Chapter 12 – Osteoporosis: Clinical Evaluation

Updated 28 Nov 2010

Introduction

Osteoporosis is a common disease characterized by low bone strength that results in an increased risk of fracture (1). Fractures are associated with serious clinical consequences, including long-term disability, increased risk of death, and high healthcare costs. Early identification and intervention with patients at high risk for fracture is needed to reduce the burden of osteoporotic fractures (2). The management of a patient with a confirmed diagnosis of osteoporosis or low bone mass (osteopenia) includes assessment of fracture risk, evaluation for secondary causes of skeletal fragility, making decisions on initiation of treatment, and identification of all relevant clinical factors that may influence patient management. This is a review of the key components in the care of patients with osteoporosis prior to treatment.

Diagnosis of osteoporosis

The World Health Organization (WHO) diagnostic classification (Figure 1) (3) is made by bone mineral density (BMD) testing with dual-energy X-ray absorptiometry (DXA) using the T-score, calculated by subtracting the mean BMD (in g/cm2) of a young-adult reference population from the patient's BMD and dividing by the standard deviation (SD) of the young-adult reference population. The International Society for Clinical Densitometry (ISCD) recommends that BMD be measured at the lumbar spine (L1L4), total hip, and femoral neck, with the 33% radius (1/3 radius) being measured when the lumbar spine and/or hip cannot be measured (e.g., obese patient who exceeds weight limit of table) or is invalid (e.g., patient with lumbar laminectomy) (4). Osteoporosis cannot be diagnosed by BMD measurement at skeletal sites other than lumbar spine, femoral neck, and 33% radius or with technologies other than DXA. The quality of DXA instrument maintenance, acquisition, analysis, interpretation, and reporting is important in obtaining valid results that can be used for making appropriate clinical decisions (4;5). In a patient with a fragility fracture, a clinical diagnosis of osteoporosis may be made independently of BMD results, assuming that other causes of skeletal fragility (e.g., osteomalacia) are not responsible for the fracture. Establishing a diagnosis of osteoporosis is clinically useful because it facilitates communication among healthcare providers and patients concerning a disease with potentially serious consequences; in some countries, such as the United States (US), a diagnosis is necessary in order to select a numerical code for submission of insurance claims for reimbursement for medical services.

Figure 1. World Health Organization criteria for classification of patients with bone mineral density measured by dual-energy X-ray absorptiometry (3).

 

The National Osteoporosis Foundation (NOF) has released indications for BMD testing in men and women in the US (Figure 2) (6), based on ISCD Official Positions (4), with consideration that BMD testing should be done only when it is likely to have an influence on patient management decisions.  Other organizations and other countries with different economic resources and health care priorities have used a variety of methodologies to develop alternative recommendations (7-9).

Figure 2. National Osteoporosis Foundation indications for bone mineral density testing in the United States (6).

• Women age 65 and older and men age 70 and older, regardless of clinical risk factors
• Younger postmenopausal women and men age 50 to 69 about whom you have concern based on their clinical risk factor profile
• Women in the menopausal transition if there is a specific risk factor associated with increased fracture risk such as low body weight, prior low-trauma fracture or high risk medication
• Adults who have a fracture after age 50
• Adults with a condition (e.g., rheumatoid arthritis) or taking a medication (e.g., glucocorticoids in a daily dose ≥ 5 mg prednisone or equivalent for ≥ three months) associated with low bone mass or bone loss
• Anyone being considered for pharmacologic therapy for osteoporosis
• Anyone being treated for osteoporosis, to monitor treatment effect
• Anyone not receiving therapy in whom evidence of bone loss would lead to treatment
• Postmenopausal women discontinuing estrogen should be considered for bone density testing

 

Fracture risk assessment

There is a robust correlation between BMD and fracture risk, with approximately a 2-fold increase in fracture risk for every 1 SD decrease in BMD (10). However, many or most patients with a hip fracture have a T-score better than -2.5 (11); although fracture risk is higher in patients with very low BMD, there are numerically many more patients with a T-score better than -2.5 than with a T-score of -2.5 or worse, therefore more fractures in those with higher T-scores. The presence of clinical risk factors (CRFs) that are independent of BMD, particularly age and prior fracture, can help identify patients at high risk for fracture by providing information on fracture risk that is complementary to BMD. The NOF has provided an extensive list of CRFs (Figure 3) for osteoporosis and fractures. Since most fractures occur as a result of a fall, it is helpful to recognize risk factors for falling (Figure 4) so that appropriate interventions can be made, when possible, to reduce the chances of falling.

Figure 3. Conditions, diseases and medications that cause or contribute to osteoporosis and fractures (6).

