Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2014 Aug 1.
Published in final edited form as: Appl Radiat Isot. 2013 Apr 17;78:57–61. doi: 10.1016/j.apradiso.2013.04.012

Calcium Isolation from Large-Volume Human Urine Samples for 41Ca Analysis by Accelerator Mass Spectrometry

James J Miller a,e, Susanta K Hui b, George S Jackson c, Sara P Clark d,e, Jane Einstein c, Connie M Weaver c, Maryka H Bhattacharyya a,d,*
PMCID: PMC4089085  NIHMSID: NIHMS479079  PMID: 23672965

Abstract

Calcium oxalate precipitation is the first step in preparation of biological samples for 41Ca analysis by accelerator mass spectrometry. A simplified protocol for large-volume human urine samples was characterized, with statistically significant increases in ion current and decreases in interference. This large-volume assay minimizes cost and effort and maximizes time after 41Ca administration during which human samples, collected over a lifetime, provide 41Ca:Ca ratios that are significantly above background.

Keywords: 41Ca analysis, accelerator mass spectrometry, biological samples, calcium oxalate, human urine

INTRODUCTION

An extremely rare isotope of calcium, 41Ca is an effective tool to track calcium release from bone (Denk et al., 2006, 2007; Elmore et al., 1990; Freeman et al., 1997; Johnson et al., 1994). 41Ca has a very low natural abundance (~10−15) and a very long half-life (1.03×105 years), which enables it to be utilized as a quasi stable isotope analyzed by accelerator mass spectrometry (AMS) (Brown et al., 2005). Once injected into the human body, virtually all of the 41Ca that remains in the body after 3 months resides in bone (Cheong et al., 2007; Weaver et al., 2009). 41Ca: Ca ratios in urine can be used in translational research to help determine, for example, whether cadmium in mainstream cigarette smoke stimulates calcium release from bone in humans (Bhattacharyya, 2009; Ebert McNeill et al., 2012). Administration of 10 to 100 nCi 41Ca via injection or oral administration is sufficient to evaluate skeletal 41Ca release for the lifetime of the subject – a feature unique to this long-lived isotope (Denk et al., 2006, 2007; Fitzgerald et al., 2005).

A number of low-volume methods to isolate calcium for AMS analysis are reported that use 30- to 50-ml samples of urine, serum, or saliva, with carrier calcium added to improve sample handling (Fitzgerald et al., 2005; Freeman et al., 1995; Freeman et al., 1997; Lin et al., 2004). However, addition of carrier calcium significantly decreases the 41Ca:Ca ratio and therefore shortens the time after 41Ca administration that a study produces samples with ratios significantly above background. Although one large-volume method is reported (Cheong et al., 2007), it requires expensive filtration devices that need to be cleaned for re-use. The purpose of this technical note is to characterize a simplified large-volume calcium isolation protocol for AMS, compare it to the existing protocol, and demonstrate its transfer to an outside laboratory, providing the scientific basis for its implementation in clinical studies using 41Ca.

MATERIALS AND METHODS

To generate the most valid comparisons, each protocol was carried out in the laboratory that was most proficient in that particular protocol, i.e., Standard Protocol, C Weaver/Purdue University; Modified Protocol, MH Bhattacharyya/Argonne National Laboratory or SK Hui/University of Minnesota.

Standard Large-Volume Protocol

Cheong et al. (2007) report the one published protocol that isolates calcium for AMS analysis using large volumes of human urine. According to this protocol, urine samples are collected in acid-washed containers, mixed thoroughly, and ammonium hydroxide (concentrated; 14.5 M) is added (~10 ml per liter of urine). Samples are mixed well and pH is checked to ensure it is at least pH10. Saturated ammonium oxalate (~5% w/v) is added (50 ml per liter of urine) and mixed well. Samples are allowed to stand overnight to allow calcium to precipitate as calcium oxalate. The supernatant is decanted, leaving the precipitate and some urine. The precipitate is collected on the filter of a vacuum filter apparatus [Nalgene polysulfone (PSF) filter holder with 1 liter receiver and AcetatePlus filter (5.0 micron, 47mm, GE Water & Process Technologies)] and rinsed at least 3 times with 1% ammonium oxalate. The filter is transferred to a weighing boat, covered partially with another one to prevent contamination, and dried in an oven at ~37° C. The dried precipitate is transferred on the filter to a container for dissolution, in preparation for 41Ca analysis by AMS at the Purdue University PRIME Lab. Each filter apparatus is cleaned free of 41Ca using 10% Contrad 70 detergent solution and rinsed with deionized water and ultrapure water prior to re-use with another sample.

