METHOD FOR THE TREATMENT OF WOMEN POSTMENOPAUSICS USING ULTRA-LOW DOSES OF ESTROGENBACKGROUND OF THE INVENTION Endogenous estrogens are dramatically reduced from natural or surgical menopause, and this decline results in a marked increase in bone loss and subsequent fractures. Endogenous estrogens are clearly important for maintaining skeletal health in young women. However, the importance of endogenous estrogens in older women is less certain. It has previously been shown that endogenous estrone levels in women over 65 years of age were not associated with subsequent hip or vertebral fractures. MR. Cummings et al., J. Bone Min. Res. 10 (Suppl. I): S174 (1995). However, estradiol is a more potent estrogen than estrone and studies of its relationship with fractures have been inconclusive. Serum estradiol levels in pre-menopausal women average more than 100 pg / ml. Serum estradiol levels in post-menopausal women who are not treated with hormone replacement therapy fall below 20 pg / ml, and as many as 30-50% of post-menopausal women have serum estradiol levels that are undetectable by conventional sensitive assay methods (ie less than 5 pg / ml). Conventional treatment for post-menopausal women includes estrogen replacement therapy in doses sufficient to maintain serum estradiol levels above 40-60 pg / ml. Conventional hormone replacement therapy has been shown to be useful for treating physical positions resulting from the decline of post-menopausal estrogen, including reducing the loss of, or even increasing the density of bones; and decrease the risk of bone fracture. However, studies have indicated that hormone replacement therapy may be linked with increased risk of cancer of the breast and endometrium, as well as blood clotting and bleeding from the uterus. Accordingly, there remains a need in the art for methods to treat the physical conditions that result from the decline or deficiency of post-menopausal estrogen. There is also a need in the art for treating these physical conditions while reducing the side effects of hormone replacement therapy. SUMMARY AND OBJECTIVE OF THE INVENTIONAs a first aspect, the present invention provides a method for treating physical conditions resulting from the decline of estrogens in a post-menopausal subject. The method comprises administering to the subject an amount of estrogen that is effective to produce a serum estradiol level of between about 5 pg / ml and 15 pg / ml. As a second aspect, the present invention provides a method for reducing the risk of osteoporotic bone fractures in a subject afflicted with or susceptible to post-menopausal osteoporosis. The method comprises administering to the subject, an amount of estrogen that is effective to produce a serum estradiol level of between about 5 pg / ml and about 15 pg / ml. As a third aspect, the present invention provides equipment for use by a consumer affected with or susceptible to physical conditions resulting from the decline of post-menopausal estrogen. The kit comprises (a) a transdermal patch capable of transdermally administering less than about 20 μg of estrogen per day; and (b) instructions describing a method for using the transdermal patch to reduce the risk of fracture or bone in the consumer. As a fourth aspect, the present invention provides another method for treating physical conditions resulting from the decline of estrogen in a post-menopausal person. The method includes administering less than about 20 μg of estrogen per day in the absence of exogenous progestin. As a fifth aspect, the present invention provides another method for reducing the risk of osteoporotic bone fractures in a person affected with or susceptible to osteoporosis. The method includes administering less than about 20 μg of estrogen per day in the absence of exogenous progestin. These and other aspects of the present invention are further described in the drawings, description of the preferred embodiment and examples of the invention that follow. BRIEF DESCRIPTION OF THE DRAWING Figure 1 is a cross-sectional view of a type of transdermal patch that can be employed in the methods of the present invention. Figure 2 is a cross-sectional view of a second type of transdermal patch that can be employed in the methods of the present invention. DESCRIPTION OF THE PREFERRED MODALITY Unless defined otherwise, all the technical and scientific terms used here have their conventional meaning in the art. As used herein, the following terms have the meanings ascribed to them. "Physical conditions that result from the decline of post-menopausal estrogen," refers to physical conditions that are common among post-menopausal women and that are caused at least in part by a decline in estrogen in the body. These conditions include but are not limited to osteoporosis, headaches, nausea, depression, hot flashes, decreased bone mineral density.; and increased risk or incidence of bone fracture, including vertebral and / or hip fracture. "Post-menopausal subject" refers to women in the period of life after menopause. Affected subjects with post-menopausal symptoms include women after menopause who exhibit any of the above physical conditions after menopause and particularly women after menopause who exhibit decreased mineral density in bones, vertebrae, hips or other site, or who have experienced either vertebral or hip fracture. Subjects susceptible to physical conditions resulting from post-menopausal decline in estrogen, include women approaching the onset of menopause, who exhibit a decrease in serum estradiol levels, compared to pre-menopausal women, and women after menopause that exhibit a decrease in serum estradiol levels, but that have not yet exhibited physical conditions caused by post-menopausal estrogen decline. Subjects exhibiting decreased serum estradiol levels include subjects exhibiting serum estradiol levels at or below 20 pg / ml, including subjects exhibiting undetectable levels of estradiol in serum. For purposes of this invention, sex hormone levels, including serum estradiol levels, were measured using radioimmunoassay after extraction and column separation. The lower limit of detectability is 5 pg / ml of estradiol. "Osteoporotic bone fractures" refer to bone fractures, typically in the vertebrae or hip, for which osteoporosis is a contributing factor. The methods of the present invention are useful for the treatment of post-menopausal subjects, particularly subjects affected with or susceptible to post-menopausal physical conditions of the type discussed above. The methods of the present invention involve the administration of estrogen in an amount effective to produce the desired serum estradiol level in the subject. As used herein the phrase "treat physical conditions" contemplates eliminating or reducing the severity or incidence of these physical conditions in the subject affected with these conditions, and also avoiding the occurrence of post-menopausal physical conditions in a subject susceptible to these conditions, as a result of the decline of post-menopausal estrogen. Although the treatment of these post-menopausal physical conditions may include the complete elimination of these conditions in an affected subject, the complete elimination of the condition is not required to meet the definition of the term contemplated by the present invention. Thus, the present invention involves the use of ultra low doses of estrogen for the treatment of physical conditions resulting from estrogen decline and to reduce the risk of osteoporotic bone fractures in a subject afflicted with or susceptible to post-menopausal osteoporosis. . The methods of the present invention may also include the additional step of testing the estradiol level in the serum of the post-menopausal subject to be treated and determining that the estradiol level in the subject's serum is normal for post-menopausal women in the same age group as the subject. In other words, the methods of the present invention are not reserved for the treatment of women having lower than normal levels of serum estradiol, than the average of post-menopausal women in the same age group. The methods of the present invention can be used to treat post-menopausal subjects whose serum estradiol is normal for post-menopausal women of the same age group. As used herein, a "normal" level of estradiol in serum is a serum estradiol level that is relatively close to the level of estradiol in the average serum for women of the same age group. The present inventors have unexpectedly discovered that loss of mineral density in bones is not the only contributing factor that leads to an increased rate of osteoporotic bone fractures in subjects afflicted with post-menopausal osteoporosis. Low levels of estradiol in serum, ie less than 5 pg / ml, especially when accompanied by levels of sex hormone binding globulin (SHBG) of 1 μg / dl or more, substantially increase the risk of hip and vertebral fracture. In addition, low serum levels of 1,25 (OH) 2 vitamin D also lead to increased risk of hip fractures. The method of estimating the risk of osteoporotic bone fractures involves the use of logistic models developed by SAS Institute Cary, North Carolina, to analyze the relationship between predictors and vertebral fractures, and proportional hazards models that take into account the design of the sample case -cohorte, to analyze the relationship in their forecasters and hip fractures. The analyzes are adjusted for baseline age and weight, and report as relative risks with 95% confidence intervals. The proportion of fractures attributed to various levels of hormones is estimated using the technique described in W.S. Browner, American Journal of Epidemiology 123: 143 (1986), the description of which is hereby incorporated by reference in its entirety. By "reducing the risk of osteoporotic bone fracture" it is meant that the risk, as measured using the prior techniques, is less for a particular subject that is treated with the methods of the present invention, as compared to the risk for the same. subject before treatment using the methods of the present invention. In this way, the risk of osteoporotic bone fracture is reduced for a subject treated with the methods of the present invention compared to an untreated post-menopausal subject affected with those susceptible to osteoporotic bone fractures.