Lifestyle factors
         Low calcium intake
         Vitamin D insufficiency
         Excess vitamin A
         High caffeine intake
         High salt intake
         Aluminum (in antacids)
         Alcohol (3 or more drinks/d)
         Inadequate physical activity
         Immobilization
         Smoking (active or passive)
         Falling
         Thinness

Genetic factors
         Cystic fibrosis
         Homocystinuria
         Osteogenesis imperfecta
         Ehlers-Danlos syndrome
         Hypophosphatasia
         Parental history of hip fracture
         Gaucher's disease
         Idiopathic hypercalciuria
         Porphyria
         Glycogen storage diseases
         Marfan syndrome
         Riley-Day syndrome
         Hemochromatosis
         Menkes steely hair syndrome
         Hypogonadal states
         Androgen insensitivity
         Hyperprolactinemia
         Turner's & Klinefelter's syndromes
         Anorexia nervosa and bulimia
         Panhypopituitarism
         Athletic amenorrhea
         Premature ovarian failure

Endocrine disorders
         Adrenal insufficiency
         Diabetes mellitus
         Thyrotoxicosis
         Cushing's syndrome
         Hyperparathyroidism
        
Gastrointestinal disorders
         Celiac disease
         Inflammatory bowel disease
         Primary biliary cirrhosis
         Gastric bypass
         Malabsorption
         GI surgery
         Pancreatic disease

Hematologic disorders
         Hemophilia
         Multiple myeloma
         Systemic mastocytosis
         Leukemia and lymphomas
         Sickle cell disease
         Thalassemia

Rheumatic and autoimmune diseases
         Ankylosing spondylitis
         Lupus
         Rheumatoid arthritis

Miscellaneous conditions and diseases
         Alcoholism
         Emphysema
         Muscular dystrophy
         Amyloidosis
         End stage renal disease
         Parenteral nutrition
         Chronic metabolic acidosis
         Epilepsy
         Post-transplant bone disease
         Congestive heart failure
         Idiopathic scoliosis
         Prior fracture as an adult
         Depression
         Multiple sclerosis
         Sarcoidosis

Medications
Anticoagulants (heparin)
Cancer chemotherapeutic drugs
Gonadotropin releasing hormone agonists
Anticonvulsants
Cyclosporine A and tacrolimus
Lithium
Aromatase inhibitors
Depo-medroxyprogesterone
Barbiturates
Glucocorticoids (≥ 5 mg/d of prednisone or equivalent for ≥ 3 mo)

Vertebral fracture assessment (VFA)

VFA is a method for imaging the thoracic and lumbar spine by DXA for the purpose of detecting vertebral fracture deformities. Identification of a previously unrecognized vertebral fracture may alter diagnostic classification, change estimation of fracture risk, and influence treatment decisions (12). In comparison with standard radiographs of the spine, the correlation for detecting moderate and severe vertebral fractures is good, with a smaller dose of ionizing irradiation, greater patient convenience (i.e., it may be done at the same visit and with the same instrument as BMD testing by DXA), and lower cost. In a study of women age 65 and older, using the Genant semi-quantitative (SC) method of classifying vertebral deformities (13), the sensitivity of VFA for diagnosing moderate (grade 2) and severe (grade 3) vertebral fractures was 87-93%, with a specificity of 93-95% (14). Indications for DXA are listed in Figure 4. Optimal use of DXA and VFA requires training and adherence to well established quality standards (4).

Figure 4. International Society for Clinical Densitometry (ISCD) indications for vertebral fracture assessment (VFA) (4;90). This is a summary of the ISCD recommendations for consideration of VFA, with the caveat that VFA should not be done unless it is likely to influence clinical decisions. 

BMD	Women	Men
N/A	Chronic glucocorticoid therapy (equivalent to 5 mg or more of prednisone for 3 months or longer)
Osteoporosis	If documentation of 1 or more vertebral fractures will alter clinical management
Osteopenia plus any one of the following	Age ≥ 70 HHL > 4 cm (1.6 in.) PHL > 2 cm (0.8 in.) Self-reported VF        (not previously documented)	Age ≥ 80 HHL > 6 cm (2.4 in.) PHL > 3 cm (1.2 in.) Self-reported VF        (not previously documented)
Osteopenia plus two or or more of the following	Age 60-69 Self-reported NVF HHL 2-4 cm Chronic systemic diseases associated with increased risk of VFs	Age 70-79 Self-reported NVF HHL 3-6 cm Chronic systemic diseases associated with increased risk of VFs On pharmacologic androgen deprivation therapy or following orchiectomy

 

N/A = not applicable
BMD = bone mineral density
HHL = historical height loss
PHL = prospective height loss
VF = vertebral fracture
NVF = nonvertebral fracture

Figure 4. Risk factors for falls adapted from guidelines of the National Osteoporosis Foundation (6). The presence of any of these risk factors should trigger consideration of further evaluation and treatment to reduce the risk of falls and fall-related injuries.

Environmental risk factors
Lack of assistive devices in bathrooms
Loose throw rugs
Low level lighting
Obstacles in the walking path
Slippery outdoor conditions

Medical risk factors
Age
Anxiety and agitation
Arrhythmias
Dehydration
Depression
Female gender
Impaired transfer and mobility
Malnutrition
Medications causing oversedation (narcotic analgesics, anticonvulsants, psychotropics)
Orthostatic hypotension
Poor vision and use of bifocals
Previous fall
Reduced problem solving or mental acuity and diminished cognitive skills
Urgent urinary incontinence
Vitamin D insufficiency [serum 25-hydroxyvitamin D (25(OH)D) < 30 ng/ml (75
nmol/L)]

Neurological and musculoskeletal risk factors
Kyphosis
Poor balance
Reduced proprioception
Weak muscles

Other risk factors
            Fear of falling

Quality of DXA and VFA

DXA and VFA should be performed by well-trained and experienced staff operating an instrument that has been maintained and calibrated according to the manufacturer's standards. Precision assessment and LSC calculation by each DXA technologist are required in order to make quantitative comparisons of serial BMD measurements. The use of the correct scan mode and proper patient positioning is important for accurate BMD measurements and essential for serial comparisons of BMD. VFA should be done by a technologist properly trained in acquisition techniques and interpreted by a clinician familiar with methods of diagnosing vertebral fractures using this technology. Bone densitometry facilities should be supervised by a clinician who knows current methods for BMD measurement and fully understands the standards for quality control, interpretation, and reporting of the findings. Poor quality studies may result in inappropriate clinical decisions, generate unnecessary healthcare expenses, and be harmful to patients (5).

