Increased Prevalence of Radiological Spinal Deformities in Active Acromegaly: A Cross-Sectional Study in Postmenopausal Women


  • The authors have no conflict of interest


This cross-sectional study shows that high numbers of postmenopausal women with acromegaly develop vertebral fractures in relation to the activity of disease. In patients with active acromegaly, vertebral fractures occur even in presence of normal BMD, whereas in patients with controlled acromegaly, vertebral fractures are always accompanied by a pathological BMD.

Introduction: We studied the frequency of radiological vertebral fractures in a cohort of postmenopausal women with active or controlled acromegaly.

Materials and Methods: Thirty-six postmenopausal acromegalic patients (15 with active and 21 with controlled disease) were evaluated for BMD, bone metabolism (serum 25-hydroxyvitamin D, PTH, bone-specific alkaline phosphatase {BSALP}, and urinary deoxypyridinoline {Dpd}), and vertebral quantitative morphometry. Thirty-six nonacromegalic postmenopausal women, matched for age, were selected among the patients consulting the Bone Center as a control group for BMD evaluation and vertebral quantitative morphometry.

Results: Vertebral fractures were shown in 19 patients (52.8%) and 11 controls (30.6%; χ2: 3.7; p = 0.06). Fractured acromegalic women were older and had higher serum IGF-1, Dpd, and BSALP and lower T score and serum vitamin D values compared with nonfractured patients. Moreover, the fractured women had a longer diagnosis and were in the postmenopausal period for a longer period than the nonfractured women. The fracture rate was significantly higher in active than in controlled acromegaly (80% versus 33.3%; χ2: 7.6; p = 0.008). The patients with active acromegaly who fractured (12 cases) had significantly higher serum IGF-1 values (356 ng/ml; range: 212–950 versus 120 ng/ml; range: 84–217; p < 0.001) and T scores (−1.3 SD, range: −2.9 to +1.3 versus −2.7 SD, range: −3.4 to −1.5, p = 0.04) compared with the fractured women whose disease was controlled (7 cases). All fractured women with controlled acromegaly had T scores <−1.0 SD (57.1% of them had osteoporosis, and 42.9% were osteopenic). In contrast, 41.7% of women whose fractures were associated with active disease had a normal T score (>−1.0 SD), whereas osteopenia and osteoporosis were found only in 33.3% and 25.0% of them, respectively.

Conclusions: This cross-sectional study shows that high numbers of postmenopausal women with acromegaly develop vertebral fractures in relation to the activity of disease. Furthermore, our study shows that, in patients with active acromegaly, vertebral fractures occur even in the presence of normal BMD, whereas in patients with controlled acromegaly, vertebral fractures are always accompanied by a pathological BMD.


GROWTH HORMONE (GH) is an important factor in the regulation of bone growth and bone metabolism during the lifespan. (1) The progressive decline of GH secretion(2) is considered to be among the factors contributing to age-related and postmenopausal bone loss. (3, 4) Moreover, GH deficiency plays an important role in determining bone loss in patients with hypopituitarism. (5)

The effects of GH excess on bone metabolism and structure have been previously evaluated with conflicting results. (6) Traditionally, acromegaly is considered as one of the causes of secondary osteoporosis. (7) In fact, several studies have suggested a GH-mediated increase of bone turnover in acromegaly based on the evaluation of biochemical markers, calcium kinetics, and bone histomorphometry. (8–11) However, data on BMD are conflicting. Some studies reported increased BMD in patients with acromegaly, (12–14) others observed reduced BMD, (15, 16) and others did not find any difference in BMD between acromegalic patients and healthy subjects. (17–19) Indeed, the effects of GH excess on bone are likely variable in relation to the site as well as to the gonadal status of the patients, being hypogonadal patients those shown to be at higher risk of osteoporosis. (20, 21) Moreover, trabecular and cortical bone show different sensitivity to GH excess at various skeletal sites. (22–24)

A recent paper has provided evidence for a reduced trabecular biochemical competence in active acromegaly despite the increased trabecular bone content of calcium. (25) Whether this feature may lead to an increased risk of fracture in acromegaly remains to be clarified. Surprisingly, in fact, very few papers have been published dealing with the hard clinical endpoint of osteoporosis (i.e., bone fractures) in acromegaly. (26, 27) Only one recent study has reported that acromegaly seemed to be a protective factor against clinical fractures. (27) Radiological vertebral deformities have emerged in the last decade as the method of choice for evaluating the true prevalence of fractures in population studies. (28) Moreover, vertebral deformities have been found to correlate well also with the clinical outcome of the osteoporotic patients. (29, 30)

