Mandibular and palatal tori, bone mineral density, and salivary cortisol in community-dwelling elderly men and women.

BACKGROUND: This investigation evaluated the relationship between the presence of tori and bone mineral density (BMD) and salivary cortisol levels. METHODS: A total of 230 healthy, community-dwelling elderly men (n = 129) and women (n = 101) aged 60 and older participated in this study. Forty-three women were on hormone replacement therapy (HRT). This was a component of a 5-year longitudinal study measuring subjects' body composition, hormone levels, physical activity, and diet every 6 months. Subjects were examined for the presence of tori by visual inspection and digital palpation. BMD at six sites was measured by dual-energy X-ray absorptiometry. Salivary cortisol levels were measured by radioimmunoassay. RESULTS: Twenty-three percent of all subjects had mandibular tori, 13% had palatal tori, and 12% had both mandibular and palatal tori. Mandibular tori were more common in men, and palatal tori were more common in women. The presence of mandibular tori was significantly correlated with BMD of the lumbar spine, femoral neck, trochanter, and Ward's triangle for all subjects, and with the femoral neck and trochanter of women not on HRT. Men with palatal tori had lower levels of salivary cortisol in the evening. CONCLUSIONS: This study documented the high prevalence of mandibular and palatal tori in a group of 230 elderly, community-dwelling subjects. Women not on HRT and all subjects taken as a group with mandibular tori had higher BMD. The presence of tori at young adulthood may be a marker of higher BMD in the future and of a lower risk for developing osteoporosis.

Elevated salivary cortisol in the evening in healthy elderly men and women: correlation with bone mineral density.

BACKGROUND: Aging is associated with a loss of bone mineral density (BMD) in men and women. Loss of BMD can also be caused by hypercortisolemia in men or women at any age. This study measured salivary cortisol at 2300 h and 0700 h as indices of cortisol secretory activity in 228 elderly, community-dwelling subjects. Salivary cortisol results were correlated with BMD. We hypothesized that salivary cortisol is elevated at 2300 h in elderly people, and that salivary cortisol will correlate negatively with BMD. METHODS: Saliva was sampled at 2300 h (nadir in circadian rhythm) and 0700 h (peak in circadian rhythm) in 130 men (70.7 +/- 0.4 years old) and 98 women (70.0 +/- 0.4 years old); approximately half of the women were receiving hormone replacement therapy (HRT). BMD was measured by dual energy x-ray absorptiometry. RESULTS: Salivary cortisol at 2300 h was significantly elevated in men (2.3 +/- 0.1 nmol/L) and women (2.1 +/- 0.1 nmol/L) as compared to 73 younger controls (1.2 +/- 0.1 nmol/L; 37 +/- 1 year old). Salivary cortisol at 0700 h was not different between older subjects and younger controls. There was a significant negative correlation of lumbar (L2-4) BMD and 2300 h salivary cortisol in older women (r = -0.20, p = .05; n = 98); this correlation was significant only in women not on HRT. There was a highly significant negative correlation of lumbar (L2-4) BMD and 0700 h salivary cortisol in older men (r = -0.31, p = .0003). CONCLUSIONS: Salivary cortisol is a simple, nonstressful method for assessing activity of the hypothalamic-pituitary-adrenal (HPA) axis in the elderly population. A major finding was an elevation in the late night nadir in cortisol secretion. We also suggest that elevated cortisol secretion in elderly people may contribute to the age-related loss in bone mineral density and that this effect is prevented by HRT.

Association of dehydroepiandrosterone sulfate, body composition, and physical fitness in independent community-dwelling older men and women.