Technologies for assessment of skeletal health

DXA. Devices that measure or estimate BMD differ according to their clinical utility, cost, portability, and use of ionizing radiation (Table 1). DXA is the "gold standard" method for measuring bone density in clinical practice (15).  There is a strong correlation between mechanical strength and BMD measured by DXA biomechanical studies (16). In observational studies of untreated patients, there is a robust relationship between fracture risk and BMD measured by DXA (10). The WHO diagnostic classification of osteoporosis is based primarily on reference data obtained by DXA (3). Most randomized clinical trials showing reduction in fracture risk with pharmacological therapy have selected study according to BMD measured by DXA (17). There is a relationship between fracture risk reduction with drug therapy and increases in BMD measured by DXA (18). Accuracy and precision of DXA are excellent (19). Radiation exposure is with DXA is very low (20). BMD of the 33% (one-third) radius, measured either by a dedicated pDXA device or a central DXA instrument with appropriate software, may be used for diagnostic classification with the WHO criteria and to assess fracture risk, but is generally not clinically useful in monitoring the effects of treatment (21). DXA measures bone mineral content (BMC in grams [g]) and bone area (cm2), then calculates and areal BMD in g/cm2 and other parameters, such as the T-score and Z-score. DXA is used for diagnostic classification, assessment of fracture risk, and for monitoring changes in BMD over time.

Table 1. Devices for measuring or estimating bone mineral density (BMD).  Clinical applications of different technologies are listed with approximate comparison of associated radiation exposure and cost, with 0 = none, + = low, ++ moderate, +++ = highest.

            DXA = dual-energy X-ray absorptiometry
            pDXA = peripheral DXA
            QUS = quantitative ultrasound
            QCT = quantitative computed ultrasound
            pQCT = peripheral QCT
            SOS = speed of sound
            BUA = broadband ultrasound attenuation

 

* World Health Organization classification

**pDXA of the distal one-third radius (33% radius) may be used with the WHO classification

 

Quantitative ultrasound (QUS). QUS devices emit inaudible high frequency sound waves in the ultrasonic range, typically between 0.1 and 1.0 megahertz (MHz). The sound waves are produced and detected by means of high-efficiency piezoelectric transducers, which must have good acoustical contact with the skin over the bone being tested. Technical differences among QUS systems are great, with different instruments using variable frequencies, different transducer sizes, and sometimes measuring different regions of interest, even at the same skeletal site. The calcaneous is the skeletal site most often tested, although other bones, including the radius, tibia, and finger phalanges, can be used. Commercial QUS systems usually measure two parameters- the speed of sound (SOS) and broadband ultrasound attenuation (BUA). A proprietary value, such as the "quantitative ultrasound index" (QUI) with the Hologic Sahara' or "stiffness index" with the GE Healthcare Achilles Express, may be calculated from a combination of these measurements. SOS varies according to the type of bone, with a typical range of 3000-3600 meters per second (m/sec) with cortical bone and 1650-2300 m/sec for trabecular bone.(22) A higher bone density is associated with a higher SOS. BUA, reported as decibels per megahertz (dB/MHz), is a measurement of the loss of energy, or attenuation, of the sound wave as it passes through bone. As with SOS, a higher bone density is associated with a higher BUA. Values obtained from calculations using ultrasound parameters may be used to generate an estimated BMD and a T-score.  The T-score derived from a QUS measurement is not the same as a T-score from a DXA. QUS cannot be used for diagnostic classification and is not clinically useful to monitor the effects of therapy (23).            

Quantitative computed tomography (QCT) and peripheral QCT (pQCT). QCT and pQCT measure trabecular and cortical volumetric BMD at the axial skeleton and peripheral skeletal sites, respectively. QCT is a useful research tool to enhance understanding of the pathophysiology of osteoporosis and the mechanism of action of pharmacological agents used to treat osteoporosis. QCT predicts fracture risk, with the correlation varying according to skeletal site and bone compartment measured, type of fracture predicted, and population assessed (24). The ISCD Official Positions state that "spinal trabecular BMD as measured by QCT has at least the same ability to predict vertebral fractures as AP spinal BMD measured by central DXA in postmenopausal women with lack of sufficient evidence to support this position in men; pQCT of the forearm at the ultradistal radius predicts hip, but not spine, fragility fractures in postmenopausal women with lack of sufficient evidence to support this position in men (24). QCT is more expensive than DXA and QUS, uses higher levels of ionizing radiation, and cannot be used for diagnostic classification. T-scores by QCT are typically lower than with DXA (25), thereby overestimating the prevalence of osteoporosis when used incorrectly with the WHO diagnostic criteria.

