Osteoporosis in COPD outpatients based on bone mineral density and vertebral fractures
One of the extrapulmonary effects of chronic obstructive pulmonary disease (COPD) is osteoporosis. Osteoporosis is characterized by a low bone mineral density (BMD) and microarchitectural deterioration. Most studies in COPD patients use dual-energy X-ray absorptiometry (DXA) only to determine osteoporosis; therefore, microarchitectural changes without a low BMD are missed. The aim of this study was to determine the prevalence and correlates of osteoporosis in COPD patients based on DXA, spinal X-rays, and combinations thereof. DXA and spinal X-rays were obtained and pulmonary function tests, body composition, 6-minute walking distance, medical history, and medication use were assessed in 255 clinically stable COPD outpatients of a large teaching hospital in the Netherlands. Half of all patients had radiologic evidence of osteoporosis. Combining the results of DXA with spinal X-rays augmented the proportion of COPD patients with osteoporosis compared with both methods separately. The prevalence of osteoporosis was not significantly different after stratification for Global Strategy for the Diagnosis, Management, and Prevention of COPD (GOLD) stage. Most patients with osteoporosis did not receive pharmacologic treatment. Age, body mass index (BMI), and parathyroid hormone (PTH) level were significant independent correlates for osteoporosis. Chest physicians should be aware of the high prevalence of osteoporosis in patients with COPD, even in the presence of a low GOLD score, as well as especially in elder COPD patients with a low BMI and/or an increased PTH level. © 2011 American Society for Bone and Mineral Research.
Although primarily a pulmonary disease, chronic obstructive pulmonary disease (COPD) has important systemic features,1 such as skeletal muscle atrophy and weakness,2 arterial stiffness,3 and osteoporosis. Indeed, COPD patients have been found to have an increased prevalence of osteoporosis compared with healthy subjects.4
Osteoporosis is a systemic skeletal disease characterized by a low bone mineral density (BMD) and/or microarchitectural deterioration resulting in an increased bone fragility and hence an increased fracture risk.5 Currently, the “gold standard” to diagnose osteoporosis is a dual-energy X-ray absorptiometry (DXA).5
The major disadvantage of DXA is that only a decreased BMD can be assessed and not changes in microarchitecture. Theoretically, in the case of vertebral fracture, the BMD of the lumbar spine as measured by DXA might even be falsely elevated. Therefore, the ideal diagnostic test for osteoporosis has the potential to assess microarchitectural changes as well as changes in BMD.
Unfortunately, microarchitectural changes can be assessed by histomorphometric analysis or micro–computed tomographic (µCT) analysis of bone biopsy samples,6 which are too invasive to be considered in a routine clinical setting. Fragility fractures can be the result of a decreased BMD as well as changes in microarchitecture. Indeed, subjects without osteoporosis as determined by DXA have been found to have vertebral fractures7–9 and therefore should be considered osteoporotic as well. Vertebral fractures can be assessed on an X-ray of the vertebral spine by a semiquantitative method according to Genant.10 Indeed, a strong relationship between severity of vertebral fracture and microarchitectural deterioration assessed by bone biopsies has been found.6
Based on DXA, the prevalence of osteoporosis in COPD patients with Global Strategy for the Diagnosis, Management, and Prevention of COPD (GOLD) stage II to IV varies between 9% and 69%.4 In addition, the prevalence of vertebral fractures of the thoracic spine was 42% in a group of 2981 COPD patients.11 Then again, the proportion of osteoporotic COPD patients will increase when local DXA scans are combined with vertebral fractures, as shown in the study by Jørgensen and colleagues.12 However, in the latter study, only 62 patients with severe to very severe COPD were studied. Moreover, patients were not characterized extensively, and patients with mild or moderate COPD were not included. Indeed, the degree of airflow limitation, measures of body composition, and the use of corticosteroids have been identified as possible clinical correlates of osteoporosis in COPD.4
To date, it remains unknown whether and to what extent these correlates of osteoporosis are interrelated in patients with COPD. Therefore, we aimed to determine the prevalence of osteoporosis in a well-characterized cohort of clinically stable COPD outpatients (GOLD stages I to IV). Moreover, we explored the importance of these factors relating to the presence of osteoporosis in a single study using local DXA scans and spinal X-rays. Early identification of correlates of osteoporosis may allow chest physicians to better monitor and stabilize the impaired BMD and/or microarchitectural changes in patients with COPD.