In this study, we evaluated the prevalence of radiological vertebral fractures, identified as deformities of vertebral bodies by a quantitative morphometric analysis, in postmenopausal women with active and controlled acromegaly defined by stringent clinical and biochemical parameters. (31) In this selected population, which based on previous reports(22, 23) could have been considered at higher risk for osteoporosis, we aimed at studying whether the prevalence and degree of spinal fractures were related to age, duration of postmenopausal period, bone turnover, and activity of acromegaly.



We studied 36 postmenopausal acromegalic patients. The median age of the patients was 60.5 years (range, 44–79 years). Acromegaly had been previously diagnosed by failure of suppression of serum GH concentrations <1 ng/ml after a 75-g oral glucose load together with fasting plasma IGF-1 concentrations above the normal ranges for age. (31) The median duration of disease, estimated on the basis of clinical history (i.e., when the patient recalled appearance of signs and symptoms of the disease) was 6.5 years (range, 2–25 years). As established in the Consensus Conference held before designing the study, (31) the patients were defined to have a controlled disease when nadir GH after an oral glucose load was <1 ng/ml and circulating IGF-1 values were reduced to an age-adjusted normal range. Twenty-one of our patients had controlled disease according to the Consensus criteria(31) after surgery (11 of these 21 patients had disease controlled by somatostatin analog treatment). (32) The remaining 15 patients had active disease (oral glucose load nonsuppressed serum GH {>1 ng/ml}, IGF-1 concentrations above the normal range corrected for age, with or without evidence at magnetic resonance imaging of persistent pituitary mass(31)). The patients with hypopituitarism as an effect of a radical treatment and those on treatment with drugs influencing bone metabolism (included estrogen substitution) were not included in this study.

Thirty-six postmenopausal women were enrolled as a control group. Each control subject was matched for age and duration of postmenopausal period with each acromegalic patient. These women were selected from a larger population of postmenopausal women who had anteroposterior and lateral X-ray examinations of the thoracic and lumbar spine performed and attended the out-patient osteoporosis clinic of University of Brescia with symptoms suggestive of a pathological involvement of thoracic-lumbar spine. Control women who took drugs influencing bone metabolism or had clinical evidence of diseases known to affect bone status were not included in the analysis.

The patients and control subjects gave informed consent to this study, which was approved by the local Ethical Committee.

Measurement of BMD and quantitative morphometrical assessment of vertebral fractures

BMD of the lumbar spine was measured by DXA (QDR-1000; Hologic, Waltham, MA, USA). The results were cross-calibrated with the European Spine Phantom. These measurement were made at the time of the spine X-ray. BMD of the lumbar spine (L2-L4) was measured separately and expressed as the mean. As applied to normal individuals, osteopenia and osteoporosis were defined with a T score below −1.0 SD and below −2.5 SD, respectively. (33) Fractured vertebrae, assessed by the following methods, were excluded from the BMD analysis.

For assessment of vertebral fractures, anteroposterior and lateral X-ray examinations of the thoracic and lumbar spine were performed and were centrally digitized and assessed by the same experienced physician. The use of fixed percentage reduction in vertebral height is the simplest and most practical method of studying vertebral deformities. In this study, a quantitative morphometric assessment of vertebral deformity in T4-L4 was performed using dedicated morphometry software (Spine-X Analyzer; ICAM Diagnostics, Milan, Italy). In brief, using a translucent digitizer and a cursor, six points were marked on each vertebral body to describe vertebral shape. Anterior (Ha), middle (Hm), and posterior (Hp) vertebral heights were measured, and height ratios (Ha/Hp, Hm/Hp, Hp/Hp of the above vertebrae, Hp/Hp of the below vertebrae) were calculated for each vertebra from T4 to L4; deformities were defined as mild, moderate, or severe based on a height ratio decrease of 20–25%, 25–35%, and >35%, respectively. (34) The morphometric analysis was performed by a single operator (MN). The intraobserver CV, evaluated on a series of 10 measurements, was between 5% and 8%. The patients and the subjects with clinical history of recent significant trauma and those with neoplastic disease or other bone disease or prolonged immobilization were excluded from the analysis. According to previous observations, (35, 36) the wedging mild vertebral fractures with the loss of anterior height alone in presence of a normal T score was considered as suspected for different pathological process than osteoporosis. Differential diagnosis, even if difficult with the radiological approach, prevents overestimation of prevalence of osteoporotic fractures. In fact, up to 30% of spinal deformities are reported to be of nonosteoporotic origin. (35)