Figure 5. National Osteoporosis Foundation indications for bone density testing (6)

- Women age 65 and older and men age 70 and older, regardless of clinical risk factors
- Younger postmenopausal women and men age 50 to 69 about whom you have concern based on their clinical risk factor profile
- Women in the menopausal transition if there is a specific risk factor associated with increased fracture risk such as low body weight, prior low-trauma fracture or high risk medication
- Adults who have a fracture after age 50
- Adults with a condition (e.g., rheumatoid arthritis) or taking a medication
           (e.g., glucocorticoids in a daily dose ≥ 5 mg prednisone or equivalent for ≥ 3 months) associated with low bone mass or bone loss
- Anyone being considered for pharmacologic therapy for osteoporosis
- Anyone being treated for osteoporosis, to monitor treatment effect
- Anyone not receiving therapy in whom evidence of bone loss would lead to treatment
- Postmenopausal women discontinuing estrogen should be considered for bone density testing
 

WHO fracture risk assessment tool (FRAX®)

The combination of BMD and CRFs predicts fracture risk better than BMD or CRFs alone (26;27)  This has been recognized for many years (28), yet until recently, CRFs have played a relatively minor role compared with BMD in the qualitative assessment of fracture risk (29). In order to assist physicians in making a more accurate quantitative assessment of fracture risk, the WHO, in cooperation with other scientific societies (e.g., ISCD, NOF), has developed FRAX, a computer-based algorithm that estimates the 10-year probability of hip fracture and major osteoporotic fracture (i.e., clinical spine, hip, proximal humerus, and distal forearm fracture). FRAX can be accessed online at http://www.shef.ac.uk/FRAX (Figure 6), on some software versions of DXA systems, and on some handheld computer devices. FRAX is based on analysis of data from 12 large prospective observational studies in about 60,000 untreated men and women in different world regions, having over 250,000 person-years of observation and more than 5,000 reported fractures reported.

Figure 6. FRAX online for US Caucasian patients. This example shows a 65 year-old woman who has no clinical risk factors for fracture and a femoral neck T-score of -2.3 with a Hologic instrument. The 10-year probability of major osteoporotic fracture is 10% and the 10-year probability of hip fracture is 2.2%. These levels do not meet the National Osteoporosis Foundation guidelines for initiation of pharmacological therapy in the US (6). Image reproduced with permission of the World Health Organization.

 

 

The input for FRAX is the patient's age, sex, height, weight, a "yes" or "no" response indicating the presence or absence for each of 7 CRFs: 1. previous 'spontaneous' or fragility fracture as an adult; 2. parent with hip fracture; 3. current tobacco smoking; 4. ever use of chronic glucocorticoids at least 5 mg prednisolone for at least 3 months; 5. confirmed rheumatoid arthritis; 6. secondary osteoporosis, such as type 1 diabetes, osteogenesis imperfecta in adults, untreated longstanding hypothyroidism and hypogonadism, or premature menopause (note: this is a "dummy" risk factor that has no effect on the fracture risk calculation unless no femoral neck BMD value is entered); 7. alcohol intake greater than 3 units per day, with a unit of alcohol defined as equivalent to a glass of beer, an ounce of spirits or a medium-sized glass of wine), and femoral neck BMD, if available. Since the introduction of FRAX, upgrades have been introduced to correct errors, enhance its usability, and incorporate new data that have become available.

Benefits of FRAX. The use of FRAX provides a quantitative estimation of fracture risk that is based on robust data in large populations of men and women with ethnic and geographic diversity. Expression of fracture risk as a probability provides greater clinical utility for than relative risk. which leads to an overestimation of risk when the comparator group is at low risk. When combined with cost-utility analysis, a fracture risk level at which it is cost-effective to treat may be derived. FRAX can be used to estimate fracture probability without femoral neck BMD, allowing it to be used when DXA in unavailable or inaccessible.  

Limitations of FRAX. To generate a valid FRAX output, the reponses to CRF questions must be correct; for example, an incorrect entry of self-reported rheumatoid arthritis or use of glucocorticoids could skew the results toward overestimation of fracture risk. FRAX may underestimate or overestimate fracture risk due to dichotomized  (yes or no) input for CRFs  that in reality are associated with a range of risk that varies according to dose or severity; for example, fracture risk may be underestimated when a patient is on high-dose glucocorticoid therapy or has had multiple recent fragility fractures, even when a "yes" response is entered for these CRFs. FRAX is validated only in untreated patients and may overestimate fracture risk when the patient is being treated; the NOF/ISCD guidance on FRAX suggests that "untreated" may be interpreted as never treated or if previously treated, no bisphosphonate for the past 2 years (unless it is an oral agent taken for less than 2 months); and no estrogen, raloxifene, calcitonin, or denosumab for the past 1 year (30).  In this context, calcium and vitamin D do not constitute treatment. FRAX in the US allows input for 4 ethnicities (Caucasian, black, Hispanic, Asia); it is not clear how to use FRAX for patients of other ethnicities or a mix of these ethnicities. Answering "yes" for the category of secondary osteoporosis has no effect on the fracture risk calculation as long as a value for femoral neck BMD is entered. The range of error for a fracture probability generated by FRAX is unknown, but may be substantial in some cases. Some important risk factors, such as falls and frailty, are not directly entered into FRAX, although they are indirectly included insofar as they are a component of aging. FRAX may underestimate fracture risk when the lumbar spine BMD is substantially lower than femoral neck BMD, as may occur in about 15% of patients (31).

