LEAD IN BONE: INTERNAL SOURCE AND TARGET ORGAN?

Denis Nash, MPH, PhD and Ellen Silbergeld, PhD*, New York City Health Department and *Department of Epidemiology and Preventive Medicine, Program in Human Health and the Environment, University of Maryland Medical School, Baltimore MD 21201

Email of corresponding author: esilber@aol.com

ABSTRACT

The toxicokinetic behavior of lead results in accumulation in bone over the life span with half-lives (depending upon bone type) 2-3 orders of magnitude longer than its half-life in blood. Lead in bone represents both a long-term dosimeter of past exposure and a potential source of internal exposure during periods of bone demineralization. We have studied the effects of menopause on lead redistribution since this period is a natural life stage during which cessation of gonadal hormone production alters bone physiology. Using data from both NHANES II and NHANES III we have found that blood lead levels in postmenopausal women are 30-35% higher than in premenopausal women. Using the NHANES III dataset, we report this pre-postmenopausal differential is not observed in postmenopausal women currently using hormone replacement therapy. Bone may also be a target for lead toxicity, as evidenced by an inverse relationship between blood lead levels and bone density measures in postmenopausal women. Overall these results suggest that bone may serve as a source of exposure later in life and that lead may have damaging effects upon bone integrity.

INTRODUCTION

Lead is a bone-seeking element, and over a lifespan of exposure, lead levels in bone increase. For that reason, measurement of lead in bone, directly by biopsy or in vivo by x-ray fluorescence (XRF), has been utilized to evaluate cumulative exposure and to test associations with long term outcomes in adults (Hu et al 1998). Since bone is a dynamic physiological compartment, lead may be mobilized from the skeleton during periods of increased bone mineral turnover (Gulson et al 1997). Several cross sectional studies have identified menopausal status as a significant predictor of blood lead levels in women (Webber et al 1995; Symanski and Hertz 1995; Silbergeld et al 1988; Staessen et al 1996). A prospective study observed that perimenopausal women had small but significant increases in blood lead over time, which were correlated with decreases in bone mineral density (Muldoon et al 1997)

Lead is incorporated into the bone mineral matrix, within the hydroxyapatite lattice (Hass et al 1967). Relativity little attention has been given to the possibility that lead might affect bone structure and function, despite the fact that its atomic weight is approximately 4 times the size of calcium. Studies of bone cells in vitro have demonstrated that lead affects response to calcium modulatory signals (Pounds et al 1991; Angle et al 1990), as well as production of bone related proteins in ROS cells (Sauk et al 1992).

In this paper, we report further information on determinants of blood lead levels in post menopausal women, and preliminary data on possible adverse effects of lead on bone in older women.

METHODS

For these analyses, we utilized data from the National Health and Nutrition Examination Survey III, conducted from 1988 to 1994 in the US. The NHANES III sample was designed to represent the US civilian, noninstitutionalized population, with some oversampling to include populations of special interest, such as Mexican-Americans and the elderly (Pirkle et al 1994). We have utilized the subset of 2574 women, aged 40-59 years, who took part in the NHANES III survey interview. After exclusions, for lack of data on critical variables (blood lead, menstrual status, unreliable bone density measurement, race/ethnicity groups too small for analysis), the analyses were run on a total of 1914 women, of which 11.2% were non-Hispanic black, 84.6% non-Hispanic white, and 4.2% were Mexican-American. Information was utilized on alcohol use, body mass index, parity, breastfeeding, education, income, smoking history, and residence. Blood lead levels were measured in venipuncture samples by CDC-NCEH, with a method limit of detection of 1 :g/dL. Menopausal status was categorized as premenopausal (intact ovarian function), surgically menopausal (oophorectomy prior to natural cessation of menses), and natural menopause. History of fracture and use of hormone replacement therapy (current, past, never) were self-reported. Bone mineral density was measured by dual energy x-ray absorptiometry (DEXA); measurements from the femoral neck were used in this analysis.