Biochemical measurements

Blood samples were collected after an overnight fast. Serum was promptly separated and stored at −20°C until assay. GH and IGF-1 were measured by Immulite 2000 (DPC, Los Angeles, CA, USA). The interassay CVs of GH and IGF-1 assays ranged from 5.5% to 6.2% and from 6.4% to 11.5%, respectively. Serum calcium (Ca), phosphate (P), and albumin were measured using auto-analyzer methods (Technicon Instruments, Tarrytown, NY, USA); Ca was corrected for albumin concentration. Serum 25-hydroxyvitamin D was measured by radioimmunoassay (RIA; DiaSorin, Saluggia, Italy). The sensitivity of the test was 1.5 ng/ml, and the intra-assay CV ranged from 8.6% to 12.5%. In our laboratory, the reference range was between 25 and 75 ng/ml. All blood samples for PTH were analyzed using an immunoradiometric assay (normal range: 7–53 pg./ml; Nichols Allegro, San Juan Capistrano, CA, USA). Urinary deoxypiridinoline (Dpd) was evaluated by Pyrilinks-D (normal range, 2.3-6.7 μM/M creatinine; Metra Biosystem); bone-specific alkaline phosphatase (BSALP) activity was measured by Alkaphase-B (normal range: 10–50 IU/liter; Metra Biosystem).

Statistical analysis

All data are expressed as the median and range. The relative rarity of acromegaly, as well as the stringent criteria of selection (i.e., female, postmenopausal period, no history of vertebral significant trauma) forced us to design the study taking into account the hypothesis of a large difference in fracture rate between the patients with active and those with controlled disease. This approach allowed us to attain an adequate power of the statistical analysis (at least 80%) with small study groups. This study was conducted sequentially to minimize the time of the study and to obtain results more rapidly. Unpaired data were compared using the Mann-Whitney test. Multiple comparisons were performed using Kruskal-Wallis' test with posthoc Bonferroni correction. Correlation between variables was evaluated calculating the Spearman rank coefficients. A logistic regression model was used in the statistical analysis of risk factors for the occurrence of spinal deformities. Frequencies were compared using χ2 test with Fisher correction, when appropriate. Statistical significance was assumed when p values were ≤0.05.


The median T score in our patients was −1.3 SD (range, −4.1 to +1.3), significantly (p = 0.001) higher than that measured in control women (−2.4 SD; range, −4.1 to −0.2). In the patients, BMD was normal (>−1.0 SD) in 16 patients (44.4%), whereas 10 patients (27.8%) had osteopenia and 10 (27.8%) had osteoporosis.

Vertebral fractures were shown in 19 patients (52.8%) and 11 controls (30.6%; χ2: 3.7; p = 0.06). In acromegalic women, 13 (68.4%) had two or more fractures, and in 6 (31.6%), single fractures were observed. In all of the patients with multiple vertebral fractures and in five of the six patients with a single fracture, the deformities consisted of a decrease in vertebral Hm and Hp with or without a decrease in Ha. In the one patient with a single wedging fracture and the loss of anterior vertebral height alone, the degree of deformity was moderate. The fractures were mild in 15 patients (78.9%), moderate in 3 (15.8%), and severe in only 1 patient (5.3%). A significant correlation was found between the degree and number of fractures (ρ = 0.55; p = 0.01). The degree of fractures was not significantly correlated with T score (ρ = 0.06; p = 0.81), serum IGF-1 values (ρ = −0.11; p = 0.57), duration of acromegaly (ρ = 0.12; p = 0.63), and duration of postmenopausal period (ρ = 0.13; p = 0.60).

Fractured women were older and had higher serum IGF-1, Dpd, and BSALP and lower T score and serum vitamin D values compared with nonfractured patients (Table 1). Moreover, the fractured women had a longer lasting disease and were in the postmenopausal period for a longer period than the nonfractured women (Table 1). No significant difference in serum PTH values was found between fractured and nonfractured women. Multivariate logistic regression analysis showed that duration of postmenopausal period (OR, 1.2; 95% CI, 1.04-1.30) and the activity of acromegaly (OR, 18.9; 95% CI, 2.2-158.9) were independently associated with the risk of vertebral fractures, whereas the other covariates did not attain significance when considered as possible additions to the baseline model.