OBJECTIVES: To determine the association of dehydroepiandrosterone sulfate (DHEAS), body composition, and physical fitness in independent community-dwelling men and women aged 60 to 80 years. DESIGN: Cross sectional analysis. PARTICIPANTS: Independent men and women, 60 years of age and older, living in urban and suburban communities of Southeastern Wisconsin. MEASUREMENTS: History, physical examination, physical activity level, and anthropometrics were measured for every subject. Total adipose mass (TAM) and lean body mass were measured using dual energy X-ray absorptiometry. Dehydroepiandrosterone sulfate, insulin-like growth factor-1 (IGF-1), total testosterone (TT), and free testosterone (FT) were measured using radioimmunoassay. Physical fitness was measured as VO2max using exercise stress tests. Blood for lipids was analyzed using standard assays. RESULTS: In men, the DHEAS was significantly correlated to age (r = -.32), TAM (r = -.27), percent fat (r = -.30), HDL cholesterol (r = .34), TT (r = .30), VO2max (r = .23), and percent lean body mass (% LBM) (r = .33). In women, the DHEAS was not significantly correlated to any of the variables examined except body mass index (BMI) (r = .23). In men, after partialling out age, DHEAS was significantly correlated to HDL, % fat, TAM, % LBM, and TT. Multivariate analysis for men revealed that high density lipoprotein cholesterol (HDL) was the strongest predictor of serum DHEAS level, followed by % LBM, BMI, and age. The men in the highest quartile of serum DHEAS levels were different from those in the lowest quartile in terms of age, TT, FT, % fat, TAM, % LBM, HDL, and low density lipoprotein (LDL) cholesterol level. No such differences were found in the two groups of women. CONCLUSION: In this group of independent community-dwelling older men, several factors were found to be associated with the serum DHEAS concentration, whereas in a group of older women, no such associations were identified with the exception of BMI. Men in the highest quartile of serum DHEAS level, compared with those with a serum DHEAS level in the lowest quartile, were younger, leaner, more fit, had higher TT and FT levels, and had a favorable lipid profile. No such differences were identified between the women in the highest and the lowest quartiles of serum DHEAS level.

Predictors of lean body mass and total adipose mass in community-dwelling elderly men and women.

As part of an ongoing longitudinal study, we analyzed cross-sectional data to identify the predictors of lean body mass (LBM) and total adipose mass (TAM) in community-dwelling elderly men and women. Body composition analysis was done using dual energy x-ray absorptiometry. A total 262 subjects (118 women and 144 men), 60 to 80 years of age, from the urban and suburban communities of southeastern Wisconsin were studied. In women, the age (r = -.18), body mass index (BMI) (r = .43), and waist-to-hip ratio (WHR) (r = .30), and in men, BMI (r = .45) and insulin-like growth factor-1 (IGF-1) (r = .32) were identified as predictors (P < .05) of LBM. In women, the BMI (r = .87), WHR (r = .21), and functional work capacity (VO2 max) (r = -.47), and in men, the BMI (r = .83), WHR (r = .52), dehydroepiandrosterone sulfate (DHEAS) (r = -.27), total testosterone (TT) (r = -.35), free testosterone (FT) (r = -.23), physical activity (LTE) (r = -.32), and VO2 peak (r = -.59) were identified as predictors of TAM. After partialling out age in addition to the predictors identified earlier, the VO2 peak was identified as a predictor (P < .05) of LBM in both women and men, and TT, FT, and LTE as predictors (P < .05) of LBM in men. We conclude that the BMI, WHR, and VO2 peak influences LBM and TAM in both women and men. Additionally, in men LBM and TAM is influenced by hormone profile.

Human growth hormone effect on serum IGF-I and muscle function in poliomyelitis survivors.

Previous work showed low insulin-like growth factor I (IGF-I) in polio survivors compared with age-matched controls and it was hypothesized that the low IGF-I was caused by the lack of growth hormone (GH) secretion. The present study asked: Is the nocturnal release of GH subnormal in polio survivors? Can the low IGF-I level be raised to the range of healthy young men (240 to 460 ng/mL) by human growth hormone (hGH) treatment? If so, what dose of hGH is required? Does the hormone treatment affect muscle function? Eleven polio survivors with evidence of postpoliomyelitis syndrome, aged 50 to 65 years, and low IGF-I levels (average IGF-I value of 170 ng/mL) were studied. The serum level of GH was measured in the first 4 hours of sleep. The serum IGF-I level was determined before and during hGH treatment at 0.0075, 0.015 or 0.03 mg/kg of ideal body weight (IBW), three times a week for successive periods of 1 month. Before and after hGH treatment, strength was determined in knee extensor and flexor muscles and the elbow flexor and elbow extensor muscles. Nocturnal GH was low in the polio survivors compared with healthy young men. Serum IGF-I was raised into the target range by either 0.0075 or 0.015 mg hGH/kg three times a week. After 3 months of hGH treatment, no consistent changes in muscle strength were observed in the study group.(ABSTRACT TRUNCATED AT 250 WORDS)

Hyposomatomedinemia and hypogonadism in hemiplegic men who live in nursing homes.

