( PDF ) Rev Osteoporos Metab Miner. 2012; 4 (2): 57-62

Gómez de Tejada Romero MJ1, García Caballero A2, Groba Marco M2, Cárdenes León A2, Lázaro Archilla J2, Sosa Henríquez M2,3
1 Universidad de Sevilla – Departamento de Medicina – Sevilla
2 Universidad de Las Palmas de Gran Canaria – Grupo de investigación en Osteoporosis y Metabolismo Mineral – Las Palmas de Gran Canaria
3 Servicio Canario de la Salud – Hospital Universitario Insular – Unidad Metabólica Ósea – Las Palmas de Gran Canaria

 

Summary
Background: Pregnancy-associated osteoporosis (PAO) is a clearly described and relatively frequent entity, although there are few studies which have carried out long term monitoring of the disease.
Material and methods: 5 women affected by osteoporosis who were monitored over the long term, between 4 and 16 years. In all the patients a questionnaire on lifestyle and risk factors was completed and a physical examination carried out, as well as densitometry after the pregnancy and subsequently every year during the follow up period. At the end of this period, a lateral X-ray of the dorsal and lumbar spine was performed.
Results: In 3 cases there was a spontaneous fracture as the first manifestation of PAO, while in 2 cases a very low densitometric value was observed, without fractures, in the immediate postpartum period. All the patients received calcium and vitamin D supplements, and in 3 cases a biphosphonate (risedronate) was indicated. No new fractures were observed in any of the cases over the follow up period. The bone mineral density (BMD) increased in the lumbar spine in all the patients, but in one a decreased measurement was observed in the hip, both in the femoral neck and the total hip. None of the patients became pregnant again in the period of the study.
Conclusions: While none of the patients with PAO studied over the long term suffered a new fracture, and in all an increase in the bone mineral density in the lumbar spine was observed, which makes us think that there is a recovery of DMO over time, in one case there was a considerable decrease in bone mineral density in the proximal femur, for which reason we believe that it is advisable to carry out long term monitoring of these patients.
Keywords: osteoporosis, pregnancy, fracture, densitometry, monitoring.

Introduction

Pregnancy-associated osteoporosis (PAO) was described for the first time more than 50 years ago1. It usually appears in women of between 25 and 30 years of age in their first pregnancy and may develop with lumbar pain, loss of vertebral height and compression fractures2-5, although in most cases the disease usually passes unrecognised, since it symptoms may be attributed to back pain similar to that produced by the exaggeration of lumbar lordosis in pregnant women. Although it appears not to be a frequent disease, its exact prevalence is unknown6, and is possibly underestimated for the aforementioned reason, combined with the avoidance of X-rays of the spine and, obviously, of densitometry in pregnant patients.
Some authors have described four types of osteoporosis associated with pregnancy and breastfeeding7: a) idiopathic osteoporosis in pregnancy, b) transitory osteoporosis in the hip during pregnancy, c) post-pregnancy or breastfeeding-associated lumbar osteoporosis, and d) drug-induced osteoporosis.
Case-based descriptions of PAO usually consist of few cases, and with a follow up normally over a limited time period6,8-11. Although it is accepted that fractures do not tend to occur in subsequent pregnancies and that this type of osteoporosis is usually self-limiting12,13, there are few descriptions which include densitometric and clinical follow up over the long term, especially in our country.
Therefore, we present in this study a series of 5 cases of as many patients diagnosed with PAO and followed up over several years.

Patients, materials and methods

In carrying out this work, between 1989 and 2010 5 patients who were diagnosed with pregnancy-associated osteoporosis in the Bone Metabolism Unit of the Insular University Hospital were studied. The diagnosis of the disease was made by means of a radiological study and densitometry carried out due to the presence of back pain on the final three months of pregnancy, or post-partum, either by the detection of a fragility fracture and/or by obtaining a densitometry T-score of ≤ –2,5. All the patients were Caucasian. Their informed consent was requested and the study was approved by the ethics committee of the Insular University Hospital of Gran Canaria.
A complete clinical history of all patients was taken, along with a detailed physical examination, in order to note risk factors for osteoporosis14. The current intake of calcium was estimated using the Cummings method15.