Table Table 1.. Demographical and Clinical Data of Acromegalic Postmenopausal Women With (19 Cases) and Without (17 Cases) Radiological Evidence of Vertebral Fractures
original image

Table 2 shows that the postmenopausal women with active acromegaly had slightly, but not statistically significant, elevated biochemical markers of bone metabolism and similar T scores with respect to the women whose acromegaly was controlled. However, the fracture rate was significantly higher in active than in controlled acromegaly (80% versus 33.3%; χ2: 7.6; p = 0.008). These controlled patients had comparable fracture rate with respect to the control postmenopausal women (33.3% versus 30.6%). The patients with active acromegaly who fractured (12 cases) had significantly higher serum IGF-1 values (356 ng/ml, range: 212–950 versus 120 ng/ml, range: 84–217; p < 0.001; Fig. 1) and T score (−1.3 SD, range: −2.9 to +1.3 versus −2.7 SD, range: −3.4 to −1.5; p = 0.04; Fig. 2) compared with the women with fractures whose disease was controlled (7 cases). Moreover, among active acromegalic patients, IGF-1 levels were significantly higher in women with fractures versus women without fractures. Interestingly, in controlled patients, IGF-1 tended to be, without reaching statistical significance, lower in women with fractures versus women without fractures (Fig. 1). Conversely, T scores were similar in acromegalic patients with fractures versus acromegalic patients without fractures, whereas they were reduced in a highly significant manner in women with fractures versus women without fractures in those with controlled disease. The analysis of individual data shows that all fractured women with controlled acromegaly had T scores below −1.0 SD; four patients (57.1%) had osteoporosis and three (42.9%) had osteopenia (Fig. 2). In contrast, a normal T score (> −1.0 SD) was found in 5 of the 12 patients (41.7%) in whom the fractures occurred with active acromegaly (Fig. 2). In the other seven patients with active acromegaly, fractures were associated with osteopenia (T score between −1.0 and −2.5 SD) in four cases (33.3%) and with osteoporosis (T score <−2.5 SD) in three cases (25%; Fig. 2).

Table Table 2.. Demographical and Clinical Data of Acromegalic Postmenopausal Women With Controlled (21 Cases) or Active Disease (15 Cases)
original image
Figure FIG. 1..

Serum IGF-1 concentrations in patients with and without vertebral fractures subdivided in relation to activity of acromegaly. The data were compared using the Kruskall-Wallis test with posthoc Bonferroni correction.20

Figure FIG. 2..

T score evaluated by vertebral DXA in patients with and without vertebral fractures subdivided in relation to activity of acromegaly. The data were compared using the Kruskall-Wallis test with posthoc Bonferroni correction.20

Active and controlled patients with fractures showed no significant difference in age (mean, 64.5 years; range, 49–71 years versus mean, 65 years; range, 54–79 years; p = 0.51), duration of disease (mean, 6.5 years; range, 2–25 years versus mean, 15 years; range, 3–25 years; p = 0.46), duration of postmenopausal period (mean, 13.5 years; range, 4–26 years versus mean, 17 years; range, 3–30 years; p = 0.22), serum BSALP (mean, 56.5 U/liter; range, 36.0-87.0 U/liter versus mean, 65.0 U/liter; range, 38.0-71.0 U/liter; p = 0.6), 25OH vitamin D (mean, 17.5 ng/ml; range, 10.0-37.0 ng/ml versus mean, 17.0 ng/ml; range, 12.0-21.0 ng/ml; p = 0.5), PTH (mean, 44.0 pg./ml; range, 23.0-73.0 pg./ml versus mean, 58.0 pg./ml; range, 29.0-66.0 pg./ml; p = 0.1), and urinary Dpd (mean, 9.5 μM/M creatinine; range, 7.5-17.9 μM/M creatinine versus mean, 11.8 μM/M creatinine; range, 7.6-19.4 μM/M creatinine; p = 0.9) values. Patients with active acromegaly had predominantly mild degree vertebral fractures (83.3% of total fractured), and in one-half of cases, the fractures were single (Table 2). In patients with controlled disease, the fractures were multiple in all cases and were of mild degree in 71.4% of the cases (Table 2). In controlled acromegalics, the more severe fractures occurred in those patients with the lowest BMD.


