limitations among older adults, although it was not designed to assess the unique effects of vertebral osteoporosis. Since only a small number of individuals reported an inability to perform each activity, we defined functional limitations in a particular activity as either having a lot of difficulty or being unable to perform the activity.

Confounding variables

Age at the time of follow-up and six selected medical conditions were included as potential confounders in multivariable models. We summed the number of the following conditions previously shown to be associated with functional limitations: depressive symptoms, hand symptoms, knee symptoms, coronary heart disease, prior stroke, and prior hip fracture, and created a measure representing illness burden (range 0–6). This measure of illness burden has been shown in prior Framingham studies to be associated with functional limitations [32, 33]. Depressive symptoms, hand symptoms, and knee symptoms were measured at the time of the follow-up examination. Depressive symptoms were determined by a score of 16 or greater on the CES-D depression score [34]. For hand symptoms, responses were determined by the answer to the question, “Do you have pain, aching or stiffness in either of your hands on most days”? For knee symptoms, the question was phrased similarly. Coronary heart disease (history of angina, myocardial infarction, coronary insufficiency, and/or congestive heart failure), stroke (sudden onset of a focal neural deficit lasting more than 24 h) and hip fracture (confirmed by medical record review) were considered to be present if they had been identified during any Framingham examination prior to or during the follow-up examination.

Statistical analysis

We calculated odds ratios (OR) and 95% confidence intervals (CI) for the association between vertebral deformity and functional limitations adjusting for age (in years) and illness burden (score 0–6). We then assessed the independent and joint contributions of vertebral deformity and back symptoms to functional limitations by calculating relative risk estimates adjusted for age (<80, ≥80 years) and illness burden (0–1, 2–6 medical conditions) for each of these associations. The formulas for calculating these estimates are based on stratified 2×2 tables and therefore require that age and illness burden be categorized. Cutoffs for age and illness burden were chosen to maximize control of confounding while also minimizing loss of precision. We calculated relative risks for the following three categories: the proportion of limitations in function among subjects with vertebral deformity who did not have back symptoms (“Deformity Only”), subjects with back symptoms who did not have vertebral deformity (“Symptoms Only”), and subjects with back symptoms and vertebral deformity (“Symptoms and Deformity”). In each of these analyses, the reference group consisted of subjects who did not have either vertebral deformity or back symptoms (“No Symptoms or Deformity”).

We also calculated the interaction contrast ratio of deformity and symptoms [35]. In the context of our study, the interaction contrast ratio is a measure of the extent to which vertebral deformity and back symptoms combine to result in greater functional limitations than would be expected from the presence of each condition in the absence of the other. This measure therefore addresses whether the presence of back symptoms and vertebral deformity interact to result in greater functional limitations than would be expected from the additive contributions of back pain unrelated to vertebral deformity and asymptomatic vertebral deformity.

The interaction contrast ratio was calculated using the formula [RR (vertebral deformity and back symptoms present) - RR (vertebral deformity and no back symptoms present) - RR (back symptoms and no vertebral deformity present) + 1] [35]. The first three sets of calculations, representing the independent and combined associations between vertebral deformity and back symptoms on functional limitations, result in an estimate of the relative risk, in which 1.00 represents no association. In the case of the interaction contrast ratio, the null value is represented by a result of 0.00. The extent to which there is evidence of interaction contributions is therefore determined by the extent to which the interaction contrast ratio differs from 0.

We also calculated the proportion of limitation attributable exclusively to back symptoms, vertebral deformity, and the interaction contributions of vertebral deformity and back symptoms among subjects with these two conditions as well as among the study population for each of the functional activities measured. A high attributable proportion (AP) in any of these categories would provide evidence that back symptoms and/or vertebral deformity account for a substantial portion of functional limitations in older adults.