The purpose of this study was to determine the prevalence of low serum insulin-like growth factor-I (IGF-I) and testosterone in men with poststroke hemiplegia. Serum concentrations of IGF-I, total testosterone, and free testosterone were compared in healthy young men, healthy old men, and old men with poststroke hemiplegia. A low IGF-I level, below the lower 2.5 percentile of the healthy young men, occurred in 85% of the healthy old men, and in 88% of the poststroke hemiplegic patients. When a low IGF-I was defined as a value below the lower 2.5 percentile of the healthy old men, the prevalence in the hemiplegic men was 5%. For total testosterone, a value below the lower 2.5 percentile in the healthy young men occurred in 78% of the healthy old men and in 79% of the stroke survivors. Low total testosterone, defined as a value below the lower 2.5 percentile of the healthy old men, occurred in 17% of the hemiplegic men. The results with free testosterone were similar. Compared with healthy young men, most healthy old men have low serum IGF-I and testosterone levels. Old hemiplegic men resemble healthy old men in their IGF-I levels, but they have more cases of severe hypogonadism (total tostosterone < 193ng/dL). Because correction of IGF-I and testosterone deficiencies in younger adults improves muscle strength, work capacity, and quality of life, treatment with human growth hormone and testosterone may be a useful adjunct to physical measures in the rehabilitation of selected hemiplegic stroke survivors.

Comparison of clinical indicators in two nursing homes.

OBJECTIVE: Pressure ulcer prevalences in 30 VA nursing homes in 1986 ranged from 0% to 15%. The institutions with lowest ("A") and highest ("B") prevalence were selected for further examination. DESIGN: Analysis of nursing home files for five study periods, each lasting 6 months. SETTING: A and B were 60-bed rural and 280-bed urban facilities, respectively. MEASUREMENTS: Eleven outcome indicators were calculated for each study period: prevalences and incidences of pressure ulcer, aggressive behavior and disruptive behavior, 6-month declines in each of the four activities of daily living (ADLs), and prevalence of underweight. RESULTS: Populations in A and B were similar with regard to age, sex, length of stay, degree of dependency, and level of nursing care. All indicators for the first study period were more favorable in A than in B. In addition, underweight (body mass index < 22 kg/M2) was significantly less prevalent in A than in B. The differences between the two institutions in the indicators were persistent over the five study periods from 1988 to 1991. CONCLUSIONS: The populations of A and B were similar in the available measures of severity of illness. Nevertheless, the residents in nursing home A were significantly less likely to experience adverse outcomes than were the residents in nursing home B. The virtual absence of pressure ulcers, physical aggression, and verbal disruption in nursing home A, despite the presence of many immobile and demented residents, suggested that these complications can mostly be prevented.

Carpal tunnel syndrome and gynaecomastia during growth hormone treatment of elderly men with low circulating IGF-I concentrations.

OBJECTIVE: We studied the relationship between plasma level of insulin-like growth hormone I (IGF-I), changes in lean body mass and in adipose mass, and adverse side-effects during human growth hormone (hGH) treatment of elderly men who had low IGF-I levels. DESIGN: The first six months was a period of baseline observation. The subjects were then randomized into two groups so that during months 7-18, men in group I received hGH, and men in group II served as untreated controls. SUBJECTS: Eighty-three overtly healthy elderly men, who were selected because their plasma IGF-I level was less than 0.35 units/ml. The men were randomly assigned in a ratio of three to one into group I (n = 62) or into group II (n = 21). MEASUREMENTS: Plasma IGF-I level was measured monthly. Lean body mass and adipose mass were measured every six months. RESULTS: Fifteen men left the study during the baseline period because of personal reasons or intercurrent medical events. In those who received drug (group I), there were a number of adverse reactions which could have been related to the hGH therapy: carpal tunnel syndrome 10, gynaecomastia 4, and hyperglycaemia 3. In total there were 27 dropouts from group I and two dropouts from group II after the six-month point, for a variety of medical and non-medical reasons, the majority probably not related to hGH therapy. During the hGH treatment of group I, plasma IGF-I increased from the range 0.10-0.35 units/ml into the range 0.5-2.2 units/ml. Among the 18 men who completed 12 months of hGH treatment without experiencing one of the three above-noted presumed hGH side-effects, mean and peak plasma IGF-I during treatment were significantly lower than among the 13 men who experienced carpal tunnel syndrome or gynaecomastia (one subject had both) while on hGH. With one exception, neither carpal tunnel syndrome nor gynaecomastia occurred in any individual with a mean IGF-I level less than 1.0 units/ml during hGH treatment. Twelve months of hGH treatment (group I) caused an increase in lean body mass to 106% of the initial baseline (month one of the protocol), and a reduction in adipose mass to 84% of the baseline. Meanwhile, the lean body mass of the untreated men in group II declined to 97% of the initial baseline. The body composition responses after 12 months of treatment in group I were larger in the men whose mean intra-treatment IGF-I level was 0.5-1.0 units/ml, than in the men whose mean intra-treatment IGF-I level was 1.0-1.5 units/ml. CONCLUSIONS: These observations show that when elderly men with low circulating IGF-I concentrations are treated continuously with hGH, elevation of plasma IGF-I above 1.0 units/ml is associated with a substantial frequency of carpal tunnel syndrome or gynaecomastia. It may be that the effects of the hormone in expanding lean body mass and reducing adipose mass can be achieved, and the side-effects avoided, by maintaining the mean IGF-I level in the range 0.5-1.0 units/ml.

