K. Michael Davies

Bone Remodeling Increases Substantially in the Years After Menopause and Remains Increased in Older Osteoporosis Patients

Bone remodeling rates (Ac.f) were measured in transilial biopsy specimens from 50 healthy premenopausal women before and 1 year after menopause, in 34 healthy women 13 years past menopause, and in 89 women with untreated osteoporosis. Ac.f nearly doubled 1 year after menopause, tripled 13 years after menopause, and remained elevated in women with osteoporosis.

Calcium and Weight: Clinical Studies

Data from six observational studies and three controlled trials in which calcium intake was the independent variable (and either bone mass or blood pressure the original outcome variable) have been reanalyzed to evaluate the effect of calcium intake on body weight and body fat. Analysis reveals a consistent effect of higher calcium intakes, expressed as lower body fat and/or body weight, and reduced weight gain at midlife. Similarly, studies relating nutrient intake to body composition report negative associations between calcium intake and body weight at midlife and between calcium and body fat accumulation during childhood. There is a fairly consistent effect size, with each 300 mg increment in regular calcium intake associated with ∼1 kg less body fat in children and 2.5–3.0 kg lower body weight in adults. Taken together these data suggest that increasing calcium intake by the equivalent of two dairy servings per day could reduce the risk of overweight substantially, perhaps by as much as 70 percent.

Determination of bone mineral density of the hip and spine in human pedigrees by genetic and life-style factors

In 40 human pedigrees with 563 subjects, we evaluated the contribution of genetic and life‐style factors (exercise, smoking, and alcohol consumption) and the interactions between non‐genetic factors in determining bone mineral density (BMD) of the hip and spine. In our analysis, we adjusted for age, weight, height, menopausal status in females, life‐style factors, and the significant interactions among these factors. For the spine and hip BMD, heritabilities (h2) (± SE) were, respectively, 0.68 (0.21) and 0.86 (0.28) in males and 0.64 (0.13) and 0.67 (0.14) in females. Exercise had significant beneficial effects for male spine BMD and female hip BMD. Alcohol consumption experienced in our sample had significant beneficial effects on hip BMD in both sexes. Although the main effect of smoking was not significant, there were significant interaction effects between smoking and other important factors (e.g., exercise, weight, alcohol consumption). For example, for female spine BMD, exercise had significant beneficial effects in smokers; however, its effect in non‐smokers was non‐significant. This result indicates that exercise may reduce deleterious effects of smoking (if any) on BMD, but may have minor effects in increasing BMD in non‐smokers. The various interaction effects among risk factors explicitly revealed here for the first time indicate that the detailed effects and direction of individual risk factors may depend on the presence and magnitude of other factors. Weight invariably affected BMD of the hip and spine in both sexes. Age effects were significant for hip BMD, but not for male spine BMD. Genet. Epidemiol. 19:160–177, 2000.

Vitamin D status in a rural postmenopausal female population

Background: Inadequate vitamin D nutritional status is increasingly recognized as common in North American and European populations, but the extent of the shortfall and the parameters of the distribution for populations of interest remain uncertain. Purpose: To report the distribution of values for serum 25-hydroxyvitamin D [25(OH)D] in a population of rural postmenopausal women, together with quantification of factors related to vitamin D status. Setting: Nine largely agrarian counties in eastern Nebraska (∼41° N). Participants: A population-based sample of 1,179 women 55 years of age and older recruited into a four-year trial of calcium and vitamin D supplementation. Methods: Baseline biochemical, dietary, and anthropometric measurements obtained on entry into trial. Results: Serum 25(OH)D concentration at baseline varied cyclically with season, with the solar cycle explaining 2.9% of the total variance (P < 0.001). Mean seasonally adjusted 25(OH)D concentration was 71.1 nmol/L. Serum 25(OH)D also exhibited the expected inverse curvilinear relationship with serum parathyroid hormone (PTH), with the inflection point of the curve located at approximately 80 nmol/L. Supplements containing vitamin D were regularly taken by 59% of the cohort (median dose: 200 IU/d). Nevertheless, approximately 4% of all women had values below the laboratory reference range and more than two-thirds fell below 80 nmol/L. Seasonally adjusted serum 25(OH)D concentration was positively correlated with the size of daily vitamin D supplement dose, and negatively with age, weight, and body mass index (P < 0.01 for all). In stepwise multiple linear regression models, weight, age, and supplement dose were independently correlated with seasonally adjusted serum 25(OH)D, and together explained 19% of the total variance of adjusted 25(OH)D concentration. Women taking supplements had only one-sixth the chance of having a 25(OH)D value below the reference limit of the assay, compared to women who did not use supplements. Conclusions: Approximately two-thirds of this rural population fell below 80 nmol/L, a value considered to be the lower end of the optimal range. Based on the slope of 25(OH)D on supplement dose observed in these women, it would require an additional vitamin D input of nearly 2000 IU/d to reach the goal of an RDA for vitamin D, i.e., to bring 97.5% of the cohort to levels of 80 nmol/L or higher.