The proportion of functional limitations attributed exclusively to back symptoms and the proportion attributed exclusively to vertebral deformity among those with both of these conditions was calculated using the formula [(RR (back symptoms and no vertebral deformity present) – 1) / RR (vertebral deformity and back symptoms present)] and [(RR (vertebral deformity and no back symptoms present) – 1) / RR (vertebral deformity and back symptoms present)] respectively. To calculate the proportion of functional limitations among those with vertebral deformity and back symptoms attributable to the interaction between the two, the interaction contrast ratio was divided by the relative risk of functional limitations among subjects with both vertebral deformity and back symptoms. The proportion of functional limitations attributable to back symptoms among the study population was calculated by multiplying the attributable proportion among the subjects with both vertebral deformity and back symptoms by the proportion of subjects with functional limitations who had vertebral deformity and/or back symptoms [35]. Data analyses were performed on SAS version 8; OR and corresponding 95% CI were determined using logistic regression. Relative risks and attributable proportions were calculated from stratified 2×2 tables.

The institutional review boards at Boston University Medical Center and the Hebrew Rehabilitation Center for Aged Research and Training Institute approved the protocol for this study. Informed consent was obtained at the time of each biennial examination.

Results

A total of 1,133 women participated in the follow-up (1992–1993) examination. Of those 1,133 women, 444 had contributed baseline radiographs and were therefore eligible for the current analysis. Female participants who attended the follow-up examination but who did not have spine X-rays were similar to subjects in regard to age, illness burden, and prevalence of back symptoms (data not shown). Ages of participants at follow-up ranged from 72 to 96 years, with a median of 78 years. The cumulative incidence of at least one vertebral deformity was 24.8%. Slightly more than 6% of women had more than two vertebral deformities. Of the 444 women, information on back symptoms reported at follow-up was available for 438. Slightly more than 37% of women with incident vertebral deformity reported back symptoms at follow-up. Of those women who did not sustain an incident vertebral deformity, 23.6% reported back symptoms at follow-up. Information on demographic characteristics of subjects by back symptoms and vertebral deformity are provided in Table 1.
Table 1

Demographic characteristics of subjects with and without back pain and vertebral deformity

Subjects with no symptoms or deformity

Subjects with deformity only

Subjects with symptoms only

Subjects with symptoms and deformity

n=252

n=68

n=78

n=40

Age

  Mean

78.5

79.3

78.3

79.0

  Standard deviation

4.7

4.2

4.4

5.6

Number of medical conditions

  0

107

28

16

12

  1

96

25

23

11

  2

36

11

23

10

  3

11

2

14

4

  4

2

2

2

3

Number of fractures

  1

40

18

  2

13

9

  3–9

15

13

Grade of fractures

  1

60

40

  2

61

43

  3

8

10

  4

0

0

The prevalence and relative risk of functional limitations according to the presence or absence of incident vertebral deformity for each functional activity is presented in Table 2. For most functional activities, limitations were more prevalent among subjects with vertebral deformity than among those without. Of note, persons with vertebral deformity had an increased risk of difficulty pushing or pulling a large object. In Table 3, data on the prevalence and relative risk of functional limitations according to each vertebral deformity and back symptoms category are presented. Functional limitations were more frequently reported among subjects with back symptoms than among those without, regardless of whether vertebral deformity was present or absent.
Table 2

The prevalence and association between incident vertebral deformity and limitations in each of the nine functional activities among female Framingham subjects. For the relative risk estimates, the comparison group consists of subjects without vertebral deformity. CI confidence interval

Functional limitation

Number of subjects with at least one vertebral deformity

Percent with functional limitations

Number of subjects with no vertebral deformity

Percent with functional limitations

Relative risk (95% CI)

Difficulty standing in one place for about 15 min

110

16.4

332

9.0

1.81 (1.05, 3.12)

Difficulty walking 1/2 mile

105

15.2

329

9.4

1.62 (0.92, 2.84)

Difficulty stooping, crouching, or kneeling

108

28.7

332

19.0

1.51 (1.04, 2.19)

Difficulty lifting a 10-lb object off the floor

108

12.0

327

5.5

2.19 (1.11, 4.31)