Materials and Methods
Patient population and study design
Clinically stable patients with a history of COPD were recruited at the outpatient respiratory department of a large clinical teaching hospital (Catharina Hospital) in Eindhoven (the Netherlands) and included in this cross-sectional study. Recruiting period was from May 2005 until July 2008. Diagnosis of COPD was made according to the American Thoracic Society (ATS)13 and categorized in accordance with the GOLD guidelines.1 Written informed consent was obtained from all patients, and the study was approved of by the Medical Ethical Committee of the Catharina Hospital Eindhoven (METC Number: M05-1522; clinical trials.gov ID number: NCT 00231127).
Medical history, medication use, smoking history, alcohol use, fractures in the past of the patient or his or her parents, and daily calcium intake were assessed with a questionnaire and by reviewing the medical charts. Cumulative oral and intravenous corticosteroid use were assessed by reviewing the medical charts from the start of referral to the chest physician until start of the study. The Charlson Comorbidity Index was used to score the presence and severity of comorbidities14 (see Online Supplement). All patients had COPD; therefore, the minimum score was 1 point.
Height and weight were measured, and body mass index (BMI) was calculated and defined as low (<21 kg/m2), normal (21 to 25 kg/m2), overweight (>25 to 30 kg/m2), and obese (>30 kg/m2). Bioimpedance analysis was done using the BODYSTAT 1500 medical, single frequency (Xitron Technologies, Eichenau, Germany), and the fat-free mass index (FFMI) was defined as depleted (men < 16 kg/m2 and women < 15 kg/m2) or normal.15
Lung function parameters were assessed using the Jaeger MASTERLAB BODY (VIASYS Healthcare, Houten, The Netherlands); postbronchodilator forced expiratory volume in the first second (FEV1) and forced vital capacity (FVC) were measured, and FEV1/FVC was calculated. All included patients had an FEV1/FVC of < 70%. According to the GOLD guidelines, patients with an FEV1 ≥ 80% predicted were classified as GOLD I, patients with an FEV1 between 50% and 80% of predicted were classified as GOLD II, patients with an FEV1 between 30% and 50% were classified as GOLD III, and finally, patients with an FEV1 < 30% were classified as GOLD IV.1 In addition, total lung capacity (TLC), residual volume (RV), and the diffusing capacity for carbon monoxide corrected for the alveolar volume (TLCOVA) were assessed, and RV as percentage of TLC (RV%TLC) was calculated.
A 6-minute walking distance (6MWD) was determined according to the most recent ATS guidelines16 to assess functional exercise tolerance. Venous blood was obtained, and total leukocytes and C-reactive protein (CRP) were determined to assess inflammatory status. In addition, vitamin D, parathyroid hormone (PTH), and thyroid-stimulating hormone (TSH) were assessed because they are known to influence bone turnover.17–19 Vitamin D deficiency was defined as 25-hydroxyvitamin D3 [25(OH)D3] < 50 nmol/L.20, 21
BMD of the neck and trochanter of the hip and of the lumbar spine (L1–L4) were measured on a Delphi upgraded to Discovery W (S/N 70991; Tromp Medical Engineering BV, Castricum, The Netherlands) DXA scan. Diagnosis of osteoporosis was based on the lowest T-score of the three measured locations and defined according to the World Health Organization (WHO; osteoporosis: T-score ≤ −2.5; osteopenia: T-score between −2.5 and −1; normal BMD: T-score > −1).5 In the case of vertebral fractures (see below) or significant osteoarthritis of the lumbar spine, BMD at the lumbar spine was not assessed.
X-rays of the lumbar spine were performed using a digital Bucky system (Digital Diagnost or H-Diagnost-Om65-CS64, Philips Medical Systems, Best, The Netherlands). For the thoracic spine, a lateral chest X-ray was used when made within 1 month of inclusion in the study. If no lateral chest X-ray was available or if the endplates of the vertebrae could not be visualized clearly, an X-ray of the thoracic spine was made. In all patients, X-rays of the lumbar spine were made. Vertebral bodies of T4 to L4 were measured independently by two investigators (LG-V and JvE), both of whom were trained by a radiologist (FJ) with special knowledge in this field.