Determination of bone mineral density by dual energy X-ray absorptiometry (DXA)

A measurement of bone mineral density (BMD) was made both in the lumbar spine (L2-L4) and in the proximal extremity of the femur using at the first stage a Hologic® QDR-1000 densitometer and subsequently Hologic® QDR-4500 Discovery. In order to calculate the T-score the following formula was applied:
T-score = (observed value – value of peak BMD)/typical deviation from BMD.
The peak BMD values were those previously published as normal in the Canarian population16. A variation coefficient of 0.75 ± 0.16% was obtained. All the determinations of BMD were carried out by the same technician (JS).

Diagnosis of fractures

The vertebral fractures were diagnosed by means of a lateral X-ray of the dorso-lumbar spine, from D3 to L5, applying Genant’s diagnostic criteria17. The presence of non-vertebral fractures was garnered from reports provided by the patients and subsequently confirmed by hospital records. Fragility fractures were considered to be those produced by a minor trauma or by a fall to the ground from a maximum height of the person themselves. Excluded were fractures resulting from traffic accidents or from falls from a greater height, as well as cranial, face, metacarpal and phalangeal fractures.

Results

Table 1 shows the baseline characteristics of the patients studied. All the women had a current intake of calcium lower than 800 mg a day and were sedentary, except one who had high levels of physical activity. None of the patients smoked or drank alcohol during their pregnancy, but one of them returned to smoking after pregnancy and four started to drink alcohol, albeit moderately. All the patients returned to their normal weight after pregnancy, with the exception of one woman who remained overweight.
Table 2 lists the year of diagnosis of PAO, the year of the last review and period of time that each patient was followed. The period of follow up was variable, varying between 4 and 16 years. In three cases the follow up was longer than 12 years.
Table 3 includes gynaecological data. The age of the patients during the pregnancy which resulted in PAO was higher than 30 years in all cases, except for one patient, who was 29 year of age. In three cases the PAO happened during the 2nd pregnancy, and in the two remaining patients, during their first gestation. All the patients, with the exception of one, breastfed their children, although lactation lasted only a few months, less than, or equal to, 6 months in all cases. Subsequent to the pregnancy which resulted in the PAO and after the follow up, none of the patients became pregnant again, or suffered a miscarriage. Over the follow up period two of the patients experienced the menopause (those with the longest period).
Table 4 show the clinical data relating to the patients. None had comorbidity: no cases of AHT, diabetes mellitus, hypercholesterolemia or urolithiasis, and in only one case was there a congenital mitral prolapse. In three cases the PAO progressed with fracture at the time of diagnosis: in two patients the fractures were costal and in one, vertebral, without previous trauma in either case. After the pregnancy, drug treatment with risedronate was indicated in three patients, and calcium and vitamin D supplements in all. None of the patients had a further fracture during the years of follow up. In three cases, new diseases appeared: hypophyseal adenoma, hypothyroidism and bilial lithiasis.
In Table 5 can be seen the changes produced in BMD, both in the lumbar spine and in the proximal extremity of the femur. The change is highly variable, since in one case the recuperation of BMD reached 137% in L2-L4 after 16 years of follow up, while in another patient a significant decrease was found in BMD in total hip, with a loss of 17.6% after 4 years of follow up. In all the women there was an increase in BMD in the lumbar spine, while in four cases the BMD in the total hip increased. In the femoral neck there was a decrease in BMD in two cases.