Prevalence of low plasma IGF-I in poliomyelitis survivors.

OBJECTIVE: To compare plasma levels of insulin-like growth factor-I (IGF-I, also termed somatomedin C) in polio survivors and healthy control subjects and to determine their relation to selected clinical characteristics. DESIGN: Cross sectional study. SETTING: Polio survivors living in the community recruited from the Wisconsin Polio Support Group. PARTICIPANTS: A total of 124 polio survivors (49 males and 75 females), ages 35 to 77 years, and 261 healthy control subjects (139 males and 122 females) of similar age. MEASUREMENTS: Plasma IGF-I levels were compared in polio survivors and age-matched control subjects. In the polio survivor group, the relation of IGF-I to selected clinical characteristics was examined before and after adjusting for covariates. RESULTS: Statistical analyses showed that the IGF-I concentrations were significantly lower in the polio survivors than in the controls. This difference was reflected in the means and standard errors of the two groups (0.45 +/- 0.02 vs 0.60 +/- 0.02 units/mL, P < 0.01). Plasma IGF-I below 0.35 units/mL in adults indicates little or no growth hormone secretion. In polio survivors, 38% of the plasma IGF-I values were < 0.35 units/mL compared with 19% in the healthy group. Univariate analysis showed that IGF-I in the polio survivors was significantly correlated with age, gender, and body mass index, and with dependency, pain, and difficulty in the activities of daily living (ADLs). The correlations with ADL dysfunction were independent of the correlations with age, gender, and body mass index. IGF-I level did not correlate with the subjective report of recent decline in functional status. CONCLUSION: Lower levels of IGF-I are seen in polio survivors, and this finding correlates with ADL dysfunction. The hyposomatomedinemic tendency of polio survivors may have an adverse effect on their neuromuscular function and quality of life.

Lipid and lipoprotein abnormalities in young quadriplegic men.

To assess the effect of severe inactivity on the serum lipid and lipoprotein profile, 21 quadriplegic men between the ages of 24 and 47 were compared with 20 age-matched healthy control men. The group of quadriplegic men had significantly lower levels of total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), HDL2-C and HDL3-C. The current recommendation for desirable TC is less than 200 mg/dl, whereas HDL-C of less than 35 mg/dl is considered a risk factor for coronary heart disease. Of the 50% (10/20) of the men in the normal control group who had a desirable TC, only 10% (1/10) had a low or undesirable HDL-C value. In comparison, although 81% (17/21) of the group of quadriplegic men had a desirable TC, 53% (9/17) of these individuals had a low HDL-C level. It is concluded that although the presence of lower TC could be beneficial in QM, the decreased values of HDL-C and HDL2-C and the increased ratio of TC/HDL-C suggest a higher risk of coronary heart disease. The findings are consistent with recent reports of an increased prevalence of coronary heart disease in spinal cord injury patients, which could be due to an abnormal lipoprotein profile related to diet, inactivity, changes in body composition, and life style. Moreover, the present data suggest that HDL-C should be measured in quadriplegic men with modifiable risk factors, even if they have desirable TC, to avoid missing an increased coronary heart disease risk status.

Hyposomatomedinemia in men with post-poliomyelitis syndrome.