Vertebral fractures were defined according to Genant.10 Fractures were classified as mild (20% to 25% reduction in anterior, middle, and/or posterior height and 10% to 20% reduction of the projected vertebral area), moderate (25% to 40% reduction in heights and 20% to 40% reduction of the projected vertebral area), and severe (≥40% reduction in heights). In addition, fractures were classified according to their shape: wedge (reduction in anterior height), biconcave (reduction in middle height), or crush (reduction in posterior height). In case of disagreement between the two assessors of vertebral fractures, the radiologist made the final decision. In the case of one or more vertebral fractures without high-energy trauma, patients were diagnosed as having osteoporosis. In addition, the 10-year fracture risk was calculated by using the FRAX score22 (see Online Supplement).
To compare patient characteristics between different GOLD stages, discrete variables were presented as percentages and compared with the chi-square test. Moreover, continuous variables were presented as means ± SD and compared with analysis of variance (ANOVA) and the post-hoc LSD test.
Univariate and multivariate multinomial logistic regression analyses (enter procedure) were performed to investigate determinants of osteopenia and osteoporosis based on DXA and spinal X-rays. Only patients without bone medication were analyzed. Univariate analyses with osteoporosis, osteopenia, and normal BMD as dependent variables were used to test for the potentially confounding effect of biomedical and demographic factors. If significant at p < .05, the variables were included in the multivariate analyses. A p value less than .05 was used to indicate statistical significance. Odds ratios (ORs) with 95% confidence intervals (CIs) are reported. All statistical analyses were performed using SPSS Version 16.0 (SPSS, Inc., Chicago, IL, USA).
Two-hundred and fifty-five patients with COPD were included (62% men; GOLD stages I–IV: n = 71, 100, 57, and 27, respectively). Most of the patients were ex-smokers and had an overweight BMI, a normal FFMI, and clear functional exercise intolerance (Table 1). Moreover, 44% of the COPD patients had one or more self-reported comorbidities. Gender distribution, current smoking status, self-reported comorbidities, mean age, mean BMI, and mean FFMI were not significantly different between the GOLD stages. Vitamin D deficiency significantly increased from 34% in GOLD I to 65% in GOLD IV patients.
Table 1. Patient Characteristics
|Age, years||68.05 ± 0.55||67.15 ± 1.02||69.90 ± 0.89||66.51 ± 1.11||66.78 ± 1.6|
|FEV1, % predicted||64 ± 1.3||90 ± 0.9||65 ± 0.8*||41 ± 0.8*,**||29 ± 2.7*,**,***|
|RV%TLC||51 ± 0.7||42 ± 0.9||51 ± 0.8*||58 ± 1.1*,**||62 ± 1.8*,**,***|
|TLCOVA, % predicted||80 ± 1.6||83 ± 2.3||84 ± 2.5||80 ± 3.5||62 ± 4.6*,**,***|
|BMI, kg/m2||27.06 ± 0.30||27.38 ± 0.49||27.26 ± 0.50||26.96 ± 0.62||26.51 ± 1.18|
| Low, %||8.6||4.2||10.0||5.3||18.5|
| Overweight, %||43.9||50.7||41.0||47.4||29.6|
| Obese, %||22.4||22.5||24.0||21.1||22.2|
| Males||19.2 ± 0.6||18.7 ± 0.3||18.6 ± 0.7||21.0 ± 2.0||19.3 ± 2.1|