Difficulty getting in and out of a car

109

1.8

332

2.1

0.87 (0.18, 4.13)

Difficulty pushing or pulling a large object

106

22.6

323

10.2

2.22 (1.37, 3.57)

Difficulty putting socks or stockings on

109

2.8

332

2.7

1.02 (0.28, 3.68)

Difficulty reaching or extending arms above shoulder level

110

4.6

332

1.8

2.52 (0.78, 8.08)

Difficulty writing, handling or fingering small objects

110

5.5

332

4.8

1.13 (0.45, 2.82)

Table 3

The prevalence and association between incident vertebral deformity/back pain and limitations in each of the nine functional activities among female Framingham subjects. For the relative risk estimates, the comparison group consisted of subjects without back symptoms or vertebral deformity. CI confidence interval

Number of subjects

Percent with functional limitations

Relative risk (95% CI)

Difficulty standing in one place for about 15 min

  No symptoms or deformity

251

5.2

  Deformity only

68

5.9

1.14 (0.38, 3.37)

  Symptoms only

77

22.1

4.26 (2.17, 8.38)

  Symptoms and deformity

40

35.0

6.76 (3.43, 13.30)

Difficulty walking 1/2 mile

  No symptoms or deformity

249

6.8

  Deformity only

65

6.2

0.90 (0.31, 2.59)

  Symptoms only

76

18.4

2.70 (1.40, 5.22)

  Symptoms and deformity

38

31.6

4.63 (2.40, 8.90)

Difficulty stooping, crouching, or kneeling

  No symptoms or deformity

251

15.5

  Deformity only

67

16.4

1.06 (0.57, 1.95)

  Symptoms only

77

29.9

1.92 (1.23, 3.01)

  Symptoms and deformity

39

48.7

3.14 (2.04, 4.83)

Difficulty lifting a 10-lb object off the floor

  No symptoms or deformity

247

3.6

  Deformity only

68

10.3

2.82 (1.09, 7.31)

  Symptoms only

76

11.8

3.25 (1.34, 7.89)

  Symptoms and deformity

38

15.8

4.33 (1.63, 11.48)

Difficulty getting in and out of a car

  No symptoms or deformity

251

0.4

  Deformity only

68

0.0

1.22 (0.05, 29.59)

  Symptoms only

77

7.8

19.57 (2.39, 158.73)

  Symptoms and deformity

39

5.1

12.87 (1.20, 138.89)

Difficulty pushing or pulling a large object

  No symptoms or deformity

243

7.8

  Deformity only

65

13.9

1.77 (0.84, 3.73)

  Symptoms only

76

18.4

2.36 (1.24, 4.47)

  Symptoms and deformity

39

35.9

4.59 (2.52, 8.38)

Difficulty putting on socks or stockings

  No symptoms or deformity

251

1.2

  Deformity only

68

1.5

1.23 (0.13, 11.64)

  Symptoms only

77

7.8

6.52 (1.67, 25.45)

  Symptoms and deformity

39

5.1

4.29 (0.74, 24.88)

Difficulty reaching or extending arms above shoulder level

  No symptoms or deformity

251

0.8

  Deformity only

68

1.5

1.85 (0.17, 20.04)

  Symptoms only

77

5.2

6.52 (1.22, 34.97)

  Symptoms and deformity

40

10.0

12.55 (2.38, 66.23)

Difficulty writing, handling or fingering small objects

  No symptoms or deformity

251

3.2

  Deformity only

68

1.5

0.46 (0.06, 3.63)

  Symptoms only

77

10.4

3.26 (1.27, 8.40)

  Symptoms and deformity

40

12.5

3.92 (1.35, 11.39)