The present inventors also unexpectedly discovered that the treatment of physical conditions resulting from the decline of estrogen can be affected by ultra-low doses of estrogen without the need for progestin administration. The administration of estrogen, excluding the administration of progestin has now been found by the present inventors to be effective for the treatment of post-menopausal women who have uterus and ovaries. Previous hormone replacement therapies have been based on co-administration of progestin for effective treatment, particularly women who have not undergone hysterectomy or ovariectomy. The source of exogenous estrogen for use in the methods of the present invention may include any convenient form of estrogen for administration to a subject. Suitable forms of exogenous estrogen include both natural and synthetic compounds, which exhibit estrogenic activity. Various forms of exogenous estrogen are commercially available. For example, convenient forms of exogenous estrogen include but are not limited to estradiols, including -estradiol, 17β-estradiol, ethinyl estradiol, estradiol benzoate, and estradiol 17β-cypionate; estrone; estriol; conjugated equine estrogens; and salts of the previous ones. The above are all examples of steroids that exhibit estrogenic activity. Examples of non-steroidal compounds that exhibit estrogenic activity include but are not limited to diethyl ilbestrol diphosphate, diethylstilbestrol dipropionate, and hexestrol.
Currently, the preferred form of exogenous estrogen for use in the methods of the present invention is estradiol. The amount of exogenous estrogen to be administered to the subject is sufficient to achieve a serum estradiol level of at least about 5 pg / ml, but not more than about 20 pg / ml and preferably not more than 15 pg / ml. In other words, according to the methods of the present invention, sufficient exogenous estrogen is administered to achieve a total serum estradiol level of at least about 5 pg / ml to about 20 pg / ml. Since the serum estradiol level of an untreated subject will differ for each individual, different individuals may require administration of different doses of estrogen, to achieve the required serum estradiol level. It is not required that the serum estradiol level of each subject to be treated be increased between about 5 and about 20 pg / ml; instead, the level of estradiol in the total serum of each treated subject should be at least about 5 and not more than about 20 pg / ml. Often, the amount of estrogen to exogenous to be administered is sufficient to achieve a serum estradiol level of between about 5 pg / ml and about 10 pg / ml.
Contrary to prior understanding, the present inventors have now discovered that serum estradiol levels of between 5 pg / ml and 15 pg / ml, advantageously produce only at the risk of vertebral and hip fracture. Administration of this lesser amount than the conventional exogenous estrogen has the additional advantage of decreasing the risk of undesirable side effects associated with hormone replacement therapies. The administration of exogenous estrogen can be achieved by any convenient route. For example, formulations for oral and parenteral administration of exogenous estrogen are known in the art, and can be employed in the methods of the present invention. Formulations suitable for oral administration may be present in discrete units, such as capsules, amylaceous capsules, troches, or tablets, each containing a predetermined amount of the active compound, such as a powder or granules; or as a solution, dispersion or suspension in an aqueous or non-aqueous liquid. These formulations can be prepared by any convenient pharmacy method that includes the step of bringing into association the active compound and a convenient carrier (which may contain one or more accessory ingredients). In general, the formulations of the invention are prepared by uniform and intimate mixing of the active compound with a liquid or solid carrier finally divided, or both, and then it is necessary to shape the resulting mixture. For example, a tablet can be prepared by compressing or molding a powder or granules containing the active compound, optionally with one or more accessory ingredients. Compressed tablets can be prepared by compressing, in a convenient machine, the compound in a free-flowing form, such as a powder or granules optionally in admixture with a binder, lubricant, inert diluent and / or dispersing agents / surfactants. Molded tablets can be made by molding in a convenient tabletting machine, the powdery compound is wetted with an inert liquid binder. The amount of exogenous estrogen in the oral formulation is an ultra-low dose of estrogen that will depend on the precise form of estrogen to be administered, but typically less than 0.5 mg / per day. Preferably, the amount of estrogen to be administered orally is between about 0.1 mg and about 0.25 mg of estrogen per day. For example, the amount of estradiol administered orally is from about 0.1 mg to about 0.25 mg per day. It is well within the skill of those in the specialty to determine the equivalent doses of other forms of estrogen equally. In the preferred embodiments of the present invention, estrogen is administered parenterally or transdermally rather than orally. The above routes of administration are preferred over oral administration, because oral administration of estrogen can lead to increased levels of sex hormone binding globulin. Globulin that binds sex hormone can decrease the beneficial effects of administering estrogen to post-menopausal subjects, particularly in subjects who exhibit signs of osteoporosis or loss of mineral density in bones. Although oral administration is not the preferred route, the methods of the present invention can be carried out using oral formulations. Formulations of the present invention suitable for parenteral administration, conveniently comprise sterile aqueous preparations of the active compound, these preparations are preferably isotonic with the blood of the intended recipient. These preparations can be administered by intravenous, intramuscular, intradermal or vaginal ring subcutaneous injection. These preparations can be conveniently prepared by mixing the active ingredient, an estrogen, with water or a glycine buffer and making the resulting solution sterile and isotonic with the blood. The amount of estrogen to exogenous in the parenteral formulation is an ultra-low dose of estrogen that will depend on the precise form of estrogen to be administered, but typically not more than 20 μg per day. Preferably, the amount of estrogen administered parenterally is between about 5 μg and about 15 μg of estrogen per day, and more preferably about 10 μg of estrogen per day. For example, the amount of estradiol administered parenterally is from about 5 μg to about 15 μg per day. It is well within the skill of those in the specialty to determine equivalent doses of other forms of estrogen equally. More preferably, the methods of the present invention include the transdermal administration of exogenous estrogen. Suitable formulations for the transdermal administration of estrogen are known in the art, and can be employed in the methods of the present invention. For example, convenient transdermal patch formulations for the administration of exogenous estrogen are described in U.S. Pat. No. 4,460,372 issued to Campbell et al., U.S. Pat. No. 4,573,996 granted to Kwiatek et al., The Patent of theE.U.A. No. 4,624,665 issued Nuwayser, U.S. Pat. No. 4,722,941 issued to Eckert et al., And US Pat. No. 5,223,261 issued to Nelson et al., The descriptions of which are hereby incorporated by reference for their discussion of transdermal patch technology. A convenient type of transdermal patch for use in the methods of the present invention is illustrated in Figure 1. In general, a convenient transdermal patch 10 includes a backing layer 12 that is not permeable, a permeable surface layer 13, a layer of adhesive (not shown) that substantially continuously coats the permeable surface layer 13, and a reservoir 16 located or sandwiched between the backing layer 12 and the permeable surface layer 13, such that the backing layer 12 extends around the sides of the reservoir 16 and joins the permeable surface layer 13 at the edges of the permeable surface layer 13. The reservoir 16 contains estrogen and is in fluid contact with the permeable surface layer 13. The transdermal patch 10 adheres to the skin by the adhesive layer in the permeable surface layer 13, such that the permeable surface layer 13 is in substantially continuous contact with the skin when the patch Transdermal 10 adheres to the skin. While the transdermal patch 10 adheres to the skin of the subject, the estrogen contained in the reservoir 16 of the transdermal patch 10 is transferred by the permeable surface layer 13, from the reservoir 16 through the adhesive layer yaya through the skin. of the subject. The transdermal patch 10 optionally may also include one or more reservoir penetration enhancing agents 16, which improve the penetration of estrogen through the skin. Examples of suitable materials that can comprise the backing layer are well known in the specialty of transdermal patch delivery and any conventional backing or backing material can be employed in the transdermal patch of the present invention. Specific examples of suitable backing layer materials include but are not limited to polyester film, such as high density polyethylene, low density polyethylene, or polyethylene compounds; polypropylene, polyvinyl chloride, polyvinylidene chloride; ethylene vinyl acetate copolymers; and similar. Examples of suitable permeable surface layer materials are equally well known in the specialty of transdermal patch delivery, and any conventional material permeable to the active ingredient to be administered, ie, estrogen, can be employed in the transdermal patch of the present invention. invention. Specific examples of materials suitable for the permeable surface layer include, but are not limited to, dense or microporous polymer films, such as those comprising polycarbonates, polyvinyl chlorides, polyamides, modacrylic copolymers, polysulfones, halogenated polymers, polychloroethers, acetal polymers, acrylic resins, and the like. Specific examples of these types of conventional permeable membrane materials are described in U.S. Pat. No. 3,797,494 granted to Zaffaroni. Examples of suitable adhesives that can be coated on the backing layer to provide the adhesive layer are equally well known in the art and include, for example, pressure sensitive adhesives, such as those comprising acrylic and methacrylic polymers. Specific examples of suitable adhesives include polymers of acrylic and methacrylic acid esters (e.g., n-butanol, n-pentanol, isopentanol, 2-methyl butanol, 1-methyl butanol,1-methyl pentanol, 3-methyl pentanol, 3-methyl pentanol, 3-ethyl butanol, isooctanol, n-decanol, or n-dodecanol esters thereof) alone or copolymerized with ethylenically unsaturated monomers such as acrylic acid, methacrylic acid, acrylamide, methacrylamide, N-alkoxymethyl acrylamides, N-alkoxymethyl methacrylamides, Nt-butylacrylamide, itaconic acid, vinyl acetate, N-alkyl acids with 10 to 24 branched-maleamic carbon atoms, glycol diacrylate, or mixtures of the above; natural or synthetic rubbers, such as silicone rubber, styrene-butadiene rubber, butyl ether rubber, neoprene rubber, nitrile rubber, polyisobutylene, polybutadiene and polyisoprene; polyurethane elastomers; vinyl polymers such as polyvinyl alcohol, polyvinyl ethers, polyvinyl pyrrolidone and polyvinyl acetate; urea formaldehyde resins; phenol formaldehyde resins; resorcinol formaldehyde resins; cellulose derivatives such as ethyl cellulose, methyl cellulose, nitrocellulose, cellulose acetate butyrate, and carboxymethyl cellulose, and natural gums such as guar, acacia, pectin, starch, destria, gelatin, casein, etc. As will be apparent to those skilled in the art, the adhesive layer should be inert to the active ingredient, estrogen, and should not interfere with the transdermal delivery of estrogen through the permeable surface layer. Pressure sensitive adhesives are preferred for the adhesive layer of the transdermal patch, to facilitate application of the patch to the skin of the subject. Suitable penetration enhancing agents are well known in the art as well. Examples of conventional penetration enhancing agents include alkanols such as ethanol, hexanol, cyclohexanol, and the like; hydrocarbons such as hexane, cyclohexane, isopropylbenzene; aldehydes and ketones such as cyclohexanone, acetamide; N, N-di (lower alkyl) acetamides such as N, N-diethylacetamide, N, N-dimethyl acetamide; N- (2-hydroxyethyl) acetamide; esters such as N, N-di-lower alkyl-sulfoxides; essential oils such as propylene glycol, glycerin, glycerol monolaurate, isopropyl myristate, and ethyl oleate; salicylates; and mixtures of any of the foregoing.