| Females||16.1 ± 0.2||17.0 ± 0.4||16.0 ± 0.4||15.8 ± 0.5||15.3 ± 0.8|
| Low, %||21.1||7.7||25.6*||18.8||40.0*|
| Normal BMD, %||30.3||32.4||33.3||22.8||29.6|
| Osteopenia, %||46.1||47.9||48.5||43.9||37.1|
| Osteoporosis, %||23.6||19.7||18.2||33.3||33.3|
| Fracture(s), %||36.5||31.0||41.0||36.8||33.3|
| Fractures TS||0.54 ± 0.06||0.51 ± 0.12||0.50 ± 0.08||0.54 ± 0.12||0.67 ± 0.23|
| Fractures LS||0.13 ± 0.03||0.07 ± 0.05||0.22 ± 0.06||0.07 ± 0.03||0.22 ± 0.10|
| Fractures total||0.67 ± 0.07||0.58 ± 0.14||0.72 ± 0.12||0.61 ± 0.13||0.89 ± 0.27|
| DXA and spinal X- ray, %||51.4||42.3||53.0||56.1||59.3|
| Current/ex, %||26.3/73.7||23.9/76.1||27.0/73.0||31.6/68.4||18.5/81.5|
| Pack-years||36.5 ± 1.6||34.2 ± 3.3||34.5 ± 1.1||44.3 ± 4.2||33.9 ± 3.3|
|Alcohol, units/day||1.1 ± 1.7||1.2 ± 1.6||1.0 ± 1.4||1.2 ± 2.2||0.8 ± 1.3|
|Leukocytes, n/nL||7.9 ± 0.2||7.3 ± 0.2||7.8 ± 0.2||8.1 ± 0.3||9.2 ± 0.7*|
|CRP, mg/L||9.4 ± 0.9||8.1 ± 0.8||10.4 ± 2.0||8.8 ± 0.8||10.2 ± 1.7|
|25(OH)D3, nmol/L||55 ± 1.6||58.0 ± 2.9||57.6 ± 2.3||51.7 ± 4.0||44.3 ± 4.3|
|Vitamin D deficiency,a %||44||33.8||39.4||54.5*||65.4*,**|
|PTH, pmol/L||5.7 ± 3.4||5.5 ± 2.6||5.5 ± 3.7||5.9 ± 3.0||6.4 ± 4.7|
|Bone medication,b %||12.5||5.6||10.0||19.3*||25.9*,**|
|Oral corticosteroids, %||9||0||9*||12*||30*|
|Cumulative dose,c mg||3424 ± 809||611 ± 149||4354 ± 1847||2780 ± 1015||8801 ± 2508*,***|
|Inhaled corticosteroids, %||64||54||64*||68*||85*|
|6MWD, m||353 ± 7.5||395 ± 9.9||342 ± 12.0*||353 ± 18.4*||293 ± 23.1*,***|
|Charlson score||1.80 ± 0.71||1.77 ± 0.15||1.73 ± 0.12||1.86 ± 0.14||2.04 ± 0.20|
|FRAX major fracture, %||10.6 ± 6.6||10.4 ± 7.6||10.5 ± 6.3||10.1 ± 5.3||12.3 ± 7.6|
|FRAX hip fracture, %||3.5 ± 4.1||3.4 ± 4.4||3.5 ± 4.0||3.3 ± 3.2||4.3 ± 5.5|
DXA and spinal X-rays
The prevalence of osteoporosis based on DXA, spinal X-ray, and a combination thereof was 23.6%, 36.5%, and 51.4%, respectively (Table 1). Seventy (36.1%) patients without osteoporosis as assessed by DXA did have vertebral fractures without an identified history of trauma as the likely cause. Of these 68 patients, 43 were osteopenic and 27 had a normal BMD. In addition, 38 (23.5%) patients without vertebral fractures had osteoporosis on DXA. The mean 10-year fracture risk was 10.6% ± 6.6% for a major fracture and 3.5% ± 4.1% for a hip fracture. Moreover, most of patients with osteoporosis did not use physician-prescribed bone medication: 78%, 78%, and 80% for osteoporosis based on DXA, spinal X-ray, or a combination thereof, respectively. In addition, 41% of patients with osteopenia should be treated with bone medication according to the National Osteoporosis Foundation (NOF) guidelines23 (10-year hip fracture risk ≥ 3% and major fracture risk ≥ 20%). Of these 48 patients, 42 were not using bone medication (87.5%).
There were 93 patients (36.5%) with one or more vertebral fractures resulting in a total of 172 fractures. Most patients had one, two, or three vertebral fractures (52.7%, 25.8%, and 14%, respectively), whereas there were some patients with four, five, six, and even seven vertebral fractures (3.2%, 1.1%, 2.2%, and 1.1%, respectively).