This cross-sectional study shows that high numbers of postmenopausal women with acromegaly develop vertebral fractures in relation to the activity of disease. Furthermore, our study shows that, in patients with active acromegaly, vertebral fractures occur even in the presence of normal BMD, whereas in patients with controlled acromegaly, vertebral fractures occur only in the presence of decreased BMD.

The effects of GH on bone have been a matter of discussion over the last years. It has been widely emphasized that GH has anabolic effects on bone. GH stimulates the proliferation of osteoblasts, and at some extent, it also stimulates differentiated functions of these cells. (23) Moreover, GH stimulates osteoclastic bone resorption through both direct and indirect actions on osteoclast differentiation and indirect activation of mature osteoclasts. (37) Bone metabolism and density improve during GH treatment in patients with hypopituitarism. (38) Recent studies have also proposed GH and IGF-1 as therapeutic tools in osteoporosis. (39) However, it is still controversial whether the GH excess, as it occurs in acromegaly, may lead to negative or positive effects on bone. Patients with acromegaly have high bone turnover with positive calcium balance. (8–11) The BMD is enhanced in the forearm and femoral region but not at the spinal level, where osteoporosis may occur, especially in the presence of hypogonadism. (19–22) In fact, estrogens modulate GH effects on bone, likely by regulating GH receptor expression on osteoblasts. (40) Our study is in line with recent reports showing that BMD measured at the vertebral level is variable in postmenopausal women with acromegaly(19–22); osteoporosis occurs in only 27% of the patients without any difference between active and controlled disease.

However, the real new element in our work is that, even if a minority of our patients had osteoporosis, radiological vertebral fractures were found in more than one-half of our population. It is noteworthy that the prevalence of fractures in our acromegalic patients was higher than that observed in the control subjects, although the latter was a selected group with apparent high risk of vertebral damage either because they had a low median T score or because all of them referred with symptoms suggestive for a spinal pathological involvement. Indeed, the high fracture rate observed in the whole group of our acromegalic patients is comparable with that reported by epidemiological surveys in women with osteoporosis after a long time since menopause. (29) In fact, it has been estimated that after 10 postmenopausal years 55% of osteoporotic women may develop vertebral fractures. (29) However, it is noteworthy that, in our population, the fracture rate was not only correlated with the duration of postmenopausal period, but also with the activity of acromegaly. In fact, if only the subgroup of patients with active acromegaly was considered, the prevalence of vertebral fractures became really impressive and by far exceeded the reported prevalence in patients with postmenopausal osteoporosis. (29, 41) Finally, concerning the influence of the biochemical control of the disease on the fracture prevalence in acromegaly, it is interesting to note that acromegalic patients with fractures were those who had the highest circulating IGF-1 levels, even in comparison with active acromegalic patients without fractures. On the contrary, the acromegalic subjects with fractures with controlled disease had a tendency to have lower IGF-1 levels versus controlled women without fractures, pointing toward osteoporotic fractures being a potential risk of chronic overtreatment of acromegaly. (42)