RESULTS

1. Menopausal status and blood lead levels

Menopausal status was highly correlated with blood lead levels among women

Figure 1. Blood lead levels in women, aged 40-59 years, by menstrual status and HRT use

aged 40 to 59, as shown in fig. 1. Geometric mean blood lead levels in naturally

menopausal women who reported never or not currently using HRT were significantly elevated as compared to premenopausal women; this relationship was seen after adjustment for other variables known to affect blood lead status (smoking, education, income level, ethnicity, alcohol use) (Brody et al 1994). Among postmenopausal women, current users of hormone replacement therapy had blood lead levels no different from premenopausal women. The relationship between blood lead level and femoral bone mineral density (BMD) was less

clear. When the population was stratified by BMD quartile, there was no consistent relationship between blood lead and BMD; however, women with BMD below the median (0.773 g/cm²) had higher blood lead levels than women with higher BMD.

2. Blood lead levels and bone

Within the same cohort, we evaluated the relationship between blood lead levels and bone status, by calculating the odds of osteoporosis (self reported as well as by DEXA) and reported history of fracture. For these analyses, we stratified the cohort by quartile of blood lead measurement: the quartiles had means of 0.9, 1.9, 3.0, and 4.8 :g/dL respectively. As shown in Table 1, with increasing levels of blood lead there was an increase in the percentage of women reporting hip fractures. More strikingly, the percentage of women with DEXA measurements in the osteoporotic or osteopenic category increased as blood lead levels increased over the quartiles of the population.

TABLE 1. BLOOD LEAD LEVELS AND BONE CHARACTERISTICS IN POST MENOPAUSAL WOMEN

	Q1		Q2	Q3	Q4
Mean blood lead (ug/dl)	0.9	1.9		3	4.8
Self-report (%)
Osteoporosis diagnosis	5.3		4.9	5.5	5.5
Osteoporosis treatment	3.2		2.3	2.6	2.7
Hip fracture diagnosis	0.7		1.6	2.2	1.5
DEXA measurement (%)
Osteoporotic	33.9		39.0	39.7	41.9
Osteopenic	14.9		20.6	31.5	32.9

DISCUSSION

These data confirm earlier studies reporting a relationship between menopausal status and blood lead levels among women (Silbergeld et al 1988). Because the data are cross sectional, it is not possible to infer that changes in physiological variables drive changes in blood lead. Nevertheless, the finding of associations between blood lead and HRT use strongly support the assumption that higher levels of blood lead in postmenopausal women are associated with bone-related variables. In contrast to earlier studies using the NHANES II dataset (Silbergeld et al 1988), we found no differences between white and black women, nor did we observe an effect of parity. This may be due to the general reductions in environmental lead exposures experienced nationally in the US since 1976 due to the removal of lead from gasoline, or to other changes.

The higher levels of blood lead among postmenopausal women, while still relatively low, may nevertheless confer some increased risks to health. In other analyses, we have reported associations between blood lead levels and hypertension (Nash et al 1999). In this analysis, we find a relationship between blood lead levels and risks of significantly decreased BMD, as evidenced by associations between elevated blood lead levels and increased percentages of women with osteopenia or osteoporosis. The possibility that lead in bone may have adverse effects on bone physiology deserves further study in a prospective design.

REFERENCES

Angle CR, Thomas DJ, Swanson SA (1990). Toxicol Appl Pharmacol 110: 281-287.

Brody D, Pirkle, JL, Kramer RA, et al (1994). JAMA 272: 277-283.

Gulson RL, Jameson CW, Mahaffey KR, et al (1997). J Lab Clin Med 130: 51-62.

Hass GM, Landerholm W, Hemmens A (1967) Am J Pathol 50: 815-845.

Hu H, Rabinowitz M, Smith D (1998). Environ Health Perspect 106: 1-8.

Muldoon SM, Cauley JA, Allen L (1997). Am J Epidemiol 144: Abstr 83.

Nash D, Silbergeld E, Magder L, et al (1998). Am J Epid 147: S93.

Nash D, Silbergeld E, Sherwin R, et al (1999). Am J Epid 149: S35.

Pounds JG, Long GJ, and Rosen JF (1991). Environ Health 91: 17-32.

Sauk JJ, Smith T, Silbergeld EK et al (1990). Toxicol Appl Pharmacol 110: 240-247.

Silbergeld EK, Schwartz J, Mahaffey KR (1988). Environ Res 47: 79-94.

Staessen JA, Roels H, Fagard R (1996). JAMA 275: 1563-1570.

Symanski E., Hertz PI (1995). Am J Epidemiol 141: 1047-1058.

Webber CE, Chettle DR, Bowins RJ et al. (1995) Environ Health Perspect 103: 1150-1153.