All Patients with Inflammatory Bowel Disease Should Have Bone Density Assessment: Con

A detailed analysis of the cost-effectiveness of diagnostic testing and treatment for osteoporosis in menopausal women undertaken by the National Osteoporosis Foundation (NOF) concluded that routine bone density testing was justified for all women after age 65 and for women after age 50 with one additional major risk factor (personal history of fracture, family history of osteoporosis, low body weight, or smoking) (1). There is also compelling evidence to consider all individuals on prolonged (more than 3 months per year) supraphysiologic doses of glucocorticosteroids as being at high risk for osteoporosis (2,3) and potentially benefiting from evidence-based therapies (4,–6). However, the majority of inflammatory bowel disease (IBD) patients are neither menopausal nor steroid dependent. Uniform bone density screening for this lower-risk group is not justified from the available evidence and is unlikely to prove cost effective. The strongest evidence to justify a screening program is provided by a randomized controlled trial in which clinically relevant end points are improved for screened versus unscreened cohorts (7,8). This is the level of evidence being used to promote regular fecal occult blood testing to detect asymptomatic colorectal cancer (9,–11). Of course, no such trials have been conducted with bone density screening in IBD (nor in menopausal women for that matter). Alternatively, evidence for or against a screening program can be derived by critically considering the elements of a cost-effective screening program such as those proposed by the World Health Organization for colorectal cancer testing: 1) the disease is common and associated with serious morbidity or mortalities; 2) screening tests are accurate in detecting early-stage disease, are acceptable to patients, and are feasible to general clinical practice; 3) treatment after the detection through screening has been shown to improve prognosis relative to treatment after usual diagnosis; and 4) evidence exists that the potential benefits outweigh the potential harms of screening (12). Applying this standard to bone density screening in IBD leads to the inevitable conclusion that the current level of understanding of diagnosis and treatment of osteoporosis in IBD is insufficient for recommending screening bone density. In the absence of fracture, reduced bone mass is asymptomatic. Reduced bone density is but one of many independent risk factors for fracture (13,14), and prevalence studies using low bone density in IBD as a surrogate for fracture rates and therapeutic antifracture effect cannot substitute for direct fracture data (15). Currently only one study has measured the fracture burden in IBD. Bernstein et al. (16) conducted a population-based matched cohort study in which 6,027 IBD patients were compared with 60,270 patients without IBD matched for age, gender, residential area, and year of observation (1984–1997). This study confirmed a statistically significant but relatively modest increased incidence of fractures for the spine (incidence rate ratio [IRR], 1.74 [CI, 1.34–2.24]), hip (IRR, 1.59 [CI, 1.27–2.00]), wrist/forearm (IRR, 1.33 [CI, 1.11–1.58]), and rib (IRR, 1.25 [CI, 1.02–1.52]). The combined fracture rate (IRR, 1.41 [CI, 1.27–1.56]) was similar for diagnostic subgroups and gender. Although the relative risk of fracture is similar for the different age strata, the absolute fracture incidence increases dramatically with advancing age. Combined fractures were more than eight-fold more common in the oldest cohort (versus the youngest). Hip fractures, widely accepted as the fracture site with the greatest effect on morbidity and mortality, was almost completely restricted to the oldest cohort (77 of 85; 91%). It is particularly noteworthy that the magnitude of IBD as a risk factor for fractures in this study was no greater than other so-called major fracture risk factors for menopausal women (1). This suggests that IBD could serve a similar role of identifying menopausal women that may benefit from bone density testing after age 50, but does not make a compelling case for screening younger women or men. Bone density measurement has a well-deserved central role in assessing osteoporosis in menopausal women, steroid-treated patients, and a large number of other conditions (17). Currently, dual energy X-ray absorptiometry (DEXA) is the preferred technique since it can noninvasively determine bone density measurement of central sites (spine, hip, total body) with high precision, low cost, and negligible radiation. Many prospective observational studies have confirmed that reduced bone density predicts fractures as well as hypertension predicts stroke or hypercholesterolemia predicts coronary artery disease (18). Indeed, the likelihood of hip