Most of these fractures were wedge-shaped (n = 123, 71.5%). In addition, 41 fractures (23.8%) were biconcave, and only 8 were crush fractures (4.7%). There were 110 mild (64%), 50 moderate (29%), and 12 severe fractures (7%). The locations of the fractures showed a bimodal distribution with a peak at the level of thoracic vertebra 8 and one peak at the level of thoracic vertebra 12. Type and location of fractures did not differ significantly between GOLD stages. Before the study, only 27 patients (10.6%) were diagnosed on routine chest X-ray; of these, only 8 patients knew that they had a vertebral fracture. This study showed that 80 patients (31.4%) should have been diagnosed with a fracture of the thoracic spine.
DXA and spinal X-ray after stratification for GOLD stages
Using DXA, GOLD stage I had the highest proportion of patients with osteopenia, whereas GOLD stages II and IV had the highest proportion of patients with osteoporosis. Nevertheless, differences between GOLD stages were nonsignificant (p = .314).
GOLD stage II had the highest proportion of patients with vertebral fractures, whereas GOLD stage IV had the highest mean number of vertebral fractures. Again, differences between GOLD stages were nonsignificant (p = .588 for proportion of patients with vertebral fractures and p = .631 for mean number of vertebral fractures).
The prevalence of osteoporosis (DXA + spinal X-ray) increased from 42% to 59% from GOLD I to GOLD IV but was not statistically significantly different between GOLD stages (p = .301; Table 1).
Determinants of osteopenia and osteoporosis in COPD
To assess risk factors for osteopenia and osteoporosis, only patients without bone medication were analyzed (n = 223, 87% of total). Of these patients, 69 had osteopenia and 105 had osteoporosis based on DXA and spinal X-ray. Univariate multinomial regression analysis showed that lower BMI and a myocardial infarction in the history significantly increased the OR for osteopenia (Table 2). In addition, higher age, lower BMI, lower 6-minute walking distance, lower FEV1, higher RV/TLC, higher PTH level, and a myocardial infarction significantly increased the OR for osteoporosis (Table 2). Multivariate multinomial regression analysis with gender, age, BMI, 6-minute walking distance (6MWD), FEV1, RV/TLC, PTH level, and myocardial infarction as covariables showed that only BMI is an independent significant predictor of osteopenia (Table 3). Age, BMI, and PTH level were independent predictors of osteoporosis (Table 3). There was no significant correlation between PTH level and corticosteroid use (r = 0.148, p = .512). PTH level had no significant influence when corticosteroid use was used as the dependent variable.
Table 2. Correlates of Osteopenia and Osteoporosis in COPD: Results From Univariate Multinomial Regression Analysis
| GOLD II||0.806||0.344–1.892||.621||1.259||0.569–2.785||.569|
| GOLD III||2.094||0.682–6.432||.197||2.414||0.816–7.137||.111|
| GOLD IV||0.924||0.215–3.965||.915||1.731||0.474–6.324||.406|
| Vitamin D||1.009||0.995–1.024||.207||0.996||0.982–1.010||.558|
| Corticosteroid||NAb|| || ||2.379||0.786–7.202||.125|
| Cummulative coricosteroids||1.000||1.000–1.000||.649||1.000||1.000–1.000||.108|
Table 3. Correlates of Osteopenia and Osteoporosis: Results of Multivariate Multinomial Regression Analysis
The current study has several main findings:
- 1.Of the patients with clinically stable COPD attending a regular visit at the respiratory outpatient consultation in a general hospital, 51% had radiologic evidence of osteoporosis.
- 2.Combining the results of local DXA with spinal X-rays augmented the proportion of COPD patients with osteoporosis compared with both methods separately.
- 3.The proportion of COPD patients with osteoporosis was not statistically significantly different after stratification for GOLD stages.
- 4.A large proportion of the osteoporotic COPD patients did not use physician-prescribed bone medication.
- 5.Age, BMI, and PTH level were clear correlates of osteoporosis in COPD patients who did not use physician-prescribed bone medication.
Prevalence of osteoporosis
In this study, the combination of local DXA and spinal X-ray resulted in an osteoporosis prevalence of 51% in outpatients with COPD (Table 1). This is completely in line with the findings of Jørgensen and colleagues,12 who only included COPD outpatients with an FEV1 of less than 45% of predicted and mainly female patients.