Despite the numerous limitations of FRAX, it is a helpful clinical tool when used with a good understanding of factors that may result in underestimation or overestimation of fracture risk. FRAX may enhance discussion of risk with the patient and help to identify those who are at sufficiently high for fracture to benefit from therapy. 

Medical history

A thorough medical history may identify risk factors for osteoporosis and fractures, suggesting that a bone density test and/or further evaluation is indicated. For example, the NOF guide recommends bone density testing in postmenopausal women under age 65 and men age 50-69 for whom there is "concern based on their clinical risk profile" and for adults with a condition or taking a medication that is associated with low bone mass or bone loss (6). The medical history may also reveal symptoms of potentially correctable causes of skeletal fragility (e.g., gluten intolerance with celiac disease) or co-morbidities that could influence treatment decisions (e.g., esophageal stricture suggests that oral bisphosphonates should not be given). A history of falls is a predictor of future falls, with that risk potentially modifiable though appropriate interventions. Finally, some symptoms may trigger further evaluation for the presence of fractures (e.g., historical height loss or development of kyphotic posture suggests the possibility of vertebral fractures that may warrant spine imaging). Table 2 provides examples of helpful information that might be obtained from a thoughtful interactive discussion with the patient.

Table 2. Medical history for patients with osteoporosis. A thorough review of systems and history of relevant familial disorders, previous surgical procedures, medications, dietary supplements, food intolerances and lifestyle provide helpful information in the management of patients with osteoporosis. Such historical information may play a role in determining who should have a bone density test, assessing fracture risk, providing input for the World Health Organization fracture risk assessment tool (FRAX®), evaluating for secondary causes of osteoporosis, selecting the most appropriate treatment to reduce fracture risk, and finding factors contributing to suboptimal response to therapy. Listed here are key components of the skeletal health history and examples of the potential impact on patient care.

 

 

Physical exam

Findings of importance on the physical exam of a patient with osteoporosis may be the sequelae of old fractures (e.g., kyphosis due to old vertebral fractures), a consequence of a recent fracture (e.g., localized vertebral spinous process tenderness with a new vertebral fracture), or abnormalities suggestive of a secondary cause of osteoporosis (e.g., thyromegaly with thyrotoxicosis). An accurate measurement of height with a wall-mounted stadiometer is a helpful office tool for evaluating patients at risk for fracture. A height loss of 1.5 inches (4.0 cm) or more compared to the historical maximum (32;33) or a loss of 0.75 inches (2.0 cm) or more compared to a previous measured height (34) suggests a high likelihood of vertebral fracture. Body weight measurement is part of the osteoporosis evaluation because low body weight (less than 127 lbs) (35), low BMI (20 or less) (36), and weight loss of 5% or more (37) are associated with increased risk of fracture. Localized tenderness of the spine, kyphosis, or diminished distance between the lower ribs and the pelvic brim may be the result of one or more vertebral fractures. Abnormalities of gait, posture, balance, muscle strength, or the presence of postural hypotension or impaired level of consciousness may be associated with increased risk of falling. Bone tenderness may be the caused by osteomalacia. Atrophic testicles suggest hypogonadism. Patients should be observed for stigmata of hyperthyroidism or Cushing's syndrome. Blue sclera, hearing loss, and yellow-brown teeth are suggestive of osteogenesis imperfecta. Joint hypermobility and skin fragility could be due to Ehlers-Danlos syndrome. Urticaria pigmentosa may occur with systemic mastocytosis. Table 3 shows examples of abnormal physical exam findings with osteoporosis.

Table 3. Focused physical examination in patient with osteoporosis. This table provides examples of findings on physical exam that may be helpful in the evaluation of skeletal health. It is not intended to show all findings of importance.

 

Evaluation for secondary causes of osteoporosis

The possibility of previously unrecognized causes of skeletal fragility should be considered in every patient with osteoporosis. After an initial medical history is taken and physical exam is performed, appropriate laboratory testing and imaging may provide information that is critical for ongoing patient care. Osteoporosis is commonly divided into two categories according to etiology. "Primary osteoporosis" is due to time-appropriate postmenopausal estrogen deficiency (type I osteoporosis, preferentially involving trabecular bone loss) or to aging in men and women (type II osteoporosis, with a combination of trabecular and cortical bone loss). 'secondary osteoporosis' is caused by other conditions, diseases, or medications, with or without the presence of primary osteoporosis.

The reported prevalence of secondary osteoporosis varies depending on the study population, the extent of the medical evaluation, and definitions for laboratory abnormalities. It is likely that many or most patients with primary osteoporosis have clinically significant contributing factors that may influence patient management. In a study of North American women receiving osteoporosis therapy, it was found that 52% had vitamin D inadequacy, defined as serum 25-hydroxyvitamin D (25-OH-D) levels less than 30 ng/ml (38). In another study of patients referred to an osteoporosis clinic, over 60% were found to have elements of secondary osteoporosis when vitamin D deficiency was very conservatively defined as serum 25-OH-D level less than 12.5 ng/ml (39;40). In the same study, the number of patients with secondary osteoporosis was much higher when vitamin D inadequacy was more appropriately defined as serum 25-OH-D less than 33 ng/ml (41;42).

