Assessment of bone mineral density in adults and children with Marfan syndrome
- First Online:
- Cite this article as:
- Giampietro, P.F., Peterson, M., Schneider, R. et al. Osteoporos Int (2003) 14: 559. doi:10.1007/s00198-003-1433-0
Recent studies indicate that decreased bone mineral density (BMD) occurs in the spine, femoral necks and greater trochanters of some adults and children with Marfan syndrome. Because there is uncertainty regarding the BMD status of patients with Marfan syndrome, we undertook an analysis of BMD in both adults and children with Marfan syndrome. Dual energy X-ray absorptiometry analysis was performed on a convenience sample of 51 patients (30 adults and 21 children) with diagnosed Marfan syndrome from 1993 to 2000. T-Scores (i.e. the number of standard deviations above or below the average normal peak bone density) were determined for comparison of adults. Mean±SD of individual BMD values were used for comparison of the data of children. Compared to standard values obtained from normal adult patients, adult males with Marfan syndrome demonstrated significantly reduced femoral neck BMD with an average T-score of -1.54 (P<0.001), diagnostic of osteopenia. Although osteopenia and osteoporosis were observed in several middle aged and pre- and postmenopausal women, the average T-score value for adult females and children were within normal limits. The etiology and full significance of decreased BMD in adult male patients with Marfan syndrome remain uncertain at the present time. Our results lead us to question the value of aggressive BMD evaluations by DXA in these patients, particularly prior to reaching mid-age. Further investigations will be required to shed insights into the natural history of BMD in adults and children with Marfan syndrome. Any application of bone mineral replacement therapy such as bisphosphonate, selective estrogen receptor modulators, hormone replacement therapy and vitamin D in these patients may be premature based on the existing evidence.
KeywordsAdult malesFemoral neckMarfan syndromeOsteopenia
Marfan syndrome is an autosomal dominant disorder of connective tissue with an estimated incidence in the general population of 1/10,000 . Approximately 15% of cases occur in the absence of a family history, representing new mutations. The major clinical findings of Marfan syndrome involve three major organ systems, including the ocular, skeletal and cardiovascular systems. Skeletal findings include increased height and arm span, anterior chest deformity (pectus excavatum or carinatum), mild-to-moderate joint laxity, vertebral column deformity (scoliosis and thoracic lordosis), a narrow, highly arched palate and a narrow jaw, resulting in crowded dentition, pes planus, joint hypermobility and arachnodactyly .
Advances in the treatment of the cardiovascular complications of Marfan syndrome, including the use of β- and calcium channel blocker medications and cardiothoracic surgery, enable patients with Marfan syndrome to have near normal life expectancies . The prolonged survival of patients with Marfan syndrome has also unveiled a number of non-cardiac health issues associated with improved survival. Among these, long-term musculoskeletal complications may significantly impact patients with Marfan syndrome. Winter  first suggested that the bone quality of patients with Marfan syndrome is poor. Since then, a number of reports have described decreased bone mineral density (BMD) in Marfan syndrome [5,6,7,8,9,10].
In this paper, we report the BMD results in 51 patients evaluated from 1993 to 2000. Our goal was to quantitatively determine whether patients with Marfan syndrome have osteoporosis and/or osteopenia.
Materials and methods
Institutional Review Board approval was obtained. A convenience sample of adults (18–65 years) and children (6–18 years) who met the stringent diagnostic criteria for Marfan syndrome  were enrolled for a multidisciplinary research study on different aspects of Marfan syndrome during the time period between 1993 and 2000.
Complete medical and family histories were obtained from each patient and/or their parent(s). Patients were also asked about their history of previous fractures, decrease in height, smoking, alcohol usage, menarche, menopause, estrogen, steroid usage and family history of osteoporosis. Each patient had a nutritional assessment including evaluation of growth, caloric intake, calcium and vitamin D intake, and diet history based on the results of a food frequency questionnaire .
An exercise history was also obtained. For purposes of statistical analysis, "yes" means that patients reported consistent participation in any of the following activities: aerobics, swimming, karate, treadmill, basketball, soccer, bowling, jogging, golf, skiing, baseball, basketball, special physical education classes, martial arts. "No" means that patients did not participate in a regular exercise program.
