| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
EDITORIAL |
School of Medicine and Pharmacology (R.L.P.), University of Western Australia, Crawley, Western Australia 6009, Australia; and Department of Endocrinology and Diabetes (K.Z.), Sir Charles Gairdner Hospital, Nedlands, Western Australia 6009, Australia
Address all correspondence and requests for reprints to: Richard L. Prince, School of Medicine and Pharmacology, University of Western Australia, Crawley, Western Australia 6009, Australia. E-mail: rlprince{at}cyllene.uwa.edu.au.
Anorexia nervosa (AN) warrants study by all endocrinologists for a variety of reasons, not the least being the devastating endocrine and structural effects of this disorder of appetite on its unfortunate victims. This is all the more poignant because those of us with more normal satiety regulation lack comprehension of the disorder. If the impact and epidemiology of AN in Westernized societies is not sufficient reason for a detailed understanding of the effects of the disorder, then its role as a physiological model for adolescent and early adulthood starvation the world over should interest both humanitarians and physiologists, involving as it does the regulation of growth and fuel metabolism.
In this great tapestry, it could be argued that the skeletal impact of the disorder is of low importance. There are two reasons why this argument is invalid. The first is that the skeleton accurately reflects the physiological effects of the disease written in its shape, size, and function, and second, the majority who survive the ravages of the adolescent phase of the disorder are often left with a permanently damaged skeleton as a result of lack of attainment of their genetically programmed optimal peak bone mass with consequent increased fracture risk (1).
Anatomy of Skeletal Growth
It is a truism to state that our shape is determined by our bone structure. The huge variety of shape and size of our species, evident to us all every day, is a result of the huge variation in skeletal structure determined in part during the few short years of our adolescent growth spurt. Assessment of adolescent growth by the measurement of height is a part of the physical assessment of teenagers and is a useful pointer to a variety of disorders of the activity of the physes, the determinants of height. However, bone enlargement is determined not only by an increase in length but also by an increase in width and depth, together with an increase in the bone within the bone. Cohort studies show that peak height velocity occurs 2–3 yr before menarche and peak bone mineral accretion in the 2 yr around menarche (2, 3). Eight percent of adult height and 22% of adult total-body bone mineral are gained in the 2 yr surrounding peak height velocity (4). Because peak bone mineral accretion occurred 1 yr later than peak height velocity, there is a decrease in size-corrected bone mineral density (BMD; the bone within the bone) before menarche (3, 5), reflecting a disassociation between linear skeletal growth and bone mineral accretion during the rapid lengthening of bones occurring at this time. Careful examination of the femoral neck site confirms that the periosteal expansion reaches its maximum about 2 yr after the attainment of peak height (6).
This latter growth cannot be assessed by physical examination but rather needs an imaging technique best supplied by dual-energy x-ray absorptiometry (DXA) morphometry, because of its low radiation dose and accuracy. The technique has been criticized by many as not supplying a true volumetric estimate of skeletal size by only estimating width but not depth. This ignores the huge potential for insights into the accurate assessment of events occurring on the periosteum of the bone as well as at the growth plate, when BMD is split into its components mass and area and studied in relation to normative values now available in the same way as height and weight are expressed on centile charts (7). These reference data will be incorporated in bone density software in the near future.
Strong support for the contention that DXA-measured size and mass provide important insights into the process of bone development comes from the papers published in this volume of the Journal on the effects of AN on the skeletons of 34 anorexic adolescents, two thirds of whom were postmenarcheal, and 33 control subjects (8, 9). The height of the cases and controls was similar, but the whole-body bone area of the anorexics was 5% lower. Thus, although the growth plates of the AN patients had functioned normally, there had been substantial reduction in periosteal bone apposition. This improved dramatically, due to an increase in bone width but not height, a median of 9 months later in the 14 AN patients who recovered (10% increase in body mass index and at least three menstrual periods in 6 months) but not in the 20 who had not. The nonrecovered patients, in addition to having lack of growth on the periosteal surfaces, were also deficient in growth of the bone within the bone. As expected in the recovered patients, bone formation within the bone did not quite keep up with the expansion of the bone.
At the lumbar spine, the concept that malnutrition in AN resulted in narrower bones needs some modification. At this site, the bones appeared to be of normal size at baseline but were deficient in the amount of bone within the bone. Successful refeeding was associated with an increase in the bone within the bone. Thus, vertebral body periosteal bone growth did not seem to be as impaired, as was the case at other skeletal sites.