It has been proposed by some that a bone density that is less than expected compared to an age- and sex-matched population, as represented by a low Z-score (e.g., less than -2.0), suggests a high likelihood of secondary osteoporosis and should be one of the triggers for further investigation (43;44). While there may be some merit to this concept, there are few if any studies validating the use of a Z-score cutoff for this purpose. Since secondary osteoporosis is common, a more effective strategy is to screen all patients with osteoporosis for contributing factors (45). The results of a metabolic evaluation may identify previously unrecognized diseases and conditions that require treatment in addition to, or instead of, standard osteoporosis pharmacological therapy.

Depending on the patient population being studied, different causes of secondary osteoporosis may predominate. Calcium deficiency, vitamin D deficiency, and sedentary lifestyle are common contributing factors for all patients. In women referred to an osteoporosis clinic with previously recognized medications or diseases contributing to osteoporosis, the most common were history of glucocorticoid use (36%), premature ovarian failure (21%), history of unintentional weight loss (10%), history of alcoholism (10%), and history of liver disease (10%) (39). When patients without previously recognized contributing factors were evaluated at the same specialty clinic, most were 55% were found to have vitamin D deficiency or insufficiency (serum 25-OH-D less than 33 ng/ml) (42), while 10% had hypercalciuria, 8% had malabsorption, and 7% had primary or secondary hyperparathyroidism (39). In men, the most common secondary causes of osteoporosis are long-term glucocorticoid use, hypogonadism, and alcoholism (46;47). The increasing use of aromatase inhibitor therapy for breast cancer in women (48) and androgen deprivation therapy for prostate cancer in men (49) is now recognized as an important factor in the development of osteoporosis in these patients. Other common causes for low BMD and fractures include multiple myeloma (50), gastric bypass surgery (51) and gastric resection (52). Treatable but easily missed secondary causes of osteoporosis include asymptomatic hyperparathyroidism (53), apathetic hyperthyroidism (54), mild Cushing's disease (55), and malabsorption due to unrecognized celiac disease (56). Table 5 lists some of the causes of low BMD by category.

Table 5. Many causes of low bone mineral density.

 

A variety of testing strategies have been proposed as screening for all patients with osteoporosis (39;42;45;57;58). There is general agreement that a minimal cost-effective work-up for all patients consists of a fasting serum calcium, 24-hour urinary calcium, and serum 25-OH-D. Other helpful studies, many of which are done for routine health maintenance independently of skeletal issues, include a creatinine with calculated or measured creatinine clearance, complete blood count (CBC), phosphorous, alkaline phosphatase, thyroid stimulating hormone (TSH), and hepatic enzymes. Other laboratory tests may be indicated according to the patient's clinical profile and the practice setting. A summary of useful common and uncommon laboratory studies with comments on their possible skeletal significance is provided below.

Basic blood tests

CBC- Anemia may be seen in patients with multiple myeloma or malnutrition.

Sedimentation rate- May be elevated with multiple myeloma.

Calcium- Among the many causes of hypercalcemia are primary and secondary hyperparathyroidism, hyperthyroidism, renal failure, vitamin D intoxication, and Paget's disease. Hypocalcemia may be seen with vitamin D deficiency and hyperphosphatemia.

Phosphorus- Hyperphosphatemia may occur with hypoparathyroidism, renal failure, and possibly with bisphosphonate therapy. Hypophosphatemia may be seen with primary or secondary hyperparathyroidism, vitamin D deficiency, and oncogenic osteomalacia.

Alkaline phosphatase- High values can be seen with healing fractures, osteomalacia, and Paget's disease, as well as occurring normally in growing children. Low values occur with hypophosphatasia, a rare pediatric disorder that causes impaired mineralization of bone and dental tissue.

Vitamin D- The test that best reflects vitamin D stores is the serum 25-OH-D. While there is no consensus on the optimal range of serum 25-OH-D, a reasonable target for good skeletal health is approximately 30-60 ng/ml. This is likely to maximize intestinal absorption of calcium and minimize serum PTH levels. Interpretation of serum 25-OH-D levels is confounded by assay variability (59). Serum 1,25-(OH)2-D  is usually not helpful in the evaluation of osteoporosis patients, unless there are concerns regarding renal conversion of 25-OH-D to 1,25(OH)2-D. Deficiency or insufficiency of vitamin D is very common and play a role in the pathogenesis of osteoporosis and osteomalacia.

Creatinine- Chronic kidney disease may cause an elevated creatinine level and renal osteodystrophy. Elderly patients with small muscle mass may have impaired renal function with a "normal" serum creatinine. An estimated creatinine clearance can be calculated using one of many formulae, such as that of Cockcroft and Gault (60). Impaired renal function not only has adverse skeletal effects but also raises considerations regarding the type and dose of pharmacologic agents used.

TSH- Hyperthyroidism from any cause, including excess thyroid replacement, can usually be recognized by a low TSH. High bone turnover associated hyperthyroidism is associated with loss of bone mass. 

Liver enzymes- Abnormalities may be caused by chronic liver disease, which is a risk factor for osteoporosis.

Basic urine tests

Urinalysis. Proteinuria may occur with multiple myeloma or chronic kidney disease. Abnormal cells may suggest kidney disease.