Figure 2 is an example of a second type of transdermal patch that is suitable for the transdermal delivery of estrogen through the present invention. In this example, the active ingredient is incorporated into the adhesive layer instead of being contained in a tank. Examples of these types of patches are conventionally known and include for example the Climera ™ patch available from Berlex. The transdermal patch 20 comprises a backing layer 22 and an adhesive / drug layer 24. The adhesive / drug layer 24 has the combined function of adhering the patch 20 to the subject's skin and containing the active ingredient estrogen that is going to manage. The active ingredient is leached from the adhesive / drug layer 24 and through the skin of the subject when the patch adheres to the skin. Any of the previously described backing layers can be used in this mode equally. In addition, any of the convenient adhesives described above can be employed. The adhesive / drug layer comprises a relatively homogeneous mixture of the selected adhesive and the active ingredient. Typically, the adhesive / drug layer comprises a coating that substantially covers one surface of the backing layer. The adhesive / drug layer may also include a penetration enhancing agent such as those described above by incorporating the penetration enhancing agent into the substantially homogeneous mixture of the adhesive and the active ingredient. As will be readily apparent to those skilled in the art, transdermal patches according to the present invention may include a variety of additional excipients that are conventionally employed to facilitate the transdermal administration of an active agent, particularly a spheroidal active agent. Examples of these excipients include but are not limited to carriers, gelling agents, suspending agents, dispersing agents, preservatives, stabilizers, wetting agents, emulsifying agents and the like. Specific examples of each of these types of excipients are well known in the art and any conventional excipients can be employed in the transdermal patches of the present invention. It is important to note, however, that the transdermal patches of the present invention exclude progestin. Accordingly, progestin is not a convenient excipient for use in the transdermal patch formulations of the present invention.
The amount of exogenous estrogen in the transdermal patch formulations is an ultra-low dose of estrogen, which depends on the precise form of estrogen to be administered, but is sufficient to deliver less than 20 μg, and typically not more than 15 μg per day . Preferably, the amount of estrogen administered by the transdermal patch is between about 5 μg and about 15 μg of estrogen per day. More preferably, the amount of estrogen administered is about 10 μg per day. Although the typical dose of estrogen according to the method of the present invention is less than 20 μg, doses as high as 25 μg may be employed. For example, the amount of estradiol administered parenterally is from about 5 μg to about 15 μg per day. It is well within the skill of those in the specialty to determine equivalent doses of other forms of estrogen equally. The ultra-low level of estrogen employed in the methods of the present invention has unexpectedly been found to substantially reduce the risk of osteoporotic bone fractures in post-menopausal women. Typically, transdermal patches are designed to be worn for several days before replacement is required. In this way, the amount of estrogen in the reservoir should be sufficient to allow administration of less than 20 μg per day for a period of several days. As an example, a transdermal patch according to the present invention which is designed to administer 10 μg of estrogen per day, for seven (7) days will contain about 1 mg of estrogen. A suitable patch for the administration ofμg per day for seven (7) days will only contain approximately 1.4 mg of estrogen. Based on these specific examples, a person skilled in the art will be able to discern the necessary amount of estrogen to include in the transdermal patch to achieve the supply of the correct daily dose of estrogen. Advantageously, the present invention also provides equipment for use by a consumer affected with or susceptible to post-menopausal symptoms including the transdermal patch and instructions describing the method for using the transdermal patch to treat post-menopausal symptoms and / or reducing the risk of osteoporotic bone fracture in the consumer. The instructions will instruct the consumer to adhere the transdermal patch using the adhesive, directly on the surface of the skin, for example on the upper arm, to achieve transdermal administration of the ultra-low dose of estrogen from the patch and in this way increase the estradiol level in the serum in the consumer between approximately 5 pg / ml and approximately 20 pg / ml. The instructions also direct the consumer to replace the patch as required, to continue the administration of estrogen as necessary to maintain this level of estradiol in serum when using the transdermal patch. In particular, the instructions may direct the consumer to replace the transdermal patch every seven (7) days to ensure administration of less than 20 μg and preferably 10 μg of estrogen per day, when a patch of seven (7) days is given. used in the team. These equipment can be advantageously packaged and sold in units of single or multiple patches. The following example is provided to illustrate the present invention and shall not be considered as limiting thereof. Example 1 The following example and data demonstrate the co-relationship between serum estradiol level and risk of osteoporotic bone fracture and also demonstrates the efficacy of using an ultra-low dose of exogenous estrogen, to reduce the risk of osteoporotic bone fracture. and for the treatment of post-menopausal symptoms. 1.- Experimental population The study of osteoporotic fractures is a prospective study of a group of 9,704 women, who were recruited from population-based lists in four communities in the U.S.A. : Baltimore, Minneapolis,Pittsburgh and Portland. See, Cummings et al., N.