Each patient had a complete physical exam. The pubertal development of all children with Marfan syndrome was assessed by physical examination and scored by the criteria of Tanner . Bone density was measured by means of dual-energy X-ray absorptiometry (DXA) using the Lunar DPXL (Lunar Corp., Madison, Wisc., USA) . Spine, proximal femur, and total body BMD data were determined using the Lunar software program. The anatomic site T-scores (i.e. the number of standard deviations above or below the average normal peak bone density) were determined by comparison of the patient to controls of the same sex and ethnicity provided by the Lunar software program. Spine BMD studies were not performed on patients with implanted spinal instrumentation for scoliosis repair surgery.
The following categories for BMD have been established by the World Health Organization: normal bone density, BMD within 1 SD of peak bone mass; osteopenia, BMD 1–2.5 SD below peak bone mass; and osteoporosis, BMD >2.5 SD below peak bone mass.
Reporting T-scores in children is not appropriate . Therefore, the mean and SD of individual BMD values only are reported for children.
The Student t-test was used to compare T-scores between patients with Marfan syndrome and control subjects.
Age height and weight breakdown of Marfan patients enrolled in study. SD standard deviation
Mean age (years)
Mean height (cm)
Mean weight (kg)
Average bone mineral density (BMD) and T-scores obtained on dual energy X-ray absorptiometry analysis. SD standard deviation
No statistically significant differences were seen in the BMD T-scores in adult subjects who reported participating in a regular program of exercise, as compared to patients who did not report participation in a regular program. This suggests that the low BMD values observed in our patient population are not due to restriction in physical activity.
Both osteoporosis and osteopenia were observed in our study population. There was no significant correlation between decreased BMD and smoking, age of menarche, and alcohol usage. Calcium intake was normal in all study patients.
Fractures were reported in 16 patients enrolled in the study. A history of trauma was obtained in 13 of these patients. Ten of 16 patients were osteopenic or osteoporotic in one of the three measured BMD parameters.
From our results it is evident that any concomitant effects of Marfan syndrome on BMD are not measurable by DXA in children and subtle in adults. In fact, some adult females in our study actually exhibited increased BMD for no apparent clinical, nutritional, or behavioral reason. We saw only a significant decrease in BMD of the femoral neck of adult males. Why only adult male patients would show this effect is unclear, but it may be related to their increased average height, weight and/or skeletal morphology as compared to females. Systematic studies comparing bone morphology between adult males with Marfan syndrome and control individuals will be required to clarify this issue.
It may prove that declines in bone density will appear only over longer periods of time. The average age of our adult patients was only 41 years. It would now be interesting to see whether the effects are accelerated as compared to the normal population, as Marfan syndrome patients reach more advanced ages.
Reduced BMD has been reported by others in the femoral neck of adult males [9,10]. However, decreased axial BMD in females has also been reported [9,10]. This was not observed in the present study. Our BMD values for L2–L4, femoral neck, and total body in females with Marfan syndrome were significantly higher than values reported in studies using Hologic instrumentation [6,10], but similar to the values of L2–L4 and femoral neck reported by Tobias et al. , who used a Lunar DXPL machine.
Additionally, T-scores are more comparable between instruments than Z-scores. LeParc et al.  had reported decreased BMD Z-scores in adult females with Marfan syndrome. If T-scores were reported in these studies, differences in the machines used to acquire BMD data could then be ignored. However, since only Z-scores were reported, accurate comparison is not possible.
Methodological concerns have been raised about the use of DXA analysis in children. Reporting indices of bone density in children is not advisable . Maynard et al.  noted inconsistencies in the diagnostic classification of osteopenia among three sex-non-specific reference data sets, resulting in misclassification of boys as osteopenic. DXA provides information on the bone mineral content, but does not provide true bone density measurements. Additionally, DXA does not account for large changes in body and skeletal size that occur during growth, thus making it difficult to use in longitudinal studies on children. Nelson and Koo  recommended the use of total body DXA scan analysis in children. This measurement provides the maximum skeletal area for measurement that results in an increased likelihood of detecting a change in overall bone mass. They also emphasized that, if indices of bone density are used in children, these data must be interpreted with caution. Even with the use of total body analyses and the avoidance of the application of T-score indices, however, we saw no significant deviation from the norm in children with Marfan syndrome.