Modified Large-Volume Protocol

A modified protocol for isolating calcium from large-volume urine samples was developed at Argonne National Laboratory. Steps in the modified protocol and the rationale for each are presented in Table 1. Four steps in the precipitation differ from the above standard protocol: using heated ammonium oxalate, adding sodium acetate, adding ammonium oxalate before adjusting the pH with ammonium hydroxide, and washing the precipitate thoroughly, by centrifugation, with dilute (0.5%) ammonium oxalate. Finally, precipitation was carried out in the original urine collection bottle, transferred to a 50-ml centrifuge tube, and all subsequent steps were carried out in that tube, including washing, drying, storing, and shipping the precipitate for analysis of 41Ca:Ca ratio by AMS at the Purdue PRIME Lab.

Table 1.

Modified Large-Volume Protocol for Calcium Isolation from Human Urine for 41Ca Analysis by AMSa

Step Detailed Description
(per 100 ml urine)		 Explanation
1. Acidify urine in collection bottle Add 7.5 ml 12 N HCl to urine collection bottle, acidifying urine to ~0.8M acid. Acidified urine can be stored at room temperature, nonbiohazardous, calcium maintained in solution
2. Add hot ammonium oxalate Add 25 ml saturated ammonium oxalate (~5%) heated to 90°C Precipitation from hot solution produces purer precipitate than from room temperature solution
3. Add sodium acetate Add 25 ml 10% sodium acetate Improves precipitate properties
4. Mix well Mix by rotation and inversion of urine bottle
5. Add ammonium hydroxide Add ~10 ml conc. ammonium hydroxide (14.5 M). Verify pH 11 using pH paper; add more NH4OH if needed. graphic file with name nihms479079t1.jpg Initiates precipitation processb
6. Mix well Mix by rotation and inversion of urine bottle
7. Place in cold room overnight Tilt urine bottle at 45° angle and place in cold room. Allow to settle overnight. Precipitate settles in corner for easy decanting of supernatantb
8. Decant supernatant into waste container, leaving precipitate (ppt) plus ~20 ml supernatant Self-explanatory
9. Transfer ppt from urine collection bottle to 50-ml centrifuge tube Suspend ppt in the ~20 ml residual supernatant and pour into tared 50-ml PE centrifuge tube graphic file with name nihms479079t2.jpg Ppt transfer need not be complete – final AMS measurement, ratio of 41Ca to Ca, is unaffected by recovery
10. Complete transfer of ppt to 50-ml tube and centrifuge Rinse urine bottle with 5% ammonium oxalate twice (2×10 ml) to complete transfer of ppt to centrifuge tube. Centrifuge (2000×g, 5 min) and decant supernatant.
11. Wash ppt 2x with dilute ammonium oxalate and centrifuge Suspend ppt in 20 ml 0.5% ammonium oxalate (1:10 dilution of 5% solution), centrifuge as above, pour off supernatant. Repeat. Removes sodium and potassium, including 41K, from calcium oxalate precipitate
12. Dry ppt with acetone Add 20–30 ml acetone and thoroughly suspend ppt, centrifuge, pour off supernatant, allow ppt to dry. Re-weigh tube and subtract tare weight to obtain weight of dried ppt. Final 50-ml centrifuge tube, with precipitate, can be capped and shipped for AMS analysis
Open in a new tab
a

Typically applied to ~700-ml urine sample stored in 1-L wide-mouth polyethylene bottle. Amounts in Steps 1–5 are ‘per 100 ml urine’; amounts in remaining steps refer to entire precipitate, independent of starting urine volume.

b

If precipitate was formed at pH 3–5, Ca was pure (no Mg), but precipitates did not consistently settle by gravity.

Split-Sample Comparison

MH Bhattacharyya provided training in the modified protocol to personnel in the laboratory of SK Hui at the University of Minnesota. After training was complete, nine individual 41Ca-containing human urine samples from a University of Minnesota study were each split. One 400-ml portion of each was shipped to Purdue University for calcium precipitation by their standard protocol in the laboratory of C Weaver; the remaining portion of each was retained at the University of Minnesota for calcium precipitation by the modified protocol in the laboratory of SK Hui. All precipitates were analyzed for 41Ca:Ca ratio by AMS at the Purdue PRIME Lab (method below). The University of Minnesota study was approved by the University of Minnesota Institutional Review Board (IRB). Signed consent forms were obtained from all study subjects.