Enl. J. Med. 332: 767-773 (1995). Women aged 65 years or older were invited by mail to participate in the study. The experimental population excluded black women, due to their low risk of hip fracture and women who had bilateral hip replacement or who required the assistance of another person to walk. Participants were asked about the current or recent use of estrogen, calcium supplements and multivitamins containing vitamin D. 2. Clinical measurements The density of minerals in bone (ie bone mass of the calcaneus) was measured using absorptiometry. simple photon obtained from the Osteo-Analyzer, Siemens-Osteon, Wahiawa, Hawaii. The average coefficient of variation of this measure between clinical centers was 1.2%. See Steiger, et al., J. Bone Miner. Res. 7: 625 (1992).3. - Serum samples All participants were instructed to comply with a fat-free diet during the night and the morning before the test to minimize lipemia that may interfere with the trials. Blood samples were taken between 8:00 A.M. and 2:00 P.M. and the serum was immediately frozen at -20 ° C. All samples were shipped in two weeks to the Biomedical Research Institute in Rockville, Maryland, where they were stored in liquid nitrogen at -190 ° C. 4. - Evaluation of fractures Women were contacted every four months by postal card or periodic clinical visits: the follow-up for vital status and occurrence of hip fractures was more than 99.7% complete. Hip fractures were confirmed by a review of per-operative radiographs. An average of 3.3 years after baseline radiographs were taken, 7,299 women (79% of survivors) returned for repeated spine films; 7,238 pairs of radiographs were judged adequate for evaluation of incident vertebral fractures. We estimate radiographs for fractures using quantitative morphometry as described in Black et al., J. "Bone Miner, Res. 10: 890 (1995) .A woman was classified as having an incident vertebral fracture if any vertebral height decreased by> 20%. and.> 4 mm between baseline and follow-up radiography See National OsteoporosisFoundation Working Group on Vertebral Fractures (Working Group of the National Osteoporosis Foundation inVertebral Fractures), J. Bone Miner. Res. 10: 5 18(nineteen ninety five) . Those who estimated the radiographs were not aware of the results of the participants' serum assays. 5.- Selection of case and group samples Using the case-cohort approach described inPrentice, R.L., Biometrika 73: 1 (1986), and excludingwomen who used hormone replacement therapy in baseline, we randomly selected serum samples from 133 participants who subsequently had a first hip fracture and 138 from the participants who had an incident vertebral fracture. We also selected a random sample of 359 women in the group, including 12 women who subsequently had a hip fracture and 14 who subsequently had vertebral fractures. For analysis of hip fracture, we excluded controls (n = 4) who had hip fractures before the baseline visit; for analysis of vertebral fracture, we excluded controls (n = 8) that did not have a follow-up radiography or whose baseline or follow-up radiography was technically inadequate for morphometry. 6.- Biochemical analyzes We determined stability of selected types of hormone measurements in 51 post-menopausal women when testing baseline levels and again after 3.5 years of storage at -190 ° C. The co-relationships(all significant at P <.001) between the two measurements were as follows: parathyroid hormone middle region (r = .99), 25 (OH) vitamin D (r = 0.88), testosterone (r =0. 99) and estrone (r = 0.98). There were few declines, if any, in the average values for this substance. An immuno radiometric assay (IRMA) for parathyroid hormone was not available when baseline measurements were made; The correlation between levels measured by IRMA and by the middle region technique was0. 78. All trials were conducted in mixed batches of cases and controls, blinded to a history of participant fracture. Samples were sent directly from storage to the analytical laboratory without thawing. Initial plans required analysis of estrone levels, but not estradiol. After completing the first batch of trials, estradiol was added to the panel of measurements and these measurements were made in 89 cases of hip fracture, 96 cases of vertebral fracture and 204 women who did not have any type of fracture.'fracture. Sex hormones and sex hormone binding globulin (SHBG), were tested by EndocrineSciences of Calabases Hills, California. Estradiol and estrone were measured by radio immunoassay after extraction and separation on 1 H20 column. Inter-assay variability for estradiol ranged from 8% to 12.5% and from 6.2 to 7.0% for estrone. The lower limit of detectability was 5 pg / ml for estradiol. Total testosterone was measured by radio immunoassay after extraction and column chromatography of Al203 with inter-assay variability from 6.1 to 13.4%. The binding capacity of SHBG was measured directly in serum using a displacement technique with an inter-assay variability of 4.1% to 14.4%. The concentration of free testosterone is determined from an ammonium sulfate precipitation process, as described in Furuyama et al., Esteroids (Steroids) 16: 415 (1970); YMayes, J. Clin. Endocrinol Metab. 28: 1169 (1968). The inter-assay variability was in the range of 10.7% to 15.5%.
Calciotropic hormones were tested by theUCSF Calciotropic Hormone Reference Laboratory of SanFrancisco, California The intact parathyroid hormone was measured by radio immunoassay, with an intra-assay variability of 5.2% and an inter-assay variation of 8.5%.
We measured 25 (OH) vitamin D and 1,25 (OH) 2 vitamin D, using radio immuno assay techniques, with an intra-assay variability of 6.6% for both and inter assay variabilities of 15% and 24.4%, respectively. Creatinine levels were measured using an automated chemistry analyzer. 1 . - Data analysis Hormone and vitamin levels were analyzed as continuous variables, by quintiles and when relevant as dichotomous variables (above and below the detection threshold). Logistic models developed by SAS Institute of Cary, North Carolina, were used to analyze the relationship between predictors and vertebral fractures and proportional hazards models that take into account the case-cohort sampling design (Epicure, Hirosoft International, Seattle, WA) for analyze the relationship between predictors and hip fracture. All analyzes were adjusted for age and baseline weight, and are reported as relative risks (approximately as hazard ratios or risk ratios) with 95% confidence intervals. The proportion of fractures attributed to various hormone levels was estimated with standard techniques described by Browner, Am. J. Epidemiol. 123: 143 (1986). We compared bone mass adjusted for age in different groups using general linear models. 8. - Results Table 1 establishes the data collected regarding the characteristics of participants who suffered incident hip or vertebral fractures. Table 1 Characteristics of participants with incidental hip fractures, incident vertebral fractures and controls with no results.
Vertebral Fracture Hip Fracture Sample Cases Cases Sample cohort fracture * cohort fracture *Characteristic (n = 138) (n = 264) (n = 133) (n = 343)Average age 73.2f 71.6 75.3f 72.1(years) Average Weight (kg) 63.5f 68.6 61.9f 68.2 Vertebral Fracture Hip Fracture Sample Cases Cases Samplefracture of decohort * fractur cohort *Current smoker of 12% 8.8% 17% í 9.1%Cigarettes Current use of 53% f 42% 51% 42%supplement ofCurrent calcium use 43% 42% 43% 48% supplementvitamin D calcium intake 695 693 659 689 per average diet (mg / day) bone mass 0.35f 0.41 0.34f 0.41Average calcaneus (g / cm2)Number of controls for hip fracture analysis varies from 240 to 343 due to missing data per estradiol test. The number of controls for vertebral fracture varies from 193 to 264; it is smaller due to the exclusion of those with technically adequate vertebral x-ray baseline and follow-up pairs. Incident fracture cases plotted with the cohort sample are included with the cases. t P < .01 compared with controls. $ P < .05 compared with controls. The data show that participants who suffered incidental hip or vertebral fractures were older, weighed less, more likely were smokers, and had lower bone mass than controls. Approximately one in three (81 or 247) women in the random sample of the group had undetectable estradiol levels (< 5 pg / ml). The data regarding the relative risk, adjusted for weight and age are established in Table 2 below. Table 2 Associations adjusted for age and weight between hormone and vitamin levels and risk of hip fracture and vertebral fracture.
Relative risk 95% C.l. (*) Variable Predominance Fracture Fracture(%) t Vertebral hipTotal of 32.8 2.48 (1.35- 2.46 (1.43-estradiol < 5 4.55) 4.22)pg / ml 95% relative risk C.l. (*)Variable Predominance Fracture Fracture (%) t Vertebral hipGlobulin that 76.4 2.04 (1.06- 2.27 (1.19-hormone link 3.85) 4.35)of sex > 0.9μg / dl Globulin 1.25 (0.97- 1.49 (1.15-hormone ligament 1.60) 1.95)of sex (by1. 0 μg / dl) Estrone 1.5 77.4 1.30 (0.76- 1.72 (1.01- ng / dl 2.17) 2.94) Testosterone 19.4 1.63 (1.00- 1.40 (0.83-free <. 0.7 2.66) 2.35) pg / ml(OH) 21.5 1.16 (0.70- 1.08 (0.64-vitamin D < 1.92) 1.80)19 ng / ml ± 1.25 (OH) 2 15.4 2.05 (1.21- 1.60 (0.92-vitamin D. < 3.46) 2.79)23 pg / ml § intact PTH 11.2 1.05 (0.52- 0.49 (0.20-greater than 50 2.12) 1.17) pg / ml * Relative risks estimated as risk ratios for vertebral fractures or proportions of risk for hip fractures. C.l. = confidence interval. f Predominance in the random sample of the group. ± To convert values of 25 -hydroxy vitaminD to p-nanomoles per liter, multiply by2. 496. § To convert values of 1, 25-dihydroxy vitamin D to picomoles per liter, multiply by 2. Four . The data show that women with undetectable levels of estradiol had a 2.5 times higher risk of hip fracture and vertebral fracture than women who had measurable levels of this hormone.
We estimate that 33% of hip fractures and 32% of vertebral fractures in this group were associated with undetectable estradiol levels. The data appear to reflect a decreased risk threshold associated with estradiol concentrations less than or equal to 5 pg / ml.
Women with total estradiol levels of 5 to 10 pg / ml had a 62% lower risk of hip fractures(risk ratio equal to 0.38, 0.20 to 0.72) and a 57% lower risk of vertebral fractures (odd ratios = 0.43, 95% confidence interval, 0.25 to 0.76) than women with levels below 5 pg / ml. Higher levels of SHBG were associated with increased risk of hip and vertebral fractures.