fracture increases 2.6-fold for every standard deviation reduction in bone density, and this is superior to hypertension for predicting stroke or hypercholesterolemia for predicting cardiovascular disease (16). As impressive as this sounds, it is important to underscore the limited diagnostic performance of such information, as there is wide overlap in the bone densities of patients who develop a fracture and those who do not (19,20). In average menopausal women with a lifetime incidence for hip fracture of 15%, Marshall et al. (16) estimated that reduced hip bone density by DEXA (cut-point 1 SD below age adjusted mean) had only a 37% sensitivity and a 36% positive predictive value. Testing applied to a younger population with lower fracture rates would be anticipated to have even worse performance. The prevalence of low bone density and of accelerated bone loss is uncertain. Osteopenia associated with Crohn’s disease has a reported prevalence as high as 41% (21). However, other groups have failed to find significant reductions in IBD (22,23), including in Crohn’s disease (24). There have been even fewer longitudinal studies of bone density in IBD. Clements et al. (25) performed measurements of the distal radius in 39 IBD patients over a mean interval of 7.9 years. Negligible change was seen in male patients (annual rate of bone loss −0.07%), and even in female patients bone loss was largely confined to the postmenopausal subgroup (−1.16%). Staun et al. (27) studied 108 patients (mostly female) with Crohn’s disease monitored over a mean 5.5 years with bone density of the spine and femoral neck. The hip decreased more rapidly than expected [annual change −2.2% with the colon preserved, −1.2% with a resected colon versus “normal” rates of decrease in hip bone mass of −1.0% per year from age 30–94 (28)] but confidence intervals were wide and included the possibility of no change. There was no significant change in the spine. Regular steroid intake was a risk factor for a decrease in femoral neck bone density in this study, but there was no effect of IBD subtype, sex, age, or biochemical bone markers. A more recent study in 80 IBD patients assessed DEXA of the spine, hip, and total body after a mean interval of 1.6 years. No significant change was seen in the spine or femoral neck, and the change in the total body (annualized change −0.52%) was similar to normal age-related loss (26). Together these longitudinal studies do not support the presence of accelerated bone loss in IBD or the need for close monitoring with bone density. Not only have there been no studies to prove that early diagnosis and treatment of asymptomatic IBD-related osteoporosis is superior to delayed treatment, but there have actually been no randomized controlled trials showing that interventions that prevent fractures in menopausal women will prevent fractures in this disorder. Although the potent bisphosphonate, alendronate, appears to be well tolerated in Crohn’s disease and can increase bone density in the lumbar spine (+4.6% alendronate group versus −0.9% placebo, p < 0.01) and hip (+3.3% alendronate group versus +0.7% placebo group, p = 0.08) after 12 months (29), this does not necessarily indicate antifracture effect (13). For example, bisphosphonate therapy leads to a significant increase in spine and hip bone density in menopausal women independent of the baseline bone density, but antifracture effect is only evident when the bone density is in the osteoporotic range (more than 2.5 standard deviations below average for a healthy young reference population) (30). Even in the treatment of menopausal osteoporosis, large clinical trials indicate that the largest treatment benefit occurs in individuals with vertebral fragility fractures, and that treating low bone density in the absence of fragility fractures is of uncertain benefit (31). To date there have not been any decision analysis or cost-effectiveness studies of bone density screening for IBD. The negligible radiation dose from bone densitometry is unlikely to have any harmful effect, and the cost is also relatively low. The same cannot be said for antiresorptive treatment, however. Currently available medications produce only a very modest increment in bone mass and, therefore, treatment initiated when bone density falls below a certain threshold is likely to be required lifelong. At hundreds of dollars per year for the newer bisphosphonates, the total cost of initiating treatment in a young patient is substantial. When considered against the low risk of fracture in young IBD patients, treatment is unlikely to be cost effective. The newer bisphosphonates are also not without significant complications (32,33). The safety of these agents in IBD patients has not been clearly established, and there is concern with their use in young patients due to the prolonged residence time in