When adjusting for age and illness burden, we found a strong association between vertebral deformity and difficulty pushing or pulling a large object (OR 2.51; 95% CI 1.40, 4.52). The data were suggestive of an association between vertebral deformity and difficulty lifting a 10-lb object off the floor (OR 2.21; 95% CI 1.03, 4.73), standing in one place for about 15 min (OR 1.84; 95% CI 0.97, 3.51), and stooping, crouching or kneeling (OR 1.65; 95% CI 0.98, 2.80). There was no increase in the odds of limitation in other functional activities associated with vertebral deformity (Table 4).
Table 4

Odds ratios and 95% confidence intervals (CI) for the association between incident vertebral deformity and limitations in each of the nine functional activities among female Framingham subjectsa

Functional limitation

Odds ratio

95% CI

Difficulty standing in one place for about 15 min

1.84

(0.97, 3.51)

Difficulty walking 1/2 mile

1.55

(0.78, 3.07)

Difficulty stooping, crouching, or kneeling

1.65

(0.98, 2.80)

Difficulty lifting a 10-lb object off the floor

2.21

(1.03, 4.73)

Difficulty getting in and out of a car

0.72

(0.14, 3.67)

Difficulty pushing or pulling a large object

2.51

(1.40, 4.52)

Difficulty putting on socks or stockings

0.94

(0.25, 3.60)

Difficulty reaching or extending arms above shoulder level

2.39

(0.69, 8.26)

Difficulty writing, handling or fingering small objects

0.98

(0.36, 2.67)

a Adjusted for age (in years) and illness burden (0–6)

Among women who had vertebral deformity without back symptoms, vertebral deformity was associated only with difficulty lifting a 10-lb object off the floor (RR 2.92, 95% CI 1.15, 7.37). Conversely, among women free of incident vertebral deformity, back symptoms were associated with increased risk of most functional limitations. For example, the risk of difficulty getting in and out of a car was 6.37 (95% CI 1.37, 29.50), and for putting on socks and stockings, the risk was 4.84 (95% CI 1.39, 16.81). The presence of both vertebral deformity and symptoms was also associated with functional limitations for most activities. Relative risks for this association ranged between 2.50 and 6.95 (Table 5).
Table 5

The independent and interaction contributions of vertebral deformity and back symptoms to each of the nine functional limitations controlling for age and illness burden among female Framingham subjectsa. CI confidence intervals

Relative risk

95% CI

Difficulty standing in one place for about 15 min

  Deformity only

1.14

(0.40, 3.20)

  Symptoms only

3.40

(1.71, 6.78)

  Symptoms and deformity

5.85

(2.99, 11.48)

  Interaction contrast ratio

2.31

Difficulty walking 1/2 mile

  Deformity only

0.99

(0.38, 2.55)

  Symptoms only

2.10

(1.08, 4.06)

  Symptoms and deformity

3.60

(1.88, 6.87)

  Interaction contrast ratio

1.51

Difficulty stooping, crouching, or kneeling

  Deformity only

1.18

(0.62, 2.23)

  Symptoms only

1.51

(0.93, 2.45)

  Symptoms and deformity

2.50

(1.63, 3.81)

  Interaction contrast ratio

0.81

Difficulty lifting a 10-lb object off the floor

  Deformity only

2.92

(1.15, 7.37)

  Symptoms only

2.83

(1.15, 6.96)

  Symptoms and deformity

3.98

(1.40, 11.30)

  Interaction contrast ratio

b

Difficulty getting in and out of a car

  Deformity only

0.73

(0.03, 16.47)

  Symptoms only

6.37

(1.37, 29.50)

  Symptoms and deformity

4.90

(0.66, 36.50)

  Interaction contrast ratio

b

Difficulty pushing or pulling a large object

  Deformity only

2.08

(0.97, 4.46)

  Symptoms only

1.95

(1.00, 3.81)

Symptoms and deformity

4.15

(2.19, 7.87)

  Interaction contrast ratio

1.12

Difficulty putting on socks or stockings

  Deformity only

2.48

(0.30, 20.92)

  Symptoms only

4.84

(1.39, 16.81)

  Symptoms and deformity

4.14

(0.72, 23.92)