After adjusting for age, each increase of 1.0 μg / dl in SHBG was associated with an increase of 1.44 times (1.16 to 1.80) in the risk of hip fracture and an increase of 1.65 times (1.30 to 2.10) in the risk of vertebral fracture. For hip fracture, this effect appears as partially mediated by weight. In a random sample of the group, 26% (63 of 244) had both an undetectable estradiol level and a SHBG level greater than or equal to 1 μg / dl; these women had a 14 times greater risk of hip fracture(95% confidence interval, 3.0 to 62) and a greater risk 12 times (3.3 to 41) of vertebral fracture than other women. The weight adjustment paved these associations somewhat for hip fracture (risk ratio = 6.9, 1.5 to 32) and for vertebral fracture (odd ratios = 7.9, 2.2 to 28). Associations between estradiol, SHBG and incident vertebral fractures were similar in the subgroups of women who had a baseline vertebral fracture and those who did not. Attributable risk for 26% of women with this combination was 60% for hip fractures and 64% for vertebral fractures. The data in Table 2 regarding vertebral fractures are expanded by the data set set forth in Table 3 below. Table 3 Predictors of vertebral fracture: muti-variable models *Odd proportion (95% interval ofconfidence) Variable (unit) Multivariable Also adjustedadjusted for bone mass Estradiol < 5 pg / ml 2.65 (1.50-4.68) 2.34 (1.30-4.22) SHBG (per 1 μg / dl of 1.62 (1.16-2.26) 1.51 (1.07-2.12)increment) Estrona 1.5 ng / dl 2.04 (1.03-4.17) 2.33 (1.12-4.76) Age (per increment 1.17 (0.89-1.54) 1.09 (0.82-1.44)years old) Weight (per increment 0.71 (0.53-0.94) 0.86 (0.64-1.16)years old) Bone mass 2.27 (1.52-3.33)calcaneus (fordecrease SD) tAll results are adjusted for other variables in the table. SD (standard deviation) of calcaneus bone is0. 098 g / cm2.t Mass of calcaneus bone is measured in baseline.
The data in Table 2 show that women with estrone values in the lowest quintile (< 1.4 ng / dl) had lower risk of vertebral fracture than women with higher levels. Although women with lower estrone tended to lower estradiol levels, the correlation between estrone and estradiol levels was modest (r = 0.57) and low estrone levels remain associated with a decreased risk of vertebral fracture after adjustment for estradiol levels, such as is illustrated in Table 3. Women whose vitamin D 1,25 (OH) 2 level was in the lowest quintile (< 23 pg / ml [< to 55.2 pm / L]) had a significantly increased risk of hip fracture (risk ratio = a 2.1, 1.2 or3. 5) as illustrated in Table 4 below. Table 4Predictors of hip fracture: multi modelsvariablesProportion of risk (95% interval ofconfidence) Variable (unit) Multivariable Also adjusted adjusted for bone mass Estradiol < 5 pg / ml 2.37 (1.27-4.45) 1.92 (1.01-3.64) Proportion of risk (95% range ofconfidence) Variable (unit) Multivariable Also adjustedadjusted for bone massSHBG (by 1 μg / dl of 1.34 (0.92-1.95) 1.25 (0.85-1.84)increase) 1,25 (OH) 2 vitamin 2.20 (1.00-4.80) 2.16 (0.93-5.00)D < . 23 pg / ml § Age (by increment 2.22 (1.50-3.28) 2.09 (1.41-3.10)years old) Weight (per increment 0.56 (0.39-0.79) 0.67 (0.46-0.98)years old) Bone mass 1.61 (1.03-2.56)calcaneus (fordecrease SD) f* All results are adjusted for other variables inthe board. SD (standard deviation) of bone masscalcaneus is 0.098 g / cm2§ To convert values for 1,25 dihydroxy vitamin D topicomoles per liter, multiply by 2.4These associations remained significant after adjusting estradiol and SHBG levels (risk ratio = 2.2, 1.0 to 4.8). Additional adjustment for serum creatinine did not represent difference(risk ratio = 2.3, 1.0 to 5.2).
After excluding women taking vitamin D supplements, the average 25 (OH) vitamin D levels plus or minus SD (were significantly higher in samples plotted from January to June (24.4 ± 10.2 ng / ml [83.6 ± 25.5 nm / L]) than from July to December (25.5 ± 10.4 ng / ml [63.6± 25.9 nm / L], P < .01). There were no significant differences when vitamin D levels in April-September were compared with those charted in October-March. There were no statistically significant associations between low serum levels of 25 (OH) vitamin D and the risk of hip or vertebral fractures (see Table 2), even after adjusting the month, season or clinic, whether women taking vitamin D supplements were included or not. Adding bone mass to multivariable models of hormones and the risk of fracture only slightly weakened the strength of association between estradiol, SHBG and risk of hip or vertebral fracture. See Tables 3 and 4. The adjustment for bone mass had no substantial effect on the strength of association between 1,25 (OH) 2 vitamin D and hip fracture or between estrone and vertebral fracture. The combined data indicated that the risks of hip and vertebral fractures in women over 65 years increase substantially when serum estradiol concentrations are belowpg / ml, the detection limit of the assay.
Approximately one third of incidental hip and vertebral fractures can be attributed to these extremely low levels of estradiol. Women with slightly higher levels of estradiol, in the range of 5 to 10 pg / ml, had a much lower risk of these fractures. Women with undetectable estradiol concentrations also had lower bone mass. However, adjusting the calcaneal mass only modestly attenuates the increased risk of fracture. This suggests that the association between very low estradiol levels and the risk of fracture is at least partially independent of bone mass. SHBG levels also exert a strong effect independent of the risks of hip and vertebral fractures. Women with SHBG concentrations greater than or equal to 1.0 μg / dl and undetectable estradiol levels had a 7 times higher risk of hip and vertebral fracture than women with lower SHBG and higher levels of estradiol. SHBG is avidly linked to circulating estradiol, such that increased levels of SHBG will be associated with decreased bioavailability of estradiol. In this way, bio-available estradiol concentrations may even be more strongly associated with fractures than total estradiol levels that we measured in this research. EXAMPLE 2 This example demonstrates a comparison of the effects of administering different amounts of estrogen using a 7-day transdermal estrogen therapeutic system to prevent bone loss in post-menopausal women. 1. - Methods Healthy women, hysterectomized, 45 to 65 years of age, 1 to 5 years after menopause with a bone mineral density baseline (BMD) measured by dual energy x-ray abscessiometry (DXA) a L2-L4 with 2.5 SD of the average for normal women under 45 years of age were eligible. The subjects were randomized to one to four doses of estradiol TTS (patches of 6.5, 12.5, 15 and 25 cm2 that provide .025, .05, .06 and .1 mg / day of estradiol, respectively) or placebo. One patch per week was used on the abdomen for the duration of the two-year trial. BMD to L2-L4, the femoral neck and forearm, serum osteocalcin (OST) and creatinine / urinary deoxypyridinoline (DPD) interlacing were measured every six months. The results after the 18-month treatment is presented. 2. - Results 176 women, with average age of 51.2 years were enrolled. A total of 103 subjects completed 18 months of treatment. The loss of BMD is avoided by all doses at all points in time. At 18 months, patients treated with placebo had a change in average percentage in lumbar BMD for -0.7 + 4.1 vs. 3.6 ± 4.7, 3.2 ± 3.1, 3.0 ± 3.6 and 4.8 ± 5.5 groups 0.025, and 0.05, 0.06, 0.1 mg / day, respectively (all differences p <0.05 vs. placebo). At 12 months, OST levels fell from baseline in all active treatment groups (percentage changes of -30.4 ± 42.9, -17.9 ± 50.7, -13.0 ± 37.6 and -26.9 + 29.8, respectively, all differences p <0.05 VS Placebo + 27.6 ± 60.4). DPD also fell from baseline to 12 months in all active treatment groups. Changes of -5.8 ± 3.4, -6.0 ± 8.5,-5.1 ± 7.4 and -1.8 + 10.8 nmol / mmol, respectively). Three of the four active treatments were significantly different from placebo. However, DPD also decreased in subjects treated with placebo.3. - Conclusions Transdermal estradiol of both currently available concentrations of TTS (0.05 and 0.1 mg / day) as well as a low dose of 6.5 cm2 CLIMARAMR TTS that delivers 0.025 mg / day prevents postmenopausal bone loss. It is expected that a CLIMARAMR TTS patch providing 10-15 μ would have effects on skeletal bone loss. EXAMPLE 3 The following example and data demonstrate the correlation between serum estradiol level and loss of mineral density in bone. The example also demonstrates the efficacy of using an ultra low dose of exogenous estrogen to reduce loss of bone mineral density. 1. - Experimental population Subjects were participants in the study of osteoporotic fractures (SOF = Study of Osteoporotic Fractures), the details of which have been described in Example 1. This study is based on a randomly sampled subgroup of 261 SOF participants who were taken blood samples and had technically adequate calcaneus BMD both in the baseline exam and a follow-up in 1993-94. Among these, 241 women had measurements of bone mineral density technically adequate both in the examination of year two in 1990 and in the follow-up in 1993-94.