In this trial, a large consecutive sample consisting of COPD outpatients representing all GOLD stages was screened for osteoporosis. Nonsignificant differences were found in the proportion of osteoporotic COPD patients after stratification for GOLD stages. These findings suggest that extrapulmonary features such as osteoporosis are poorly predicted by the degree of airflow limitation.
Indeed, 42% of the GOLD stage I outpatients had osteoporosis. Even though an age-matched control group is lacking in the current design, it seems reasonable to conclude that the prevalence of osteoporosis is higher than in healthy elderly subjects.4 These results suggest that the increased prevalence of osteoporosis in COPD patients is not or only partly dependent on the degree of airflow limitation. Indeed, there are more factors involved in the prognosis and morbidity in COPD patients.2 The question is what the link is between COPD and extrapulmonary features such as osteoporosis, arterial stiffness, and muscle atrophy, if any, besides airflow limitation. This could be chronic inflammation, reduced physical activity, or a common genetic susceptibility, or maybe these features are just comorbidities in COPD patients. More research is needed to clarify these links.
The prevalence of osteoporosis in patients in GOLD stages I to IV in this study was higher than the prevalence of osteoporosis in COPD patients in the same GOLD stages entering pulmonary rehabilitation24: 58% versus 16%, 56% versus 30%, and 59% versus 54% in GOLD stages I–II, III, and IV, respectively. These differences may be due to a referral bias, which probably will occur when referring complex and symptomatic COPD patients from a general hospital to a specialized rehabilitation center.25 However, in the pulmonary rehabilitation patients, vertebral fractures were not assessed and therefore not included in the diagnosis of osteoporosis. In addition, whole-body DXA was used as opposed to DXA of the hip and lumbar spine in this study, which is a more sensitive method to assess osteoporosis in COPD patients.26
The prevalence of vitamin D deficiency increased from 34% to 65% in patients in GOLD stages I through IV. This is in line with the recent study by Janssens and colleagues, who found vitamin D prevalences of 39%, 47%, 60%, and 77% for patients in GOLD stages I through IV, respectively.27 The prevalence in COPD patients was higher than in healthy subjects with a minimum of 15 pack-years of smoking (31%).
About a third of the patients had one or more vertebral fractures, mostly located at thoracic vertebrae 8 and 12. This is in line with previous studies in patients with COPD and inflammatory bowel disease, as well as in postmenopausal women and healthy men.11, 28–33 This bimodal distribution is probably due to different patterns of spinal loading in different segments of the spine and to variations in biomechanical and material properties of vertebral bodies.28, 32
Based on routine chest X-ray, only 34% of the thoracic vertebral fractures were diagnosed. Moreover, only 8 patients were made aware of these vertebral fractures. Radiologists and (chest) physicians should be made aware of the high prevalence of vertebral fractures in COPD patients so that they know that they should be paying attention to the vertebrae.
Treatment of osteoporosis
Undertreatment of patients with osteoporosis was high. This is in line with COPD patients entering pulmonary rehabilitation.24 In the Netherlands, there is no FRAX cutoff recommended for treatment. Therefore, we used the NOF cutoff values to assess the percentage of osteopenic COPD patients that should receive treatment to prevent osteoporotic fractures.23 However, these are guidelines for the United States, and treatment threshold should take into account that the 10-year fracture risk differs by country.34 The 10-year probability of a hip fracture for men and women at ages 50, 60, 70, and 80 does not differ that much between the United States and the Netherlands (difference of 0.1% to 1.3% for men and 0.2% to 1.7% for women).34 Therefore, we felt that using the NOF cutoff values can be justified. Adjustment of the Dutch Guidelines for Prevention and Treatment of Osteoporosis is necessary to implicate the FRAX score.
Correlates of osteopenia
Increasing BMI independently decreased the risk of osteopenia in this study. Indeed, in COPD patients entering pulmonary rehabilitation, cachectic body composition independently increased the OR for osteopenia.24 No other studies were found that investigate correlates of osteopenia in COPD patients.
Univariate analyses showed that previous myocardial infarction was a significant predictor for osteopenia as well as osteoporosis. The link between osteoporosis and cardiovascular diseases and osteoporosis and COPD, respectively, could be age, a hormonal link, a common susceptibility owing to the same risk factors (eg, smoking), chronic systemic inflammation, or even a common genetic susceptibility. More research is needed to clarify this link. In multivariate analyses, the significance disappeared.