  Interaction contrast ratio

b

Difficulty reaching or extending arms above shoulder level

  Deformity only

1.64

(0.16, 17.24)

  Symptoms only

3.35

(0.59, 19.01)

  Symptoms and deformity

6.95

(1.27, 38.02)

  Interaction contrast ratio

2.96

Difficulty writing, handling or fingering small objects

  Deformity only

0.92

(0.18, 4.67)

  Symptoms only

1.65

(0.65, 4.15)

  Symptoms and deformity

2.74

(0.87, 8.67)

  Interaction contrast ratio

1.17

a Adjusted for age (<80, ≥80 years) and illness burden (0–1, 2–6 medical conditions)

b The interaction contrast ratio was less than 0.00, indicating that the joint contribution of vertebral deformity and back symptoms was less than additive

A primary reason for combining the presence of vertebral deformity and back symptoms was to determine whether the presence of these two conditions results in greater functional limitations than would be expected from back pain in the absence of vertebral deformity and asymptomatic deformity; however, CIs for this measure were wide (data not shown). Conclusions from these analyses should therefore be regarded as exploratory. Results suggest that both vertebral deformity and back symptoms are associated with a greater degree of limitations in at least two functional activities: reaching or extending arms above shoulder level and standing in one place for about 15 min (Table 5).

Limitations attributable to vertebral deformity were evident for lifting a 10-lb object off the floor (AP among subjects with vertebral deformity and back symptoms, 0.48; AP among all subjects, 0.21), putting on socks or stockings (AP among subjects with vertebral deformity and back symptoms, 0.36; AP among all subjects, 0.09), and pushing or pulling a large object (AP among subjects with vertebral deformity and back symptoms, 0.26; AP among all subjects, 0.08). For most functional activities, the proportion of limitation attributable to deformity was smaller than the proportion attributable to back symptoms and/or the proportion attributable to the interaction contributions of both of these conditions. Conversely, a large proportion of women with back symptoms and vertebral deformity had functional limitations attributable to back symptoms unrelated to vertebral deformity. For example, among subjects who had back symptoms and vertebral deformity, fully 46% of difficulty in lifting a 10-lb object off the floor was attributable exclusively to back symptoms (Table 6).
Table 6

Proportion of functional limitations attributable to the independent and interaction contributions of vertebral deformity and back symptoms among female Framingham subjectsa

Attributable proportion among subjects with back symptoms and vertebral deformity

Attributable proportion among the total population

Difficulty standing in one place for about 15 min

  Proportion attributable to vertebral deformity

0.02

0.00

  Proportion attributable to back symptoms

0.41

0.23

  Proportion attributable to interaction

0.39

0.17

Difficulty walking 1/2 mile

  Proportion attributable to vertebral deformity

b

b

  Proportion attributable to back symptoms

0.31

0.14

  Proportion attributable to interaction

0.42

0.17

Difficulty stooping, crouching, or kneeling

  Proportion attributable to vertebral deformity

0.07

0.02

  Proportion attributable to back symptoms

0.20

0.07

  Proportion attributable to interaction

0.32

0.10

Difficulty lifting a 10-lb object off the floor

  Proportion attributable to vertebral deformity

0.48

0.21

  Proportion attributable to back symptoms

0.46

0.23

  Proportion attributable to interaction

b

b

Difficulty getting in and out of a car

  Proportion attributable to vertebral deformity

b

b

  Proportion attributable to back symptoms

>1.0

0.94

  Proportion attributable to interaction

b

b

Difficulty pushing or pulling a large object

  Proportion attributable to vertebral deformity

0.26

0.08

  Proportion attributable to back symptoms

0.23

0.10

  Proportion attributable to interaction

0.27

0.11

Difficulty putting socks or stockings on

  Proportion attributable to vertebral deformity

0.36

0.09

  Proportion attributable to back symptoms

0.93

0.62

  Proportion attributable to interaction

b

b

Difficulty reaching or extending arms above shoulder level

  Proportion attributable to vertebral deformity

0.09

0.03

  Proportion attributable to back symptoms

0.34

0.23

  Proportion attributable to interaction

0.43

0.29

Difficulty writing, handling or fingering small objects

  Proportion attributable to vertebral deformity

b

b

  Proportion attributable to back symptoms

0.24

0.12

  Proportion attributable to interaction

0.43

0.17

a Adjusted for age (<80, ≥80 years) and illness burden (0–1, 2–6 medical conditions)