Between-Study Comparison

This comparison evaluated the extent to which improvements achieved via the modified protocol persisted when samples were obtained from separate studies. Thirty-six 41Ca-containing human urine samples (500–700 ml per sample), obtained from a study of postmenopausal women conducted by MH Bhattacharyya at Argonne National Laboratory, were processed by the modified large-volume protocol in MH Bhattacharyya’s laboratory. Twenty-six 41Ca-containing human urine samples (≥ 400 ml urine per sample), obtained from a study of postmenopausal women conducted by C Weaver at Purdue University, were processed by the standard large-volume protocol in C Weaver’s laboratory. All precipitates were analyzed for 41Ca:Ca ratio by AMS at the Purdue PRIME Lab. The samples for this comparison, identified retrospectively, comprised the 62 samples from the latter two studies that were analyzed during the same AMS run on the same two sample wheels. Each study was approved by the appropriate Institutional Review Board (IRB). Signed consent forms were obtained from all study subjects.

Analysis of 41Ca:Ca Ratios by AMS

Each calcium precipitate was dissolved in 50 mL of 0.25M HNO3 solution and chromatographically purified by cation-exchange using Bio-Rad AG 50W-X8 resin. Calcium fluoride (CaF2) was precipitated with hydrofluoric acid (HF), washed, dried in a vacuum oven, loaded into aluminum sample holders, and inserted into the ion source of the accelerator mass spectrometer (AMS) to obtain the 41Ca:Ca ratio in the Purdue PRIME Laboratory (Jackson et al. 2001). The mean value of each 41Ca:Ca isotope ratio was determined by normalizing the measured value against the measured values of two standards, one measured before and one measured after each unknown. The two standards used had a 41Ca:Ca ratio of either 3.7×10−10 or 2.0×10−9. Each sample was measured multiple times, and the uncertainty was calculated from counting statistics and systematic errors (Elmore et al., 1984).

Samples for the Between-Study Comparison were analyzed before those for the Split-Sample Study, and a new detector/software combination was in place at the Purdue AMS when samples for the Split-Sample Comparison were run. This new detector/software combination tolerated higher levels of interference and did not generate a separate value for tails, as reflected in results reported in Table 2, Split-Sample Comparison. All AMS analyses within a given comparison (i.e., Split-Sample Comparison, Between-Study Comparison) were conducted over a single two-day period, with each set of samples present on the same one or two sample wheels. This approach ensured that all samples within a given comparison were run under similar instrument conditions. In addition, for the Split-Sample comparison, a sample prepared by one method was analyzed and then immediately followed by a sample prepared by the other method, with the order randomized, to decrease any effects of fluctuations in the run conditions.

Table 2.

AMS results for urine samples processed by modified vs. standard protocola

Parameter Modified
Protocol		 Standard
Protocol
Split-Sample Comparisonb
Samples Analyzed: n 18 18
Ion Current: nA 543 ± 21* 452 ± 34
Low Current<150 nA: n (%) 0 0
High Current>500 nA: n (%) 11 (61) 5 (28)
Unreliabled<75 nA: n (%) 0 0
Interferencee: cps 452 ± 45* 694 ± 65
Between-Study Comparisonc
Samples Analyzed: n 36 26
Ion Current: nA 431 ± 24* 307 ± 34
Low Current (<150 nA): n (%) 2 (6) 5 (19)
High Current (>500 nA): n (%) 15 (42) 3 (12)
Unreliabled (<75 nA): n (%) 0 (0) 2 (8)
Interferencee: cps 105 ± 5 104 ± 3
Tailsf: cpm 116 ± 10* 183 ± 30
Open in a new tab
*

Mean for modified protocol significantly different from mean for standard protocol (twotailed, independent T-test for groups with unequal variances; p<0.05).

a

Values with uncertainties are mean ± SE for number of samples analyzed, n. For values given as number of samples, n, percentages are shown in parentheses

b

Urines for Split-Sample Comparison were nine individual samples each split and processed by the two protocols, as described in text. Number of samples analyzed is 18 because each sample for a given protocol was again split, and each split was analyzed individually by AMS.