Correlates of osteoporosis
Increasing age, PTH level, and a decreasing BMI independently increased the OR for osteoporosis. Higher age is a risk factor for osteoporosis in the general population.5 When osteoporosis is based on DXA only, age is usually not an independent predictor of osteoporosis.4 However, increasing age is associated with an increasing risk of vertebral fracture in COPD patients.11, 35–37 Therefore, the increased risk of osteoporosis in older COPD patients might be the result of more vertebral fractures owing to a deteriorated microarchitecture at this age.
The finding that a decreasing BMI is associated with an increasing risk of osteoporosis is in line with several other studies in COPD patients.24, 31, 38–42 The link between low BMI and osteoporosis in COPD patients is not entirely clear yet. It might be decreased physical activity, increased inflammation, or other mechanisms leading to proteolysis.43–45 More studies are needed investigating this link.
Increasing PTH significantly increased the OR for osteoporosis. Patients with high PTH levels all had normal calcium levels; hence they had secondary hyperparathyroidism. Secondary hyperparathyroidism is caused by vitamin D deficiency.46 Indeed, we found a significant negative correlation between vitamin D and PTH in our study population (r = −0.28, p < .001). This is in line with Kuchuc and colleagues, who found an increasing vitamin D level significantly decreasing PTH.19 In addition, markers of bone turnover decreased significantly. Combined with our results, a decreasing PTH level may cause decreasing speed of bone turnover, resulting in a better microarchitecture and hence less osteoporosis.
Because there is still a lot of discussion about the best cutoff value for vitamin D deficiency, we also used 80 nmol/L as the cutoff, which is used in postmenopausal women with osteoporosis. The prevalence of vitamin D deficiency increased to 86%. The significant difference between GOLD stages disappeared (GOLD I 87%, GOLD II 84%, GOLD III 89%, GOLD IV 85%; p = .829); however, the possible correlation with osteoporosis was still not found.
There was no significant influence of FEV1 on osteoporosis. Indeed, the prevalence of osteoporosis did not differ between GOLD stages. This result suggests that there are more factors involved in the prognosis and morbidity of COPD patients than airflow limitation.47
This cross-sectional study has some methodologic considerations. Interpretation of clinical correlates of osteoporosis should be with caution in a cross-sectional design. Second, there could be a selection bias because we included COPD outpatients treated by chest physicians and not patients treated by family physicians. However, the current consecutive sample of clinically stable COPD patients has been collected prospectively in a general hospital, including all GOLD stages. Consequently, it can be considered to be a representative sample of Dutch COPD outpatients attending a regular outpatient COPD clinic.
Conclusions and future research
Half the patients with clinically stable COPD attending a regular outpatient COPD clinic have osteoporosis, as determined by local DXA and spinal X-ray independent of GOLD classification. Correlates of osteoporosis in COPD are age, BMI, and PTH level. Given the current definition of COPD as a multicomponent disease, the awareness of a high prevalence of osteoporosis should be raised. Chest physicians not only should describe the lung parenchyma on a chest X-ray but also the vertebrae. In addition, especially in older patients with a low BMI and/or secondary hyperparathyroidism, DXA and an X-ray of the total spine should be done. Future studies are needed to assess the external validity of our findings in a primary-care setting. In addition, follow-up studies are needed to determine risk factors for osteoporosis in COPD patients based DXA and spinal X-ray.
EFMW is a member of the scientific advisory boards for GSK, Boehringer Ingelheim, AstraZeneca, and Numico and received lecture fees from GSK, AstraZeneca, and Boehringer Ingelheim. He received research grants between 2004 and 2007 from GSK, AstraZeneca, Boehringer Ingelheim, Centocor and Numico. All the authors state that they have no conflicts of interest.
We thank the Department of Pulmonary Function Testing of the Catharina Hospital Eindhoven and especially Loes van den Nieuwenhuizen for the assessment with pulmonary function testing and determination of body composition and 6-minute walking distance. We thank the Diagnostic Centre Eindhoven for the DXA scans. We thank Frits Jansen, radiologist of the Catharina Hospital Eindhoven, for his help with assessment of vertebral fractures. We thank Joris van Enschot, of Catharina Hospital Eindhoven, for his assessment of vertebral fractures.