b The attributable proportion was less than 0.00, indicating protective contributions. In the case of the proportion attributable to interaction, the contribution of both vertebral deformity and back symptoms was less than additive

Data suggest interaction contributions between symptoms and deformity for limitations in reaching or extending arms above shoulder level (AP among subjects with vertebral deformity and back symptoms, 0.43; AP among all subjects, 0.29), writing, handling or fingering small objects (AP among subjects with vertebral deformity and back symptoms, 0.43; AP among all subjects, 0.17), walking 1/2 mile (AP among subjects with vertebral deformity and back symptoms, 0.42; AP among all subjects, 0.17), standing in one place for about 15 min (AP among subjects with vertebral deformity and back symptoms, 0.39; AP among all subjects, 0.17), and stooping, crouching, or kneeling (AP among subjects with vertebral deformity and back symptoms, 0.32; AP among all subjects, 0.10). This finding supports the hypothesis that these functional activities are adversely affected by the combined presence of vertebral deformity and back symptoms to a greater extent than would be expected with either individual condition. Although vertebral deformity in and of itself had only modest associations with function, vertebral deformity appeared to be more strongly associated with functional limitations when back symptoms were also present (Table 6).

Discussion

We found an association between the incidence of vertebral deformity and limitations in several functional activities. The risk of functional limitations was greater among women with incident vertebral deformity and back symptoms compared with women with incident vertebral deformity who reported no recent back symptoms. For several functional activities, vertebral deformity appeared to interact with back symptoms, with greater limitations than would be expected from the additive effect of the two conditions. Furthermore, women with recent back symptoms without evidence of incident vertebral deformity during the long follow-up period had increased risk of limitations in most of the nine functional activities.

Our findings are consistent with several other studies in which vertebral deformity was associated with increased functional limitations among older adults. In those studies in which prevalent deformities were assessed, severe deformities were associated with functional limitations. The association was less evident among subjects with moderate or mild deformities [13, 17, 19, 20]. Several investigators measured the association between incident deformities and functional limitations [15, 21]. Greendale et al. reported that the 19-year incidence among women aged 55 and older of at least one osteoporotic vertebral deformity, determined by self-report and medical review, was associated with difficulty bending, lifting, descending stairs, shopping, and cooking [15]. In a study by Nevitt et al., the severity of baseline-prevalent deformities was marginally associated with limitations in functional activities [21]. Among subjects with prevalent deformities, only those with at least one incident deformity over an average of 3.7 years of follow-up had an increased risk of functional limitations. The number of new deformities was also associated with disability.

Several studies analyzed the effect of vertebral deformity on function adjusting for back pain [14, 16, 18, 22, 23]. Data analyses in each of these studies differed from ours, as our interest in back pain itself as a predictor of functional limitations in women with vertebral deformity led us to focus our analyses on the individual and joint contribution of vertebral deformity and back symptoms on function. Two of these studies investigated a group of postmenopausal Hawaiian women of Japanese ancestry [14, 16]. Ross et al. reported that the number of prevalent deformities was associated with disability when adjusting for age, but this association became nonsignificant when incident deformities, disc disease, and traumatic back injury were included in the model [14]. Incident deformities were associated with disability in both age- and multiply adjusted models. Huang et al. reported that the number of incident vertebral deformities occurring within an average of 4 years was associated with a decrease in bending and walking-related activities and remained so after adjusting for back pain [16]. The authors concluded that recent deformities influence the ability to perform functional activities primarily through mechanisms other than back pain.