The final results are based on subgroups of 261 and218 women with BMD examination pairs of hip and full calcaneus respectively, who do not report current use of estrogen replacement therapy during the baseline interview. Sample sizes for individual trials vary due to missing values.
Also, sample sizes are smaller for estradiol than for other trials because we added estradiol to the trial set some time after the study was started. 2. - Measurements At the baseline visit in 1986-1988, a detailed questionnaire was administered in which subjects were asked about current or previous use of estrogen, calcium and multivitamins containing vitaminD. Subjects were examined to obtain height and weight measurements. The mass of the calcaneus bone was measured using simple photon absorptiometry at baseline and at a follow-up visit in 1993-1994, after an average of 5.9 years. The bone density of the hip and its sub-regions was measured using dual x-ray absorptiometry at a second visit in 1988-1990, among 82% of the original group that survived. Repeated hip bone density was measured at follow-up in 1993-1994, after an average of 3.5 years. The mean coefficient of variation (CV) of 1.2% between centers for both the femoral and calcaneal neck was estimated using multiple measurements in team members who visited each site. No bone density measurements of the spine were made in a sufficient number of subjects to ensure inclusion in this study. 3. - Serum samples In the baseline test, blood was withdrawn between 8:00 A.M. and 2:00 P.M. and the sera were stored at -20 ° C. Within two weeks, the samples were shipped to the Biomedical Research Institute (Rockville, MD) and stored in liquid nitrogen at -190 ° C. In 1994, sera were thawed for select participants and tested for circulating hormone levels and other biochemical measurements. The laboratories performing the tests were blinded to the fracture status of the participants. The stability of selected hormone measurements over time was determined in 51 women by testing baseline levels and retesting after 3.5 years of storage at -190 ° C. Correlations between the measurement pairs were as follows: estrone (r = .98), total testosterone (r = 0.99), middle-range parathyroid hormone (r = 0.99), and 25 (OH)D (r = 0.88). All correlations were significant at p < 0.001. At the time the baseline measurements were made, an immunoradiometric assay for PTH was not available. The correlation betweenPTH (IRMA) and the middle region in the samples measured after 3.5 years of storage was 0.78. The average hormone levels do not differ substantially between the two samples. 4. - Biochemical assays Concentrations of serum estradiol, estrone and total testosterone were measured by radio immuno assay. The binding capacity of globulin that binds sex hormone is measured using a displacement technique. Testosterone free levels were calculated as the total testosterone product and the non-SHBG bound steroid percent as determined from an ammonium sulfate precipitation procedure. Intact parathyroid hormone, 25 (OH) D and 1.25(OH) 2D were measured by radio immunoassay. Accuracy for all trials is reported in Table 5, along with descriptive data on averages and standard deviations as well as the quartile ranges for each hormone.
Table 5 Descriptive data in bio-chemical assays* N, average (SD) and quartile cut points were calculated for the group with calcaneal bone loss. N's for the group of. Total hip bone loss is reported in Table 4. 10 t Range of values for indicated quartile.
. Statistical analysis All biochemical variables except estradiol were analyzed by quartiles and as continuous variables. Since 36% of estradiol levels were lower than the minimum detection level (5 pg / ml), estradiol values were divided into four groups by adjusting undetectable levels as the lowest category and dividing the remaining detectable levels into tertiles.
Categories of estradiol are reported as quartiles in the tables for simplicity. For continuous models, indetectable estradiol levels were assigned zero values. Linear regression analysis was used to analyze the association between hormone levels and the annual percent change in BMD, adjusting age and weight. The results for annual absolute change in BMD and including clinical site as co-variant, were similar and not presented. The associations between hormone levels and bone loss were weaker for the femoral neck and similar for the trochanteric regions compared to the total hip and therefore we present results only for total hip bone loss. The change in average percent adjusted for age and weight in BMD and 95% confidence intervals were calculated by quartiles of each hormone using the least squares procedure, and a test for trend across quartiles (adjusting for age and weight). used to estimate statistical significance. The association between continuous hormone levels and the annual percentage change in BMD is reported in normalized ß coefficients, representing the difference in annual change in BMD by increase in standard deviation in the level of each hormone. Multivariable models were constructed to determine which biometric variables were independent predictors of bone loss after controlling the levels of other hormones. Since the levels of estradiol and estrone correlate modestly in this group (r = 0.64), these variables were only analyzed in separate models: 6. - Results Characteristics of the random sub-sample used in this study are established in Table 6, and they do not differ substantially from those of the group from which the sample was taken. The average age of women in this sub-sample was 71.3 years. The average annual bone loss was approximately 1.5% of the calcaneus and 0.5% of both the femoral neck and total hip.
TABLE 6 Characteristics of cohort participants of calcaneal bone loss.
Associations of hormones with change in calcaneal BMD are reported in Table 7. In linear regression models adjusted for age and weight, only SHBG was a significant predictor of calcaneal bone loss.
TABLE 7 Hormonal associations with change Adjusted calcaneal BMD * average annual change in BMD (95% Cl).
* Adjusted for age and weight, excluding usersERT. ¥ Difference in annual percent change in BMD due to an increase in SD in hormone / factor level. t p < 0.10 § p < 0.05 ± p < 0.01 Significance tests for differences by SD are based on the H0 test; ß = 0 in the estimate of linear parameters for each hormone. Tests of significance through quartiles are based on tests for trends. Women with SHBG levels in the highest quartile(&.2.4 ug / dL) experienced twice the loss of calcaneal bone annually (average change = -2.2 percent; 95% CI = -2.9%, - 1.6%), compared to women with levels in the highest quartile. low (Cl. 1 ug / dL; average change = -1.2%; - 1.6%, -0.7). This association remains strongly significant even after adjustment for serum estradiol, estrone and testosterone levels (p <0.002). Trends in increased loss of calcaneus bone with decreased hormone levels are apparent for all sex steroids (see Table7), but none was statistically significant. There was a reverse line border statistical association (p <0.10) between serum estrone concentrations modeled as a continuous variable and loss of calcaneus bone. The data regarding associations of hormones with BMP change of total hip (see Table 8) show that lower levels of both estrone and estradiol were significantly associated with greater loss of hip bone.
TABLE 8 Hormone associations with BMD change in total hip, change in adjusted annual average percent * in BMD (95% of Cl).fifteentwenty* Adjusted for age and weight, excluding ERT users. ¥ Difference in annual percent change in BMD due to an increase in SD in hormone / factor level. t p < 0.10 § p < 0.05 t p < 0.01 Significance tests for differences by SD are based on the H0 test; ß = 0 in the estimate of linear parameters for each hormone. 10 Significance tests through quartiles are based on tests for trends.
For example, after adjusting for age and weight, women with estradiol levels 10 pg / ml or higher experienced a slight increase in bone mass during follow-up (average annual change 0.1 percent; -0.5, -0.7) while women with levels of estradiol below 5 pg / mL experienced speeds above the average hip bone loss (average annual change -0.8 percent; - 1.2, -0.3). Examination of the average annual change in hip bone density by quartiles of testosterone levels suggested a threshold: women with total testosterone levels 26 ng / dL or higher (highest quartile) experienced a change in average annual percent of only -0.1% (-0.5%, 0.3%) compared to -0.6% (-0.8%, -0.4%) among women with less than 26 ng / dL (quartiles 1 to 3 combined). This difference was statistically significant (p <0.05). An increase in hip bone loss through the SHBG quartiles was observed, although this trend was not statistically significant in regression models adjusted for age and weight (p = 0.27). In a muiti-variable model including SHBG, estradiol and testosterone levels in addition to age and weight, lower total testosterone and higher SHBG concentrations were independently associated with change in hip BMD. Data are reported in Table 9. TABLE 9 Multivariable model: sex hormones, SHBG and total hip bone loss under change in percent BMD * of hip.