Potassium intake and the calcium economy

Background: Dietary potassium intake (K) lowers urinary calcium (Ca) excretion and, in short-term studies, may improve Ca balance. Purpose: Our objective was to assess K effects on the Ca economy under steady-state conditions. Design: 8-day, inpatient metabolic studies of nitrogen, phosphorus, and Ca balance, combined with dual isotopic Ca tracer kinetics studies. Study diet matched to prestudy nutrient intakes. Subjects: 191 single women studied from 1–5 times at ∼5-year intervals, for a total of 644 inpatient studies. Median age at time of study: 50.2 yrs; 301 studies were performed postmenopausally without hormone replacement; 343 were either premenopausal or postmenopausal but on estrogen replacement therapy. Results: Dietary K was highly significantly associated with urinary Ca excretion, with a coefficient of −0.0109 mmol urine Ca/mmol diet K. However, dietary K was negatively correlated with dual-tracer Ca absorption (coefficient for Ca absorption fraction: −0.00094/mmol dietary K), and was not associated with urine Ca after adjustment for Ca absorption. Conclusion: While a high K diet (i.e., one rich in fruits, vegetables, and dairy products) has multiple health benefits and clearly lowers urine Ca, it does not seem to exert any appreciable net influence on the Ca economy, largely because the reduced calciuria is offset by reduction of intestinal absorption. We note, however, that since the high K intakes in our studies come more from milk and meat than from fruits and vegetables, we cannot exclude a possible balance effect for different food sources of K.

Differences in bone mineral density, bone mineral content, and bone areal size in fracturing and non-fracturing women, and their interrelationships at the spine and hip.

Abstract. Osteoporotic fractures are a major public health problem, particularly in women. Bone mineral density (BMD), bone mineral content (BMC), and bone size have been regarded as important determinants of osteoporotic fractures. In 1449 women over age 30 years, we studied the detailed relationship, at the spine and hip, between BMD, BMC, and bone areal size (all measured by dual-energy X-ray absorptiometry) and compared their relative magnitudes in fracturing and non-fracturing individuals. We find that, (1) BMD and BMC are significantly higher at the spine and hip in non-fracturing women. Bone areal size is significantly larger at the spine in non-fracturing women; however, the significance disappears when adjustment is made for the significant difference of height (stature) between fracturing and non-fracturing women. In contrast to the spine, bone areal size is always significantly larger in fracturing women at the hip. (2) The relationship among BMD, BMC, and bone areal size is different at the spine and hip. Specifically, at the spine, BMD increases with bone areal size linearly. At the hip, BMD has a quadratic relationship with bone areal size, so that BMD increases at lower bone areal sizes, then (after an intermediate zone of values) decreases with increasing bone areal size. However, BMD adjusted for BMC always decreases with increasing bone areal size, as expected by the definition of BMD. With no adjustment for BMC, the increase in BMD with bone areal size is due to a more rapid increase of BMC than increasing bone areal size, thus explaining the observations of association of both larger BMD and larger bone areal size with stronger bone. (3) At the spine, 86.2% of BMD variation is attributable to BMC and 12.6% to bone areal size. At the hip, 98.0% of BMD variation is due to BMC and 1.1% due to bone areal size. The current study may be important in understanding the relationship among BMD, BMC, and bone size as risk determinants of osteoporotic fractures.