Discussion

PAO is a syndrome known since 1955, since the initial description of Nordin and Roper1, which was followed that same year by another article from Bret18. In the first publications only isolated cases were presented, until 1985 when Smith et al. put together a list of eight women who had been followed up for 10 years5. In this British series the patients mainly began with a vertebral fracture, which was not reproduced in successive pregnancies. A bone biopsy was carried out on these patients and in only one case were signs of increased bone resorption observed, the other seven cases being normal. Also normal were the analysis data, including vitamin D and calcitonin. The authors concluded that in these patients there could have been a transitory failure in the calciotropic hormones which prepare the maternal skeleton for birth.
Almost 60 years after Nordin and Roper’s description, we still do not know what is the physiopathological mechanism by which PAO is produced. Many possible causes have been suggested, such as an insufficient ingestion of calcium and vitamin D19, a change in the regulation of parathyroid hormone (PTH) or of the peptide related to PTH (PTH-RP)7 or even the coexistence of a previous disease which could cause the alteration of bone mineral homeostasis, such as the coexisting treatment with corticoids or with heparin, a weak form of osteogenesis imperfecta or anorexia nervosa3.
In our series of 5 cases, the average age of the patients during the pregnancy producing the disease was 34.4 years. In a larger series concerning PAO, the average age of the patients was 27 years3, but this was published in 1995 after 24 years of follow up, which means that most of the patients had had their pregnancy in the 1970s. It is well known that nowadays, for sociocultural reasons, women usually become pregnant later in life, which means that these data are not compatible. The age of occurrence of PAO in other series varied widely, between 25 and 40 years of age1-10,20-23.
Our patients did not have a high risk profile for osteoporosis. At the start of the pregnancy which resulted in the disease, only one was a smoker, after which she stopped smoking. None had family history of fragility fractures (particularly maternal), none had taken corticoids, and while their dietary intake of calcium was somewhat low, with a median of 750 mg a day, in all cases they had received calcium and mineral supplements during pregnancy.
One of the most controversial matters in relation to this syndrome is the understanding of whether the skeleton recuperates in the long term, if there is a high risk of new fractures in future pregnancies and if it is necessary or not to indicate treatment, and with which drug. The initial publications of Smith et al.3-5 suggested a favourable outcome the patients. In fact, in a series of 24 patients followed over 24 years, only one patient continued to have new fragility fracture, but suffered from a weak form of osteogenesis imperfecta. 10 patients had 14 pregnancies after that in which the PAO happened, which progressed completely normally without new fractures. However, other cases have been published in which the development was not so favourable. Thus, in a series of 11 patients, followed up over between 1 and 19 years, 3 patients had new fractures after the pregnancy24. In our series, none of the patients had any new fractures. In spite of this, none became pregnant again.
The BMD was determined in all our patients after pregnancy and in no cases were earlier densitometries available. In all cases we found low values of BMD in the lumbar spine, in four of the five cases below a T-score of -2.5. In one case the T-score value was -2.2, which, combined with the presence of costal fractures, justified the diagnosis of PAO. The changes in BMD in our patients over time was disparate. As can be seen in Table 5, in the lumbar spine a gain was seen in all cases, with an increase in BMD which varied between 1.7 and 137%. However, in the BMD in the femoral neck, in some patients a marked improvement was observed, whilst in others a fall of up to 17.6% was observed, despite the fact of four years having passed since the pregnancy and the patients having received treatment with risedronate, calcium and vitamin D. This wide variation in the development of BMD has also been described in other series2-4,7-9,12,22 and suggests a significant heterogeneity in the clinical characteristics of those patients who have suffered from PAO25.
In conclusion, a long term follow up has been carried out of a series of 5 women who suffered from PAO. None of them became pregnant again, nor did they suffer new fractures. Although the BMD improved in all cases in the lumbar spine (L2-L4), in two cases a decrease was found in different locations of the proximal femur. While PAO appears to be a transitory disease, which progresses favourably, given the absence of new fractures, it is advisable to carry out a follow up of these patients, since there are cases in which BMD is not normalised and could result in an increased risk of fracture.
Conflict of interest: The authors declare that they have no conflicts of interest of any kind.