A number of prevalence studies have found that the risk of vertebral deformity is related to BMD. (35) An intriguing aspect raised from our data is that we found a different relationship between BMD and fractures in patients with controlled and those with active acromegaly. In the former, the fractures always occurred, as expected, in the presence of a decreased BMD, with the more severe lesions occurring in the patients with the lowest BMD. In contrast, in about one-half of patients with active acromegaly, the vertebral fractures were associated with a normal BMD, as assessed at the spinal level by DXA. It has been argued that the fractures associated with normal BMD derive in a large proportion of cases from a different pathological process than osteoporosis. (35) The different patterns of vertebral deformities of patients with controlled versus those with active acromegaly (i.e., multiple and more severe in the former, single and mild-moderate in the latter) would suggest the presence of additional pathogenetic factors predisposing to fractures in patients with active acromegaly. However, this hypothesis could not be shown in this cross-sectional study. In fact, although the radiological approach does not allow any clarification in this context, it has been suggested that wedging fractures, in which the deformity is sustained by a mild decrease of anterior vertebral height alone, may be caused by a different process than osteoporosis. (35, 36) In our patients with fractures, either with controlled or active acromegaly, almost all vertebral deformities were characterized by a loss of mid and/or posterior heights, whereas just one patient had a single wedging fracture with a loss of anterior vertebral height alone. In this case, however, the fracture was moderate rather than mild in degree and was associated with a low T score (data not shown). However, it can be hypothesized that the exaggeration in bone turnover caused by superimposing GH/IGF-1 axis hyperactivity in estrogen deficiency could lead to a slower but more generalized process with respect to estrogen deficiency alone that may lead to fracture mostly and/or prevalently in severely osteoporotic patients. Nevertheless, it is also well known that the relationship between BMD and risk of fracture is complex. (41, 43) Fracture risk is related to bone strength that is dependent on two main physical/structural factors: quantity and quality. BMD reflects bone quantity but not bone quality, which consists of structural and materials properties. (43–45) The structure of trabecular bone has a major impact on fracture risk independent of measures of bone quantity. In acromegaly, bone quality is impaired, as suggested by clinical and histomorphometric data. (25, 46) Therefore, it is not surprising that, in acromegaly, BMD may not reflect the true risk of vertebral fracture; measurement of BMD in the axial skeleton may be overestimated in conditions of GH excess because of the influence of cortical bone and bone size. (25) In this context, the volumetric measurement of BMD in acromegaly may provide a more reliable estimate of the trabecular vertebral tissue, such as already suggested by others. (6) Our findings are in line with the results reported in another form of secondary osteoporosis (i.e., glucocorticoid-induced osteoporosis), in which BMD, as assessed by DXA, is a poor predictor of the risk of fractures. (47) Inasmuch, also under high-dose fluoride treatment (thought to have anabolic actions on bone) in postmenopausal osteoporosis, BMD increases a lot without a concomitant decrease in the risk of fractures. (48) Finally, bone turnover was increased in our acromegalic patients with fractures, regardless of the activity of disease. It could be concluded that, particularly in postmenopausal patients with active acromegaly, concomitant increase in circulating IGF-1 levels and decrease in circulating estrogens may lead to an exaggeration of bone resorption, which may in turn impair the strength of newly formed bone under GH/IGF-1 stimulus. The above discussed pathophysiological aspects were derived from a selected group of acromegalic patients that consisted of women in their postmenopausal period, a population that could have been considered at risk for osteoporosis. (22, 23) In this specific group of patients, our data would suggest that elective treatment of this form of osteoporosis would be with agents known to reduce the risk of fractures by decreasing osteoclast activity such as bisphosphonates(49) or raloxifene. (50)

Besides the pathophysiological aspects, some clinical considerations could be made based on our data. Epidemiological studies have pointed out the necessity to identify early vertebral fractures, because the presence of a single fracture, even if mild and without clinical symptoms, predisposes one for more severe fractures with consequent greater clinical impact. (30, 50, 51) Another intriguing aspect concerns the survival of subjects with vertebral fractures. It has been shown that individuals with prevalent bone deformities at the spinal level, whether clinically apparent or not, have an increased mortality. (52) These aspects may be particularly important in acromegalic patients, who are a frail population with decreased survival and impaired quality of life. (53, 54) Indeed, in our patients with active acromegaly, the fractures tended to be mild and were single, both characteristics that are not correlated with an impaired quality of life in the general population. (30, 50, 51) However, although the clinical consequences of an increased fracture rate in acromegalic patients must be clarified in future studies, a critical question concerns the identification of acromegalic patients at risk to have vertebral fractures. The results of our study would suggest that DXA measurement of spinal BMD could be helpful in the presence of controlled acromegaly but not when the disease is still active. Quantitative ultrasound may provide more useful data on the quality of bone in patients with active acromegaly, although soft tissue swelling may produce false results. (46) High bone turnover may be an important laboratory marker for identifying those at risk for fractures among active acromegalic menopausal patients. Radiological evaluation of vertebral bodies is the simplest and most reliable tool to identify patients with active acromegaly in whom a specific treatment would counteract the negative effects of vertebral fractures on quality of life and survival. Finally, it can also be suggested, based on our data, that chronic overtreatment with recombinant human GH in adult patients with hypopituitarism may be potentially exposing these patients to a higher risk of fracture despite an increase in BMD found with DXA. (5)

The results of our study suggest that active acromegaly is associated with a high risk of radiological vertebral fractures, even in the presence of a normal BMD as assessed by spinal DXA. Future studies will clarify whether the early identification, treatment, and prevention of spinal deformities may modify the quality of life of acromegalic patients.