Based on the genetic and biochemical defects in Marfan syndrome, it is biologically feasible that patients with Marfan syndrome would exhibit decreased BMD and be at increased risk of fracture. Marfan syndrome is caused by an autosomal dominant mutation in the fibrillin (FBN1) gene located on chromosome 15q(1.5–2.1) . Fibrillin-1 is a glycoprotein found in many types of connective tissue, including bone . In cultured human fetal 72-day-old long bones, Dallas et al.  co-localized fibrillin-1 to the outer periosteum and the osteoblastic layer adjacent to the new bone surface. Carter et al.  postulated that mutations in the epidermal growth factor-like repeats of fibrillin could interfere with calcium binding and therefore lead to a reduction in peak bone mass. Kohlmeier et al.  postulated that abnormal fibrillin might affect the biomechanical properties of bone by interfering with the distribution of mechanically induced strain imposed during skeletal development, or by altering the mineralization of bone matrix. Specific fibrillin-1 mutations have not, however, been identified in osteoporotic individuals with Marfan syndrome. The effects of fibrillin deficiency on the material properties of bone and its association with fracture risk are unknown.
Another issue that has not been addressed is whether long-term use of medications that these patients typically take, such as β-blockers or calcium channel blockers or other medications, have an influence on BMD, independent of any effects of the syndrome.
In fact, no correlation between decreased BMD and fractures has ever been reported [8,10]. Bone densitometry studies provide a measure of bone mineral content, but do not account for bone volume . Skeletal morphology, including cross-sectional area, may be a more significant parameter in fracture risk.
While we noted the activity and fracture history of our patient population, we found no evident propensity for increased fractures in general, or more specifically, femoral neck fractures in adult males. Many of the general fractures in our patients occurred several years before the study occurred. Many of the fractures were minor or associated with expected trauma. Our yes/no questionnaires were not designed to determine actual physical activity levels of the individuals. Nearly one-quarter of the patients had a fracture. It is not clear how this rate compares to the general population. Nor is it certain whether the fractures observed in our patient sample are coincidental and not due to decreased BMD. Previous studies have reported no significant fracture risk in Marfan syndrome. Our study cannot clarify this issue. Long-term, controlled studies that address fracture incidence in patients with Marfan syndrome could be of value in determining whether some patients with Marfan syndrome are predisposed to fractures.
In conclusion, the present study results confirmed the presence of osteopenia of the femoral neck in adult males with Marfan syndrome, but our study results did not show evidence for osteopenia in adult females or children with Marfan syndrome. We could not explain the sex-related difference, nor rule out other causes for a reduction in BMD in the Marfan syndrome patients. However, our results lead us to question the value of aggressive BMD evaluations by DXA in these patients, particularly prior to reaching mid-age. Based on the existing literature regarding BMD in children, we believe that it is not possible to make definite conclusions about the BMD status of children with Marfan syndrome. Further investigations will be required to shed insights into the natural history of BMD in adults and children with Marfan syndrome.
For the practicing clinician, the management and prevention of osseous complications in patients with Marfan syndrome has been of long-standing concern. The need to orthopedically assess and reduce fracture risk is an important issue in the management of patients with Marfan syndrome. However, we conclude that any application of bone mineral replacement therapy such as bisphosphonate, selective estrogen receptor modulators, hormone replacement therapy, and vitamin D in these patients may be premature based on the existing evidence.
We acknowledge the Children's Clinical Research Center of Weill Medical College of Cornell University (GCRCMO1RR06020), Adult Clinical Research Center of Weill Medical College (M01RR00047) the John R. Cobb Scoliosis Research Fund and the assistance of Elizabeth Wood with construction of the database. We also appreciate the suggestions provided by Drs. David B. Schowalter and Christopher Bibbo. We also thank Marshfield Clinic Research Foundation for its support through the assistance of Alice Stargardt, Doreen Luepke, and Graig Eldred in the preparation of this manuscript.