24-hour urine for calcium- A well-collected 24-hour urine for calcium is a helpful screening test for identifying patients with common disorders of calcium metabolism. The "normal" range of urinary calcium is not well established, and varies according to many dietary factors and estrogen status in women (61;62). As a "rule of thumb," urinary calcium may be considered elevated when it is greater than 250 mg per 24 hours in women; greater than 300 mg per 24 hours in men; or greater than 4 mg/kg body weight per 24 hours in either sex. It has been proposed that hypercalciuria can be easily classified as "renal" (renal calcium leak), "resorptive" (excess skeletal loss of calcium) or "absorptive" (increased intestinal absorption of calcium) (63). However, in clinical practice, these distinctions are not so easily established. Idiopathic hypercalciuria, perhaps the most common type of hypercalciuria (64), may be diagnosed if there are no underlying medical disorders (e.g., hyperparathyroidism, vitamin D toxicity, Paget's disease of bone, multiple myeloma, sarcoidosis) and no obvious dietary excesses (e.g., calcium, sodium, protein, carbohydrates, alcohol) or deficiencies (e.g., phosphate, potassium) that are associated with hypercalciuria (62). In the absence of dietary calcium deficiency, vitamin D deficiency, malabsorption, liver disease, or chronic renal failure, low urinary calcium (less than 50 mg per 24 hours in women or men) is suggestive of calcium malabsorption and warrants further investigation. Celiac disease is a common (65) cause of asymptomatic malabsorption  in osteoporosis that is treatable with a gluten-free diet (56).

Additional Studies in Selected Patients

Celiac antibodies- Antiendomysial antibody and tissue transglutaminase antibody are currently the serological markers of choice, with a higher sensitivity and specificity than antigliadin antibody and antireticulin antibody (66). If a serological marker is abnormal, or if there is a high clinical suspicion for celiac disease, the patient should be referred for endoscopy and small bowel biopsy.

Intact PTH- This may be elevated in patients with primary hyperparathyroidism, vitamin D deficiency, or renal failure.

Serum and urine protein electrophoresis- These are helpful tests to screen for possible multiple myeloma. If an M-component is identified, referral for bone marrow aspiration may be indicated.

Dexamethasone suppression test or 24-hour urinary free cortisol- This is helpful to evaluate patients with suspected Cushing's syndrome.

Serum total or free testosterone level- A useful test for all men with osteoporosis.

Serum homocysteine- Elevated circulating homocysteine levels are associated with increased risk of fracture (67;68). It is not know whether reduction of homocysteine levels by increasing dietary intake of folic acid and vitamins B6 and B12 reduces the risk of fracture.

Serum tryptase and 24-hour urine for N-methylhistamine- Systemic mastocytosis is a rare cause of osteoporosis that can be diagnosed by a biopsy of typical skin lesions of urticaria pigmentosa, when present. Patients with systemic mastocytosis may sometimes present with osteoporosis and no other manifestations of the disease (69;70) . When this disorder is suspected but skin lesions are not present, the finding of an elevated serum typtase and/or urinary N-methyl histamine can be helpful, especially during or soon after a symptomatic episode of histamine release. However, normal values do not exclude the diagnosis. Bone marrow aspiration or biopsy, or non-decalcified double tetracycline labeled transiliac bone biopsy, may be necessary to confirm the diagnosis.

Serum bicarbonate- Renal tubular acidosis (RTA) has been associated with osteoporosis (71). With distal (type I) RTA, the serum bicarbonate is usually less than 15 mmol/l with a urine pH greater than 5.5.

Bone turnover markers

Bone turnover markers (BTMs) are noninvasive laboratory tests of serum and urine that are readily available in clinical practice. While BTMs cannot be used to diagnose osteoporosis or determine the cause to osteoporosis, they have been very helpful in the research to understand the pathophysiology of osteoporosis and other skeletal diseases and the mechanism of action of interventions used in the treatment of osteoporosis. In clinical practice, BTMs offer the potential of predicting fracture risk independently of BMD and may be useful in monitoring the metabolic effects of therapy (72). Drugs that are approved for the management of osteoporosis modulate bone remodeling in ways that are reflected by changes in BTMs. A decrease in BTMs with antiresorptive therapy is predictive of a subsequent increase in BMD (73) and reduction in fracture risk (74-77). The magnitude of BTM decrease with antiresorptive therapy is significantly associated with the level of fracture risk reduction, although the proportion of treatment effect due to the reduction in BTMs appears to vary according to the type of drug used (78). With teriparatide, a bone anabolic agent, an early increase in BTM levels is predictive of a subsequent increase in BMD (79).

Markers of bone resorption are mostly fragments of type I collagen, the main component of the organic bone matrix, that are released during osteoclastic bone resorption. These are measured in the serum or urine, with those available for clinical use including N-telopeptide of type I collagen (NTX), C-telopeptide of type I collagen (CTX), deoxypyridinoline (DPD), and pyridinoline (PYD). Bone formation markers are proteins secreted by osteoblasts or byproducts of type I collagen production by osteoblasts. They are measured in the serum and include bone specific alkaline phosphatase (BSAP), N-terminal propeptide of type I collagen (P1NP), and osteocalcin.