The studies by Nevitt et al., Ross et al. and Huang et al. [21, 14, 16] suggest that recent deformities have a greater impact on function than longstanding deformities. In our study, the incidence of vertebral deformity was determined over a relatively long 18- to 25-year period and included both acute and longstanding fractures. We were therefore not able to distinguish between recent and longstanding vertebral deformity with respect to current functional limitations. If acute deformities were more strongly associated with functional limitations than longstanding deformities, the association between incident deformities and functional limitations would be underestimated in our study.

In a study of Japanese women aged 40 to 89, Jinbayashi et al. [23] concluded that two or more prevalent vertebral deformities were associated with impaired function. Adjustment for back pain did not alter these findings. In our study, we were unable to analyze the effect of multiple fractures on functional limitations as there was an insufficient number of subjects with more than one vertebral deformity. It is therefore possible that the association between incident deformity and functional limitations were underestimated in our study.

Results of these latter three studies suggest that asymptomatic changes are associated with much of the limitation in function that occurs with vertebral deformity. In our study, there was no association between asymptomatic vertebral deformity and functional activities. Each of these three studies was performed on women of Japanese ancestry whereas results of the Framingham Study are most applicable to women of European American descent. Disparities in study results could be explained by ethnic and/or cultural differences or the type of functional activity studied. These disparities can also be explained by differences in the determination of vertebral deformity. We compared women with and without incident vertebral deformity whereas in these other studies, the number of deformities was analyzed. Quantifying vertebral deformity might result in greater sensitivity to differences in disability levels as subjects with more fractures could be more likely to have greater mechanical changes in spinal alignment. However, the number of women with multiple vertebral fractures in our sample was too small to examine this issue.

Leidig-Bruckner suggested that clinical measures of kyphosis be considered in addition to radiological indices of vertebral deformity in determining the relation between osteoporosis and functional limitations [18]. If we had also been able to identify kyphosis in our subjects, effects on function might have been more evident.

There are several potential methodological concerns with our study. Several of these concerns relate to study design and timing of data collection. There is the possibility that functional limitations preceded and contributed to the incidence of vertebral deformity—especially given the association between lower levels of activity in middle age that can occur when back symptoms are present, and osteoporotic changes in later years [36]. Additionally, since the duration of this study was 18–25 years, subjects who did not survive from baseline to follow-up were excluded. Our results might therefore have been affected by survivor bias. While other studies have demonstrated that women with vertebral deformity have increased mortality compared with those without vertebral deformity [37], the proportion of back symptoms, vertebral deformity, and functional limitations among cohort members who did not survive to the follow-up examination is unknown. It is therefore difficult to determine how this potential survival bias might have affected our results. Lastly, we inferred associations of back symptoms and deformities with functional limitations when these factors per se might not have been the cause of the limitations.

Other concerns relate to our variable measurements. Our question addressing back symptoms might not have identified subjects with mild or intermittent back symptoms. Since mild symptoms are less likely to result in functional limitations, this misclassification would result in an overestimation of the effects of back pain on function. Additionally, kyphosis angle was not measured for these analyses so we were not able to evaluate the contribution of kyphosis to disability.

In conclusion, we found an association between the 18- to 25-year cumulative incidence of vertebral deformity and limitations in several functional activities, most notably, pushing or pulling a large object. Functional activities were more limited when both back symptoms and deformity were present than in the sole presence of deformity. For several functional activities, vertebral deformity appeared to interact with back symptoms, resulting in greater limitations than would be expected from the joint contribution of these two conditions. Back symptoms were strongly related to limitations in most functional activities, even in the absence of vertebral deformity. Future interventions to reduce functional limitations with aging should include efforts to prevent and treat back pain, independent of vertebral deformity status.

Acknowledgements

We would like to acknowledge Harry Genant, M.D., for reading all radiographs used in this study.

Copyright information

© International Osteoporosis Foundation and National Osteoporosis Foundation 2005