This developmental sequence of bone growth, first physeal growth and then modeling on bone surfaces, is helpful in timing the onset of nutritional and other disorders. If it is before the onset of puberty, the result will be height stunting when compared with age norms; if after menarche, it will be in modeling (bone formation on previously quiescent surfaces) occurring on bone surfaces on the outside and inside of bone. It is interesting to note that malnutrition results in delayed puberty with a reduction in the expected height for age; however, the prolonged activity of the growth plate means that these individuals are not markedly stunted. Indeed, their final height, although often lower than that determined by genetic potential, is similar to those with constitutional delayed puberty (10). On the other hand, malnutrition, depending on its timing and duration, may permanently impair bone modeling so that the individual ends up with a substantial deficit in bone mass, especially at more cortical sites (11).
Physiology of Skeletal Growth
The physiological determinants of normal skeletal growth and the impact of malnutrition on the growth plate, periosteum, and bone within the bone are complex. The currently recognized major regulators of bone growth are GH via IGF-I signaling and estrogen. During the starvation of AN, estrogen levels are low, and markers of bone resorption are high, consistent with the major effect of estrogen on bone to reduce bone turnover. Disappointingly, randomized clinical trials of gonadal hormone replacement in AN after puberty, although showing suppression of bone resorption, do not show a long-term increase in DXA areal BMD (12, 13), perhaps related to the fact that the initial estrogenic effect on stimulation of the growth plate had already occurred. If hypogonadism occurs in older patients with AN, estrogen replacement may play a larger role in skeletal preservation (14).
In patients with AN, GH levels are high, but IGF-I production is low, due to resistance to GH action. Markers of bone formation are also low. These changes are reversed by refeeding, which is associated with increased IGF-I (15, 16). This may explain the observed increase in bone size and bone within the bone, because IGF-I promotes not only longitudinal bone growth but also cortical and trabecular bone formation (17, 18). IGF-I replacement, although partially successful in increasing areal BMD especially if combined with gonadal hormone therapy (12), does not appear to be as successful as refeeding. Unfortunately, in these studies, changes in bone area and bone mass have not been analyzed independently.
Into this complex hormonal milieu of nutrition and skeletal development, the second paper in this issue of the Journal reports data on new hormonal regulators of nutritional intake, ghrelin and protein YY (PYY), which affect behavioral function in regard to satiety (9). Ghrelin, a 28-amino-acid hormone, is produced by the stomach mucosa and is considered to be a stimulator of food consumption behavior as well as regulating nutrient disposal in the body by a variety of effects including the stimulation of GH and ACTH release. In AN, not surprisingly, ghrelin and cortisol levels were high, and in multivariate analysis, the higher the ghrelin, the worse the increase in bone area over the 9 months of the study, suggesting a periosteal effect in this condition. Another gut-related hormone, PYY is produced in endocrine L cells of the distal intestine. In adults, PYY levels are lower in obese patients than controls and lower during fasting than after food. Infusion of PYY induces a substantial reduction in food consumption (19). In AN, PYY levels are inappropriately high, suggesting a possible pathological role and, like ghrelin, are associated negatively with the rise in whole-body bone area in multivariate analysis. Whether these findings indicate a short-loop incretin-like effect of these hormones directly on bone cells or, more likely, are a co-correlate of the numerous effects of the response to refeeding, which in this study varied, remains to be determined.
Clinical and Research Implications
DXA adds information to height data because it gives information on the periosteum and the bone within the bone, the other critical determinants of bone structure, in addition to the growth plate. Unfortunately, combining the area and bone mineral content to produce the areal BMD has served to confuse many by producing one shade of gray. This is because they are both changing rapidly and independently during puberty. Thus, the assessment of disorders of nutrition during growth should involve whole-body DXA body composition, analyzed as bone area and bone content as well as BMD, height, and weight. Modern software programs have simplified these analyses so that the clinician can get output on percentile charts that can be discussed with patients, ensuring full understanding of the problem by the patient and those involved in their care. These outputs can be used to assess response to therapy, remembering that AN persists as a chronic disorder that may relapse during adult years. The research implications are equally important. If low-radiation-dose DXA is used correctly, it can give reasonable data on the bone size, both in length and width, and on the bone within the bone, so it may be possible to dissect out the regulators of specific aspects of bone development, which may then form therapeutic targets.
It is interesting to speculate on the role of technological advances in health care on patient outcome. It could be argued that the glucometer has improved the lifestyle and prognosis of diabetes more than any other management intervention. In the management of growth disorders, DXA body and bone composition may enter the same league in years to come.
Footnotes
Disclosure Statement: R.L.P. reports having received honorarium from Hologic Inc. K.Z. has nothing to disclose.
For articles see pages 1231 and 1292
Abbreviations: AN, Anorexia nervosa; BMD, bone mineral density; DXA, dual-energy x-ray absorptiometry; PYY, protein YY.
Received February 11, 2008.
Accepted February 25, 2008.
References
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