Clinical use of BTMs requires knowledge of their limitations as well as benefits. BTMs are subject to pre-analytical (biological) and analytical variability.  Uncontrollable sources of pre-analytical variability include age, sex, menopausal status, pregnancy, lactation, fractures, co-existing diseases (e.g., diabetes mellitus, impaired renal function, and liver disease), drugs (e.g., glucocorticoids, anticonvulsants, and gonadotropin hormone releasing agonists)  and immobility (80). Controllable pre-analytical sources of variability include time of day (circadian variability), fasting status, and exercise (80). Analytical sources of variability include specimen processing (e.g., collection, handling, and storage) (81). Between-laboratory variability  may be large (reported to be as much as a 7.3-fold difference), casting doubt on the validity of comparing specimens sent to different labs  (82). Reference ranges for BTMs are not well established and may vary according to the population tested, the type of BTM, and the circumstances under which it is collected and processed.

In order to compare BTMs measurements longitudinally, it would be ideal to know the LSC and use this in a manner similar to what should be (but is probably not) common practice with DXA. However, the standards for calculating an LSC for a BTM are not as clear as with DXA, and the opportunity to do precision assessment for a BTM may not present itself. The NOF recommends calculating the LSC with a 95% level of confidence for each BTM used by multiplying the laboratory-provided precision error by 2.77 (6). The NOF also recommends that specimens be obtained in the early morning following an overnight fast to reduce biological variability, with serial measurements to be obtained at the same time of day and ideally during the same season of the year. The Belgian Bone Club suggests using an estimated LSC of assuming an LSC of about 30% for serum BTMs and about 50-60% for urine BTMs (72).  While the LSC for BTMs is almost always greater than for DXA, the magnitude of likely change (83)  is greater than DXA, with the 'signal to noise ratio' that may be as good or even better than DXA.

Evidence-based guidelines for the clinical use of BTMs have been developed by organizations that include the NOF (6), Belgian Bone Club (72), and the Japan Osteoporosis Society (84). The NOF guidelines state that BTMs "may predict bone loss and, when repeated after 3-6 months of treatment with FDA approved antiresorptive therapies, may be predictive of fracture risk reduction." The Belgian Bone Club suggests that "early changes in BTM can be used to measure the clinical efficacy of an antiresorptive treatment and to reinforce patient compliance," with goal of decreasing the BTM to the premenopausal range or at least achieving a decrease as great as the LSC. The Japanese guidelines indicate that "the argument for measuring bone turnover markers to evaluate the therapeutic effects of bone antiresorptive medications can be justified," but go on to state that there is insufficient evidence for their use with medications having other mechanisms of action (84).

A significant change of a BTM level in the appropriate direction following therapy is evidence that the patient is taking the drug regularly, taking it correctly, and that it is being absorbed and having the expected effect in modulating bone remodeling. Failure to achieve such a change in the BTM level is cause for concern and suggests that evaluation and possibly a reconsideration of treatment should be considered (85). The use of BTMs allows assessment of drug effect sooner than with DXA, so that evaluation and corrective action, if needed, can be taken early in the course of therapy rather than later. Monitoring BTMs, especially in association with regular contact by a healthcare provider, may improve persistence with therapy (86). Despite the well-described limitations of BTMs (87) , there is emerging support for their use in clinical practice, particularly in the assessment of response to therapy (88;89).  Clinicians who are familiar with the benefits and limitations of BTMs may find them a helpful tool, in association with BMD testing, for managing patients with osteoporosis. 

Imaging studies

Standard X-rays are used to diagnose fractures of all types and may sometimes suggest secondary causes of osteoporosis. Pseudofractures (Looser's zones), radiolucent lines running perpendicular to the bone cortex, may be seen in patients with osteomalacia. These probably represent stress fractures that have healed with poorly mineralized osteoid. Punctate radiolucencies may be seen in bone X-rays of patients with systemic mastocytosis. Primary hyperparathyroidism may cause bone cysts, subperiosteal bone resorption, brown tumors, and demineralization ('salt and pepper' pattern) of the skull. MRI, CT scanning, or nuclear imaging may be used to detect stress fractures not visible on X-ray. MRI of the spine is commonly used prior to vertebroplasty or kyphoplasty to determine the age of the fracture, the likelihood of the fracture being from causes other than osteoporosis, and whether there is retropulsion of bony fragments than could impair neurological function.

Bone biopsy

Non-decalcified double tetracycline labeled iliac crest bone biopsy is rarely used in clinical practice, but may be helpful with difficult diagnostic problems. In the evaluation of renal osteodystrophy, a bone biopsy can distinguish between high turnover and low turnover bone disease, and possibly be an aid in the selection of therapy. With infiltrative disorders of bone, such as systemic mastocytosis, a bone biopsy or bone marrow aspiration may sometimes be the only way to make the diagnosis. In patients who are not responding to therapy as expected, or in patients with unusual presentations of osteoporosis, a bone biopsy may be indicated. Bone biopsies are required by the FDA for safety monitoring in clinical trials of osteoporosis drugs.

Summary

Osteoporosis is a common skeletal disease with serious clinical consequences. Effective management of skeletal health includes appropriate selection of patients for bone density testing and assessment of risk factors for fracture. Prior to treatment, and when response to treatment is suboptimal, patients should be evaluated for secondary causes of osteoporosis. All reversible factors should be corrected and treatment should be individualized based on the clinical circumstances.