c

Urines for Between-Study Comparison encompass 62 samples: 36 ANL-Study samples and 26 Purdue-Study, as described in text.

d

”Unreliable’ indicates sample produced < 75 nA of ion current, which translates to < 10 nA of analyzed beam on the high energy side. These currents are too low to give reliable results.

e

Interference is the total count rate of everything in the detector. Note that the new detector/software combination in place in the AMS when split-sample urines were run tolerated higher interference rates.

f

Tail value is the counts per minute of interferences that made it past the Purdue PRIME Lab hardware gates and were gated out by the data analysis software. Note that the new detector/software combination in place in the AMS when split-sample urines were run did not provide a separate tail value.

Additional Methods

To determine total recovery of urine calcium in the precipitate, samples of urine were diluted with 0.1 N HCl/0.9% LaCl3, and urine calcium concentrations were determined by flame atomic absorption spectrometry (Varian SpectrAA 220, air/acetylene flame, deuterium background correction) (Varian Analytical Instruments, Sugar Land, TX) (Trudeau and Freier, 1967). Ca standards were prepared in 0.1 N HCl/0.9% LaCl3. Elemental analyses for Mg and K were conducted by ICP-MS (Thermo Fisher iCAP 6300 ICP mass spectrometer).

RESULTS

The recovery of urine calcium in the oxalate precipitate using the modified protocol was nearly complete: 96% ± 3% (mean ± SD, n=3 separate urine samples, each from a different individual). Regarding purity, calcium oxalate monohydrate accounted for 66 ± 14% of the total weight of the precipitate (mean ± SD, n=3 separate urine samples). Magnesium precipitated along with calcium for both protocols: Mg:Ca ratios (weight percent) [mean ± SE (n)] for the modified vs. standard protocols were 9% ± 6% (9) vs. 37% ± 5% (9) for the urines from the Split-Sample Comparison and 34% ± 8% (33) vs. 76% ± 31% (8) for urines from the earlier Between-Study Comparison.

Table 2 presents AMS results that were obtained when calcium oxalate was precipitated by the modified protocol vs. the standard protocol, for two separate comparisons. Beyond its simplicity, one of the biggest advantages of the modified protocol was the significant increase in average ion current. For the split-sample comparison, the average ion current was 20% higher for the modified than the standard protocol samples (p<0.05, two-tailed, independent t-test for groups with unequal variances), and twice as many samples prepared by the modified protocol had high ion currents (>500 nA) (Table 2). For the between-study comparison, this advantage resulted in about half as many modified protocol samples with low current (<150 nA) and five times as many with high current (>500 nA). Most important, no samples prepared by the modified protocol fell in the ‘unreliable’ category. In contrast, two of the 26 samples prepared by the standard protocol had currents <75 nA, typical of the ~5–10% of standard protocol samples whose results are not used because ion current values are too low.

The overall interference rate in the detector – counts from everything that is not 41Ca – was significantly lower for the modified protocol samples in the Split-Samples Study, even though their total ion current values were higher (Table 2). Similarly for the Between-Study Comparison, the relative amount of interference (cps per nA ion current) was also lower for the modified protocol samples, because their ion currents were higher: 0.24 cps/nA vs. 0.34 cps/nA for modified vs. standard protocol samples, respectively. In addition, for the Between-Study Comparison, which gave output on tails rate, samples prepared by the modified method had a count rate in the tails region that was significantly lower than did those prepared by the standard method (116 cpm vs. 183cpm) (p < 0.05, two-tailed, independent t-test for groups with unequal variances) (Table 2). This decrease in tail was largely due to the significantly lower K:Ca ratio of the calcium precipitate for these samples (weight percent): 0.44% ± 0.1% (33) vs. 3.6% ± 1.0% (8) for the modified vs. standard protocol samples, respectively (mean ± SE for number of samples in parentheses) (p<0.05). In fact, the spectra from anode 3 in the gas ionization detector (Knies and Elmore, 1994) showed that the 41K peak was much smaller than the 41Ca peak for a modified protocol sample, while the 41K peak for a comparable standard protocol sample was somewhat larger than the 41Ca peak (data not shown). When the two protocols were applied to the same samples as in the split-sample comparison, a substantive decrease in K:Ca ratio was still achieved with the modified protocol (mean weight percent): 0.08% ± 0.03% (8) vs. 0.54% ± 0.20% (8) for the modified vs. standard protocol samples, respectively (mean ± SE for number of samples in parentheses).