All analyzes included only women who had estradiol measurements for compatibility. The first column presents results of separate regression models for each hormone adjusted for age and weight.t The second column presents results from a single regression model containing SHBG, estradiol, total testosterone, age and weight.± p < 0.05. Because there was a substantial number of women who did not have estradiol measurements, it was not clear whether the significant association between SHBG and hip bone loss in the multivariable model resulted from adjustment of other hormone levels or if a stronger association between SHBG and bone loss existed in this subgroup of women. Therefore, the adjusted association of age and weight between SHBG and hip bone loss was re-examined among women who had estradiol measurements and was found to be statistically significant. See Table 9. When expressed as a continuous variable, estradiol was not significantly associated with change in hip BMD, either before or after adjustment for testosterone and SHBG. However, the test for estradiol quartile trend remained significant after controlling levels of SHBG and testosterone, while the trend across estrone quartiles was border line significance (p <0.06).
Only levels of 25 (OH) D were significantly different during winter (January to June) compared to the rest of the years (July to December), with average values of 24.0 during the winter and 26.9 in another way (p = 0.01). there were no significant differences in levels of calcitropic hormone comparing measures that are taken from May to October compared to November to April. None of the calciotropic hormones or calcium levels were significantly associated with change in calcaneal BMD after controlling for age and weight. See Table 7. This remains true even after adjustment for seasonal (July to December vs. January to June) and clinical and after any adjustment for, or exclusion from, current users of calcium or multi-vitamins that contain vitamin D. Lower levels of 25 (0H) D were significantly associated with increased bone loss in the hip, see Table 9. Women with 25 (OH) D levels in the highest quartile (>; 32 pg / mL) experienced an average annual change in BMD of -0.1 percent (-0.5, 0.3) compared to -0.7 (-1.1, -0.4) among women with levels of 25 (OH) D in the lowest quartile ( < 21 pg / mL). This trend remained significant after adjustment for clinical, seasonal and use of calcium and multivitamins containing vitamin D. There was no consistent association between parathyroid hormone (PTH) levels and bone loss at any site. See Tables 7 and 8. Examination of the tendency toThrough the PTH quartiles reveals an unusual pattern with higher hip bone loss for PTH levels in the central range (second and third quartiles) and less bone loss among women with more extreme values (lower and higher quartiles). 7. - Conclusion The data show that lower levels of estrogen in the serum are significantly associated with increased hip bone loss in older women, even after control for age, weight and levels of SHBG and testosterone. In fact, women with estradiol levels 10 pg / mL or higher experience, on average, a small increase in bone mass during follow-up (average annual change = 0.1%, 95% CI = -0.5, 0.7), while women with levels of undetectable estradiol (<5 pg / mL) lost hip bone at a rate of almost 1% per year (-0.8%, -1.2 ', -0.3). Since bone loss may be a major risk factor for hip fracture in older women, these findings suggest that even modest increases in circulating hormone levels may be effective in reducing rates of bone loss and fractures. SHBG concentrations were strongly associated with both hip and calcaneal bone loss, independent of age, weight and sex hormone levels. SHBG binds serum estradiol and testosterone with enormous affinity, and this association may reflect an increased risk of bone loss with decreasing levels of bio-available sex hormones. However, the strength of this association even after adjusting levels of estradiol in serum and testosterone suggest that there may be other influences of SHBG on the bones. Lower levels of 25 (OH) D are related to faster bone loss of the hip, but not of the calcaneus. Previous results of studies in cross section have been mixed: two studies have found post-menopausal women with lower levels 25 (OH) D have lower bone mineral density (See Villareal, J. Clin Endocrin, Metab 72 (3) : 628-634 (1991), Martínez et al., Calcified Tissue International 55 (4): 253-256 (1994)), but at least one cross-sectional study found no significant association between 25 (OH) D and the density of minerals in bone in the distant radius, the mid-radius or the lumbar spine (seeTsai et al., Calcif ied Tissue International40 (5): 241-243 (1987)). However, the Tsai et al. Study has a broad age range (33-94) and does not report associations separately from older women. PTH levels were not associated with bone loss of the calcaneus or hip in our study. This was true even after adjusting for seasonal and exclusion of women taking calcium or vitamin D supplements. This study is a prospective study, which measures rates of change in bone mass instead of measurements of bone mass in cross section. Serum samples were obtained before measurement of change in bone mass. The sample size was considerably larger than in most other related studies and the subjects were sampled from a group of older women living in four separate communities in the U.S.A. In conclusion, the results show that endogenous estrogens and SHBG are important determinants of bone loss in older women. Levels lower than 25 (OH) D are associated with more rapid loss of bone from the hip, but not from the calcaneus. PTH and other calciotropic hormones do not significantly influence bone loss in older women.
EXAMPLE 4 The following example and data demonstrate the correlation between serum estradiol levels and bone density. 1. - Experimental population Subjects in this analysis were participants in the study of osteoporotic fractures (SOF = Study of Osteoporotic Fractures), which are described in Example 1. All 9704 SOF participants were interviewed and examined in one of the clinical centers during the baseline visit in 1986-1988. On that visit, detailed data regarding the medical conditions diagnosed was the doctor and previous medications were collected. WE obtained standardized assessments of height, weight and strength. We measured BMD of the calcaneus and the next third of the radius using simple photon absorptiometry. Lateral radiographs of the lumbar and thoracic spine were obtained. Study samples were collected from each participant. Participants completed questionnaires annually and had three biennial follow-up visits to the clinic. At the first follow-up visit (conducted during 1988-1990) we measured BMD from the proximal end of the femur using dual-energy x-ray absorptiometry (DXA). Details of these measurement methods and quality control procedures for densitometry are described in Steiger et al.
J. Bone Miner. Res. 7: 625-632 (1992). 2. - Serum samples and biochemical tests All participants were instructed to undergo fat-free diet during the night and to minimize that lipemia on the morning of the test interfere with the trials. Blood was taken between 8 A.M. and 2 P.M and the serum was frozen at -20 ° C. Within two weeks, all samples were shipped to the Biomedical Research Institute (Rockville, MD), where they were stored in liquid nitrogen (approximately -190 ° C). The long-term stability of these samples was determined by comparing estradiol results obtained after two weeks of storage at -20 ° C with those obtained after 3.5 years of storage at -190 ° C; for 51 samples, the correlation coefficient was .94 and the averages (± SD) were 11.8 (9.0) after two weeks and 10.9 (9.0) after 3.5 years. In the initial group study, biomedical tests using baseline serum were available from 400 randomly selected participants. For this trial, we excluded women who did not have measurements of estradiol in serum (n = 134) and those who reported current use of systemic estrogen therapy (n = 39); 247 women remained available for the current analysis. The following biomedical analysis was performed for each subject: calcitropic factors included calcium, 25-hydroxy vitamin D, 1.25 -dihydroxy vitamin D and parathyroid hormone, growth factors included insulin-like growth factor-1 and insulin-like growth factor-that bind protein 3; markers of bone formation include bone-specific alkaline phosphatase and osteocalcin; sex and adrenal hormones included estradiol, estrone, testosterone (total and free), SHBG, dehydroepiandrosterone sulfate (DHEAS); Other endocrine tests include styroid stimulation (TSH). Methods for these biochemical analyzes are described in Register et al., Calcif. Tissue Intl. 51: 340-343 (1992).
Immunoassays for estradiol were performed by Endocrine Sciences Laboratory (Tarzana, CA), which uses a highly sensitive assay with a lower detection limit of 5-pg / mL. To validate our findings in an independent group, we measured baseline serum estradiol from a second random sample of 222 SOF participants, which met the same inclusion and exclusion criteria as the initial analysis sample. The measurements of estradiol in serum in the second group were made in Corning Nichols Institute (San JuanCapistrano, CA), which use a highly sensitive assay with a lower sensitivity limit of 2-pg / mL.
Results of 22 specimens of divided serum measured in two reference laboratories, showed similar results (correlation coefficient a .9, slope1. 3) . 3. - Evaluation of vertebral deformation The graphic radiographic morphometry was used to diagnose vertebral deformity predominant in baseline, criteria for predominant deformity were > 3 SD reduction in any of the antero-, medial- or post-vertebral height of at least one vertebra, see Black et al., J. Bone Miner. Res. 10: 890-902 (1995). 4.- Statistical analysis Analysis was carried out in the SOF coordination center at the University of California, SanFrancisco . In the first sample (N = 247), women were stratified into four groups based on their serum estradiol level: in one stratum, all women with undetected estradiol (<5 pg / mL, n = 81, 32.8%) were included; the remaining women were stratified into three approximately equal groups based on estradiol levels of 5-6 pg / mL, 7-9 pg / mL and 10-25 pg / mL; The number of subjects in these strata was 55 (22.3%), 63 (25.5%), and 48 (19.4%). In the second sample (validation) (n =222), the same serum estradiol cut-off points were used; the proportion of subjects in these strata was 25.2%, 27.5%, 24.8%, and 22.5%. The results were examined, adjusting for weight and age of the subjects. Because the weight and BMD are highly correlated and because the weights show the strongest relationship with BMD. See Bauer et al., Arm. Intern. Med. 118: 657-665 (1993).