the skeleton. Hormonal therapy with estrogen or the new selective estrogen receptor modulators (SERM) is not applicable to men or menstruating females, and is associated with a significantly increased rate of thromboembolic events (34,–36). IBD patients are already at high risk for thromboembolic disease (37,38) and an additive (or greater) risk is a distinct possibility. Given these major uncertainties in the benefit and safety of pharmacotherapies for IBD-related osteoporosis, even greater caution must be exercised before advocating widespread diagnostic screening. The temptation to extrapolate treatments and treatment thresholds from menopausal women to IBD patients is enticing but may be deeply flawed. Several studies indicate that bone density in IBD correlates with body weight and previous bowel resections (21,22,39,–42), and suggests that nutritional deficiencies and malabsorption may play a central role. An antiresorptive agent may be ineffective (and actually harmful) in the presence of nutritional osteomalacia, a diagnosis that is too frequently overlooked (43). A closer parallel to IBD-related bone disease may be found in anorexia nervosa (44). The pathogenesis of bone loss in anorexia nervosa is not completely understood and may result from a number of mechanisms, including estrogen deficiency, inadequate vitamin and calcium intake, and nutritional effects on bone formation (45). As in IBD, weight appears to be a strong predictor of bone mass (46,47). Reduced lean body weight is associated with bone loss (42) and weight gain is the most important predictor of the recovery from osteoporosis (48). Osteopenia of this disorder appears to be a low turnover state characterized by increased bone resorption without concomitant bone formation (49), and is reversed by addressing the malnutrition (50). This pattern differs from osteopenia in menopausal women and may explain why estrogen replacement does not prevent or correct the bone disease (41,43,51). Treating a nutritional disorder with an antiresorptive agent and hoping for antifracture benefit may be overly simplistic. An important confounding factor is the role of bone size in fracture susceptibility and bone density measurement. Bone density measurement with DEXA relies upon a two-dimensional projection of bone to calculate areal bone density (in units of g/cm2). Several investigators have demonstrated that projected bone area does not adequately adjust for skeletal size (52). Larger bones, which are also larger in the unmeasured third dimension, will therefore give a higher areal bone density measurement than if bone density is corrected for skeletal volume (usually expressed in mg/cm3). This affects the interpretation of results for children, adolescents, different genders, and ethnicity (53,–57). Much of the reduction in bone density seen with anorexia nervosa is actually a function of this volumetric artifact (58). Growth failure is a common and serious sequela of childhood IBD, affecting 19–35% of patients and resulting in permanent deficits in adult height (59,60). Although no studies to date have directly assessed bone size in IBD, growth failure arising from disease activity during the time of peak adolescent skeletal development would be expected to result in reduced skeletal size. Reduced areal bone density could then be as much a function of bone size as of abnormal bone metabolism. If it is inappropriate to test all IBD patients, is it possible to identify subgroups at relatively greater risk? Certainly overt fractures and older age (coinciding with menopause in women) are two such markers (14). Cumulative cortical steroid exposure has also emerged as an important variable and may account for much of the observed difference between Crohn’s disease and ulcerative colitis (35,38,39,61,62). I would suggest that published guidelines for bone density testing in menopausal women and anyone with overt osteoporotic fractures or receiving prolonged steroid therapy are useful guides to testing in IBD patients. This offers a rational response to the relatively modest excess fracture rates seen in these disorders. As always, individual considerations combined with clinical judgement need to be factored into the decision-making process. How does this relate to young IBD patients concerned about their future risk of osteoporotic fractures? The future is bright and anabolic agents are poised to make major therapeutic contributions with the real potential of reversing established osteoporosis (63). This alone justifies a cautious approach in the care of young IBD patients who will not be entering their age of major fracture risk for decades. Clearly much more research is needed on the pathophysiology, diagnosis, and treatment of this disorder before broad screening recommendations can be endorsed.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.