Heterogeneity of Bone Mineral Density Across Skeletal Sites and Its Clinical Implications

Low trauma fractures in the elderly are highly predictable by measurement of bone mineral density (BMD). Preventive measures for low BMD, such as hormone replace therapy (HRT), have potential risks. Thus, a rational decision on HRT or other therapy critically depends on an accurate diagnosis of osteopenia/osteoporosis. We assessed the degree of diagnostic heterogeneity based on spine and hip BMD for 2313 women. We found: 1. In ~30.0% of cases, the difference between spine and hip Z- and T-scores was >1.0, and in 20.8% (Z-scores) and 15.2% (T-scores) the difference was >2.0. 2. With increasing age, the proportions of women with Z- or T-scores greater at the hip than the spine generally decreased. 3. The correlation between hip and spine and Z- and T-scores ranged from 0.50 to 0.72, and generally decreased with increasing age. 4. If screened only at the hip or spine, 17.9/27.3% with osteopenia and 1.3/2.9% with osteoporosis at either site would be diagnosed as normal. Corresponding analyses of 143 men yielded similar results. Therefore, if possible, dual X-ray absorptiometry (DXA) of both the spine and hip should be performed for an accurate assessment of osteoporosis at these two most frequently fractured sites. If only one site is chosen, measurement of the hip is preferred to measurement of the spine.

Characterization of Genetic and Lifestyle Factors for Determining Variation in Body Mass Index, Fat Mass, Percentage of Fat Mass, and Lean Mass

In this study, we simultaneously characterized genetic and lifestyle factors (exercise, smoking, and alcohol consumption) in determining variation in body mass index (BMI), fat mass, percentage of fat mass (PFM), and lean mass while adjusting for the effects of age and sex. Six hundred fifty-eight Caucasian individuals from 48 pedigrees were studied for BMI. Among these individuals, 289 from 38 pedigrees were studied for fat mass, PFM, and lean mass measured by dual X-ray absorptiometry (DXA). After adjusting for age, sex, and lifestyle factors, the heritabilities (h(2)) of BMI, fat mass, PFM, and lean mass ranged from 0.52 to 0.57 with associated standard errors ranging from 0.09 to 0.14. After accounting for significant sex and age effects, exercise had significant effects for all the phenotypes studied, and the effects of smoking and alcohol consumption were not significant. Therefore, significant proportions of variation in BMI, fat mass, PFM, and lean mass were under genetic control, and exercise had a significant effect in reducing BMI, fat mass, and PFM and in increasing lean mass. This study warrants further genetic linkage analyses to search for genes for the obesity-related phenotypes measured by DXA in our population.

Genetic dissection of human stature in a large sample of multiplex pedigrees.

Recently, we reported a whole genome scan on a sample of 630 Caucasian subjects from 53 human pedigrees. Several genomic regions were suggested to be linked to height. In an attempt to confirm the identified genomic regions, as well as to identify new genomic regions linked to height, we conducted a whole genome linkage study on an extended sample of 1,816 subjects from 79 pedigrees, which includes the 53 pedigrees containing the original 630 subjects from our previous whole genome study and an additional 128 new subjects, and 26 further pedigrees containing 1,058 subjects. Several regions achieved suggestive linkage signals, such as 9q22.32 [MLS (multipoint LOD score) = 2.74], 9q34.3 [MLS = 2.66], Xq24 [two‐point LOD score = 2.64 at the marker DXS8067], and 7p14.2 [MLS = 2.05]. The importance of the above regions is supported either by other whole genome studies or by candidate genes within these regions relevant to linear growth or pathogenesis of short stature. In addition, this study has tentatively confirmed the Xq24 regions linkage to height, as this region was also detected in the previous whole genome study. To date, our study has achieved the largest sample size in the field of genetic linkage studies of human height. Together with the findings of other studies, the current study has further delineated the genetic basis of human stature.

Linkage and association of the CA repeat polymorphism of the IL6 gene, obesity-related phenotypes, and bone mineral density (BMD) in two independent Caucasian populations

AbstractGenetic factors play an important role in osteoporosis and obesity, two serious public health problems in the world. We investigated the relationships between obesity-related phenotypes, bone mineral density (BMD) and the CA repeat polymorphism of the IL6 gene in two large independent samples using the quantitative transmission disequilibrium test (QTDT). The first sample consisted of 1,816 individuals from 79 multigenerational pedigrees. Each pedigree was identified through a proband with BMD Z-scores ≤−1.28 at the hip or spine. The second sample was a randomly ascertained set of 636 individuals from 157 nuclear families. Ten alleles containing 9–18 CA repeats were identified in our Caucasian populations. For body mass index (BMI), fat mass and percentage fat mass (PFM), highly significant (P<0.01) or significant (P<0.05) results were found for linkage in our sample of nuclear families and for association in the multigenerational pedigrees. We also observed weak evidence for linkage (P=0.069) with spine BMD and for association with hip BMD in the sample of multigenerational pedigrees. Our results suggest that genetic variation in or near the IL6 locus may be involved in the etiology of obesity and osteoporosis.