The age of onset of the post-poliomyelitis syndrome (PPS) coincides with the tendency for declining activity of the growth hormone/somatomedin C (GH/SmC) axis. The normal plasma SmC range in men before the age of 40 is 0.50 to 1.50 units/mL. After age 40 about 30% of men have a plasma SmC level below 0.35 units/mL, signifying no detectable spontaneous GH secretory pulses. Because the GH/SmC axis stimulates DNA, RNA, and protein synthesis in muscle cells and increases their size and number, a deficiency of the GH/SmC axis could theoretically contribute as a secondary factor to the occurrence or severity of the PPS. Accordingly, the authors measured the plasma SmC level in 10 men with PPS, ages 35 to 63, and in 94 healthy men of similar age. In the PPS men, 100% of the values were less than or equal to 0.40 units/mL, and 90% were less than or equal to 0.35 units/mL. The corresponding proportions in the healthy men were 40% and 27%. Analysis of variance including age as a factor showed SmC to be significantly lower in the PPS men than in the healthy men. In an additional comparison, totally immobile nursing home men did not have lowered SmC values. In fact their SmC values were slightly higher than those of healthy men of similar age. The data revealed a new biochemical feature of PPS, hyposomatomedinemia, which might play a contributory role in the pathogenesis of the syndrome.

Effects of human growth hormone on body composition in elderly men.

Body composition changes progressively in mid and late adulthood. Lean body mass in men over 50 years old contracts at an average rate of -0.6% per year. Body weight tends to remain stable because of a reciprocal expansion of adipose mass. The shrinkage of the lean body mass reflects the atrophy of skeletal muscles, skin and visceral organs. Because growth hormone causes expansion of the lean body mass and contraction of the adipose mass, and because growth hormone secretion tends to diminish in late adulthood, it has been postulated that geriatric hyposomatotropism is a contributory cause to the body composition changes described above. The authors have tested this hypothesis by recruiting 45 independent men over 61 years old with plasma somatomedin C level below 0.35 U/ml, indicating little or no detectable growth hormone secretion. The 21-month protocol was as follows: baseline period 0-6 months, experimental period 6-18 months and post-experimental period 18-21 months. During the experimental period, 26 men (group I) received approximately 0.03 mg/kg of biosynthetic human growth hormone (hGH) subcutaneously 3 times a week, while 19 men (group II) received no treatment. Plasma somatomedin C was measured monthly. The following outcome variables were measured at 0, 6, 12 and 18 months: lean body mass, adipose mass, skin thickness (dermis plus epidermis), sizes of the liver, spleen and kidneys, the cross sectional areas of ten muscle groups, and bone density at 9 skeletal sites. Lean body mass and adipose mass were also measured at 21 months. In group I, hGH treatment raised the plasma somatomedin C level and maintained it in the range 0.5-1.5 U/ml.(ABSTRACT TRUNCATED AT 250 WORDS)

Effects of human growth hormone in men over 60 years old.

BACKGROUND: The declining activity of the growth hormone--insulin-like growth factor I (IGF-I) axis with advancing age may contribute to the decrease in lean body mass and the increase in mass of adipose tissue that occur with aging. METHODS: To test this hypothesis, we studied 21 healthy men from 61 to 81 years old who had plasma IGF-I concentrations of less than 350 U per liter during a six-month base-line period and a six-month treatment period that followed. During the treatment period, 12 men (group 1) received approximately 0.03 mg of biosynthetic human growth hormone per kilogram of body weight subcutaneously three times a week, and 9 men (group 2) received no treatment. Plasma IGF-I levels were measured monthly. At the end of each period we measured lean body mass, the mass of adipose tissue, skin thickness (epidermis plus dermis), and bone density at nine skeletal sites. RESULTS: In group 1, the mean plasma IGF-I level rose into the youthful range of 500 to 1500 U per liter during treatment, whereas in group 2 it remained below 350 U per liter. The administration of human growth hormone for six months in group 1 was accompanied by an 8.8 percent increase in lean body mass, a 14.4 percent decrease in adipose-tissue mass, and a 1.6 percent increase in average lumbar vertebral bone density (P less than 0.05 in each instance). Skin thickness increased 7.1 percent (P = 0.07). There was no significant change in the bone density of the radius or proximal femur. In group 2 there was no significant change in lean body mass, the mass of adipose tissue, skin thickness, or bone density during treatment. CONCLUSIONS: Diminished secretion of growth hormone is responsible in part for the decrease of lean body mass, the expansion of adipose-tissue mass, and the thinning of the skin that occur in old age.

Easily measurable adverse outcome indicators in a Veterans Affairs nursing home.