We select weight for inclusion in subsequent models. We then evaluate the possible contribution of co-variants by first examining variables that we consider to be potential predictors of bone density or osteoporotic fracture. If any potential co-founder showed a consistent trend (analysis of variance with test for linear trend) through the estradiol strata, we further evaluated their effects in multivariate models for which the results were bone density in the four skeletal sites of interest. . These models include age, clinical site, and serum estradiol as well as the candidate predictor variable. If the candidate variable is retained in most of these models, it is considered a co-founder. We also added gripping resistance and current smoker to the model, because these variables proved to be predictors of BMD in these previous studies (see Bauer et al., Ann. Intern. Med.118: 657-665 (1993)) and fracture (see Cummings andcollaborators, N. Eng. J. Med. 332: 767-773 (1995)). Using analysis of variance, the adjusted average BMD is calculated for each stratum of estradiol. The Dunnett test is used to determine the statistical significance of the difference between the adjusted BMD of the lowest stratum (unobserved estradiol) against each of the three upper strata. The logistic regression model is used to calculate adjusted multiple OBI ratios (OF) and the 95% confidence intervals for risk of vertebral deformity predominant for each of the three higher strata of estradiol against the lowest stratum. Analyzes were performed using SAS (SAS Institute, Cary,? C). 5.- Characteristics of the subject by estradiol level. Clinical measurements of women in this study are set forth in Table 10.
TABLE 10 Clinical and laboratory values measured in 247 older women grouped by serum estradiol level.
On average, the women in this study were about 72 years old (Table 10). Average estradiol varies with age, those aged 65 to 74 years had average estradiol (SD) 5.8 (5.1) pg / mL compared to 6.3 (4.9) pg / mL for those greater than or equal to 80 years. The height does not differ through the estradiol strata. Body weight was statistically significantly higher among women in the highest estradiol stratum. Biochemical tests that do not show large or consistent differences across estradiol strata include calciotropic factors, growth factors, bone-specific alkaline phosphatase, and DHEAS (demonstrated results). Estrone and testosterone in total were approximately twice as high in women with the highest estradiol levels compared to those with undetectable estradiol. Various variables, including those referring to body weight, those that affect and with consistent difference through estradiol strata. Osteocalcin tended to be lower at higher levels of estradiol, the highest stratum was 10.9% lower than the lowest stratum(undetectable) of estradiol. Because they also showed difference through estradiol strata and substantially consistent relationships to the results, 4 BMD, weight and SHBG were included in the model along with age, grip strength and whether she is a current smoker. Baseline BMD adjusted for age and weight of all four skeletal sites showed a similar tendency to be higher with higher estradiol levels. In comparison with women with less than 5 pg / mL of estradiol in serum, those with levels of 10 to 25 pg / mL had higher average BMD statistically significant in all the skeletal sites, the difference was 4.6%, 6.1%, 5.8%, and 7.1% total hip, calcaneus, proximal radius and spine (p <.05 for each comparison) After multiple adjustment, the difference remained statistically significant: 5.7%, 6.3%,6. 5%, and 6.9% for the total hip, calcaneus, proximal radius and spine. The women in the validation study showed similar characteristics to the women in our initial study including serum estradiol levels. Similar associations between serum estradiol level and BMD were found in the validation group. BMD adjusted for age and weight of all four skeletal sites showed a trend similar to the original group. In comparison with women who had < 5 pg / mL of estradiol in serum, those with levels of 10 to 25 pg / mL had 4.9%, 9.6%, 7.3%, and 6.8% higher BMD in total of hip, calcaneus, proximal radius and spine. After multiple adjustment, the difference remained statistically significant: 3.8%, 7.0%, 5.4%, 6.9% higher BMD in total hip, calcaneus, proximal radius and spine. Of women in the lowest estradiol stratum(< 5 pg / mL), 30% had greater than or equal to 1 predominant vertebral deformities in contrast to the one not prevalent in the other three strata that were in the range of 7 to 19%. After adjusting for age and weight, predominant vertebral deformities were 60% less likely among women with estradiol levels between 5 and 25 pg / mL (compared to those who had undetectable estradiol levels) (0R = 0.4, 95% range). confidence (Cl) 0.2-0.8); this proportion is affected in a minimal way by multiple adjustment (0R = 0.4, Cl0. 2-0.7). In the validation group, we found no tendency towards an association between the level of estradiol and the predominance of vertebral deformity; the predominance was similar across all strata and was in the range of 15 to 19%. 6. - Conclusions The data demonstrate that serum estradiol levels below 5 pg / mL are harmful for skeletal health in older women. After adjusting for age and weight, the data showed that women with estradiol < 5 pg / mL had substantially less BMD in all the skeletal sites. Additionally, the data show that osteocalcin, an indicator of bone turnover, tends to be higher in women with lower levels of serum estradiol. In this way, low levels of estradiol exert clinically important effects on the skeleton of an older woman. A likely explanation is that estradiol, when present in low concentrations, reduces skeletal remodeling, allows for better quality as well as bone mass and thus reduces fracture proportions. The difference in BMD that we observed was substantial and corresponds to approximately 0.4 standard deviations (SD) even after adjusting for multiple factors. This difference is expected to reduce 30% of the risk of hip fracture(considering 1.0 SD = .11 g / cm2 and 1 .0 SD difference= RR 2.8). Using similar calculations, we estimate that the benefit of spine density associated with serum estradiol 10-25 pg / mL, equivalent to approximately 0.4 SD, will be expected to reduce the risk of spine fracture by 23% (considering 1. O SD = .17 g / cm2 and 1.0 SD difference = RR 2.1). Estrone has been widely used as a predictor of skeletal health among post-menopausal women. BMD has been found to be related cross-sectionally with serum estrone levels in both white and black women. (See Cauley et al., Am. J. Epid. 139: 1035-1046 (1994)). However, this study has been limited because estradiol levels were below the detection limit in more than half of the women. Estrone was not predictive of incidental hip fractures. In post-menopausal women, estrone is quantitatively the predominant estrogen and occurs primarily from the conversion of adrenal androstenedione. Estradiol is produced through reduction of estrone and through aromatization of adrenal and ovarian testosterone and that derived from the conversion of antroestendione and DHEA (See, Cauley et al., Am. J. Epid. 139: 1035-1046 (1994)). The data show a relatively high correlation between serum levels of estrone and estradiol both in the initial sample (r = 0.65) and in the validation sample (r = 0.78). Estradiol, not estrone is considered to be the effector hormone in the nuclear receptor (See Grodin et al., J. Clin.Ende.Metab.36: 207-214 (1973)). Estradiol is also 4 to 10 times more potent than estrone. That estradiol was the main sex steroid hormone that has a strong, consistent and positive relationship with the skeletal result could therefore be expected. Estradiol can produce beneficial skeletal effects through several possible mechanisms: it reduces the activation of bone metabolic units; antagonizes the stimulation of parathyroid hormones of bone resorption; can improve the survival of osteoblasts by local cytosines or other growth factors; improves both the absorption efficiency of gastrointestinal calcium and the conservation of renal calcium. Some or all of these actions can respond to very low levels of estradiol. Our data also support the effect of estradiol to reduce bone turnover. Indirectly, estradiol can influence bones through body weight, but our data adjusted for body weight and still found a strong association. In the past, the study of the effects of estradiol has been hampered by insensitive tests. Only very sensitive and precise methods can distinguish between low (<30 pg / mL) and very low (<5 pg / mL) estradiol levels. Few commercial laboratories provide estradiol assays that have lower detection limits of less than 20 pg / mL. Lack of sensitive trials may have prevented others from finding these associations. The impact of endogenous low to very low estradiol production is subtle and can take many years to manifest differences in BMD. Our cross-sectional data suggest that estradiol levels do not decrease with aging after 65 years of age. Our results indicate a protective effect of low levels of estradiol against low bone mass and fracture, this effect is consistent because it can be observed for bone density in several skeletal sites for loss of bone in the hip and calcaneus and for risk of spine fracture . These findings were reproduced using a highly sensitive estradiol assay in a different laboratory.
The beneficial skeletal effects of estradiol in the elderly occurred at levels that previously had been considered without physiological impact. The above description of the preferred embodiments and examples is illustrative of the present invention and should not be considered as limiting thereof. The invention is defined by the following claims, with equivalents of the included claims.