BIBLIOGRAPHY
1. Nordin BE, Roper A. Post-pregnancy osteoporosis; a syndrome? Lancet 1955;268:431-4.
2. Di Gregorio S, Danilowicz K, Rubin Z, Mautalen C. Osteoporosis with vertebral fractures associated with pregnancy and lactation. Nutrition 2000;16:1052-5.
3. Smith R, Athanasou NA, Ostlere SJ, Vipond SE. Pregnancy-associated osteoporosis. QJM 1995;88:865-78.
4. Smith R, Phillips AJ. Osteoporosis during pregnancy and its management. Scand J Rheumatol 1998;107:66-7.
5. Smith R, Stevenson JC, Winearls CG, Woods CG, Wordsworth BP. Osteoporosis of pregnancy. Lancet 1985;1:1178-80.
6. Stumpf UC, Kurth AA, Windolf J, Fassbender WJ. Pregnancy-associated osteoporosis: an underestimated and underdiagnosed severe disease. A review of two cases in short- and long-term follow-up. Adv Med Sci 2007;52:94-7.
7. Glerean M, Plantalech L. Osteoporosis en embarazo y lactancia. Medicina 2000;60:973-81.
8. Blanch J, Pacifici R, Chines A. Pregnancy-associated osteoporosis: report of two cases with long-term bone density follow-up. Br J Rheumatol 199433:269-72.
9. Carbone LD, Palmieri GM, Graves SC, Smull K. Osteoporosis of pregnancy: long-term follow-up of patients and their offspring. Obstet Gynecol 1995;86:664-6.
10. Iwamoto J, Sato Y, Uzawa M, Matsumoto H. Five-year follow-up of a woman with pregnancy and lactation-associated osteoporosis and vertebral fractures. Ther Clin Risk Manag 2012;8:195-9.
11. Vujasinovic-Stupar N, Pejnovic N, Markovic L, Zlatanovic M. Pregnancy-associated spinal osteoporosis treated with bisphosphonates: long-term follow-up of maternal and infants outcome. Rheumatol Int 2012;32:819-23.
12. Phillips AJ, Ostlere SJ, Smith R. Pregnancy-associated osteoporosis: does the skeleton recover? Osteoporos Int 2000;11:449-54.
13. Uematsu N, Nakayama Y, Shirai Y, Tamai K, Hashiguchi H, Banzai Y. Transient osteoporosis of the hip during pregnancy. J Nihon Med Sch 2000;67:459-63.
14. Sosa Henriquez M. Working group on clinical practice and protocols. Basic data on osteoporosis. Rev Esp Enf Metab Óseas 2000;9:84-5.
15. Cummings SR, Block G, McHenry K, Baron RB. Evaluation of two food frequency methods of measuring dietary calcium intake. Am J Epidemiol 1987;126:796-802.
16. Sosa M, Hernández D, Estévez S, Rodríguez M, Liminana JM, Saavedra P, et al. The range of bone mineral density in healthy Canarian women by dual X-ray absorptiometry radiography and quantitative computer tomography. J Clin Densitom 1998;1:385-93.
17. Genant HK, Wu CY, van Kuijk C, Nevitt MC. Vertebral fracture assessment using a semiquantitative technique. J Bone Miner Res 1993;8:1137-48.
18. Bret J. Un nouveau syndrome; osteoporose gravidique ou osteoporose post-gravidique. Presse Med 1955;63:1549.
19. Kovacs C, Kronenberg H. Pregnancy and Lactation. Primer on the Metabolic Bone Diseases and Disorders of Mineral Metabolism. 7th Edition. 2008;Cap.18:90-9.
20. Gruber HE, Gutteridge DH, Baylink DJ. Osteoporosis associated with pregnancy and lactation: bone biopsy and skeletal features in three patients. Metab Bone Dis Relat Res 1984;5:159-65.
21. He XD, Xia WB, Xing XP, Li M, Jiang Y, Wang O, et al. Clinical features of pregnancy and lactation-associated osteoporosis: analysis of 4 cases. Zhonghua Yi Xue Za Zhi 2009;89:983-5.
22. Liel Y, Atar D, Ohana N. Pregnancy-associated osteoporosis: preliminary densitometric evidence of extremely rapid recovery of bone mineral density. South Med J 1998;91:33-5.
23. Vandecandelaere M, Cortet B, Flipo RM, Duquesnoy B, Delcambre B. Osteoporosis in pregnancy: apropos of 2 cases. Rev Med Interne 1997;18:571-4.
24. O’Sullivan SM, Grey AB, Singh R, Reid IR. Bisphos phonates in pregnancy and lactation-associated osteoporosis. Osteoporos Int 2006;17:1008-12.
25. Cano-Marquina AJ, Cano A. Osteoporosis del embarazo. Semin Fund Esp Reumatol 2012. doi:10.1016/j.semreu.2011.12.001.