Indicator data were collected from October 1986 through April 1987 for 356 residents of the nursing home at the North Chicago, Illinois, Veterans Administration Medical Center. Measures of prevalence, incidence, and rates of change were studied for 17 adverse outcome indicators of four main types: death, undernutrition, skin breakdown, and loss of activities of daily living (eating, mobility, transfer, and toileting). Indicator values can be calculated from data routinely collected for administrative and other clinical purposes and can be used to help nursing home administrators monitor trends in the physical status of residents and to establish and track compliance with quality assurance goals.

Effects of three liquid diets on nutrition-sensitive plasma proteins of tube-fed elderly men.

The effects on three nutrition-sensitive plasma proteins of isocaloric feedings with three enteral formulas were compared in 10 tube-fed male nursing home residents. The enteral products were Isocal (based on whole protein), Peptamen (based on a mixture of oligopeptides), and Vivonex T.E.N. (based on free amino acids). The nutrition-sensitive plasma proteins were albumin, transferrin, and retinol-binding protein. After observation during four weeks of feeding with Isocal, each subject was then monitored during four weeks of Peptamen and four weeks of Vivonex T.E.N. The latter two products were alternated in a crossover design. The shift of Isocal to Peptamen did not significantly (P greater than .05) influence the serum level of albumin, transferrin, or retinol-binding protein. In contrast, the shift of Isocal to Vivonex T.E.N. or of Peptamen to Vivonex caused a significant (P less than .05) decline in all three plasma proteins, the kinetics of their reductions corresponding to their known half-lives. The behavior of the three nutrition-sensitive plasma proteins suggests that in elderly nursing home men without gastrointestinal disease the nutritional value of the protein component of the three formulas follows the order Isocal = Peptamen greater than Vivonex T.E.N. However, this conclusion will require confirmation by nitrogen balance studies.

Osteopenia in the men of a Veterans Administration nursing home.

We measured by photon absorptiometry the bone density at six sites in 65 nursing home men aged 57-85 y and in 25 independent community men aged 57-80 y. Average bone density in the community men ranged from 97% to 105% of age-matched normal men. In the nursing home men these values ranged from 71% to 92% of age-matched normal men (p less than 0.05 for comparison with the community men). About 50% of the nursing home men but none of the community men had a value less than 70% of age-matched normal men at one or more sites. Among the institutionalized men bone densities at all six sites (in g/cm2) were significantly (p less than 0.05) and directly correlated with body weight but were not significantly correlated with height, age, principal or secondary diagnoses, continuing medications, or functional level.

Fractures in the men of a Veterans Administration Nursing Home: relation to 1,25-dihydroxyvitamin D.

One hundred fifty-three men, age 48-96, 86% white, had resided in this Nursing Home for an average of 6.3 years (range 1.3-36) as of August 1984. At that time, we reviewed their medical charts to record the numbers and sites of fractures which had been diagnosed during the preceding 1 to 5 years of Nursing Home residence, the duration of this period depending on the duration of institutionalization. In addition, a clinical database was compiled comprising 70 attributes, including diagnoses, drugs, plasma (serum) chemistries, and measures of hematologic, nutritional, and functional status. Fractures during the studied period of Nursing Home residence had occurred in 24 of 153 men; six residents had experienced two or more fractures. Fracture rates in hip, spine, and wrist were 2564, 366, and 549 per 100,000 patient years, respectively. The total fracture rate, hip fracture rate, and limb fracture rate were five to 11 times higher than in the age-matched general population of white men in the United States; in Rochester, MN; in Dundee, England; in Oxford, England; or in Finland. Univariate statistical analysis showed that the rates for hip fracture or for fracture at any site were significantly associated with 13 attributes: directly with age, plasma somatomedin C, blood urea N, serum creatinine, serum uric acid, serum 25-hydroxyvitamin D (25-OH-D), degree of functional impairment, and chronic urinary tract infection, and inversely with serum 1,25-dihydroxyvitamin D [1,25-(OH)2-D], serum albumin, hematocrit, and hemoglobin. There was not a significant correlation with the number of falls/month which occurred during the 7 months after August 1984. After the effect of age was partialed out, somatomedin C, 25-OH-D, 1,25-(OH)2-D, and the diagnosis of urinary tract infection were still significantly related to the occurrence of fractures. The fact that Nursing Home fracture cases had significantly higher blood urea nitrogen and 25-OH-D, and significantly lower 1,25-(OH)2-D, than their non-fracture counterparts suggests that impaired renal production of the latter vitamin D metabolite contributed to the excessive rate of fractures.