As shown in Table 3 (top), all nine Split-Urine samples gave the same 41Ca: Ca ratio independent of the method applied for preparation of the calcium precipitate. In addition, the ratio measurements were highly precise independent of protocol, with mean RUC values of 3.5% to 4.0%. In contrast, the 41Ca: Ca ratio measurements for the Between-Study Comparison samples were not as precise, in particular for samples prepared by the standard protocol. The latter samples had mean RUC values of 8.2% to 9.0% -- two- to three-fold higher than RUC values for the modified protocol samples (Table 3, bottom). The standard protocol samples retained the higher uncertainty values even when the 41Ca:Ca ratios were matched as closely as the two datasets would allow (Subset 1 vs. Subset 1A, Table 3, bottom).

Table 3.

41Ca:Ca ratios for calcium precipitates prepared by modified vs. standard protocola

Protocol Sample 41Ca:Ca ×1010 %RUC
Split-Sample Comparisonb
  Modified 1 0.641 ± 0.022 (2) 3.5
  Standard 1 0.656 ± 0.025 (2) 3.8
  Modified 2 0.528 ± 0.019 (2) 3.7
  Standard 2 0.560 ± 0.036 (2) 6.8
  Modified 3 0.475 ± 0.022 (2) 4.6
  Standard 3 0.478 ± 0.033 (2) 6.8
  Modified 4 0.523 ± 0.014 (2) 2.7
  Standard 4 0.538 ± 0.015 (2) 2.7
  Modified 5 0.346 ± 0.013 (2) 3.8
  Standard 5 0.332 ± 0.012 (2) 3.6
  Modified 6 0.416 ± 0.025 (2) 6.0
  Standard 6 0.423 ± 0.019 (2) 4.5
  Modified 7 32.30 ± 0.75 (2) 2.3
  Standard 7 28.76 ± 0.74 (2) 2.6
  Modified 8 54.9 ± 1.4(2) 2.6
  Standard 8 52.3 ± 1.3 (2) 2.6
  Modified 9 3.18 ± 0.077 (2) 2.4
  Standard 9 3.26 ± 0.092 (2) 2.8
Between-Study Comparisonc
  Modified All samples (ANL study) 4.05 ± 0.150 (36) 3.7
  Standard All samples (Purdue study) 1.06 ± 0.096 (26) 9.0
  Modified Subset 1 (ANL study)d 2.01 ± 0.060 (3) 3.0
  Standard Subset 1 (Purdue study)d 1.46 ± 0.132 (12) 9.0
  Modified Subset 1 (ANL study)d 2.01 ± 0.060 (3) 3.0
  Standard Subset 1A (Purdue study)e 1.87 ± 0.153 (3) 8.2
Open in a new tab
a

Values for 41Ca:Ca ratios are mean ± UC (n), with UC = uncertainty of AMS measurement, as described in text; n = number of samples. %RUC = relative uncertainty expressed as percentage of mean.

b

Split urine samples are the nine urine samples identified in Table 2, footnote b.

c

Between-Study samples are the 62 urine samples identified in Table 2, footnote c. Because mean 41Ca:Ca ratio was higher for Argonne National Laboratory (ANL) study, Subsets 1 and 1A, which have closer mean values (as described in footnotes d and e), were also evaluated

d

Subsets 1 include all Between-Study urine samples with 41Ca:Ca ratios (×1010) between 1.0 and 2.1; they provide largest number of samples with better equality of 41Ca:Ca means for the two protocols.

e

Subset 1A includes the three samples from the Purdue set with the highest ratios; it provides the best possible equality of 41Ca:Ca means, although with a small number of samples in each group (3).

DISCUSSION

The modified protocol reported here was devised based on the early literature on gravimetric analysis of calcium in biological samples (McCrudden, 1910; Stohl and Pedley, 1921; Washburn and Shear, 1932). It offers an optimal method to obtain a calcium oxalate precipitate from a large volume of human urine that served well for 41Ca analysis by accelerator mass spectrometry at Purdue University. A cumbersome and expensive filtration step was eliminated, while significantly increasing the quality of the AMS measurement.

Compared to previously published low-volume protocols (Fitzgerald et al., 2005; Freeman et al., 1995; Freeman et al., 1997; Lin et al. 2004), the modified protocol generates a large amount of calcium precipitate without addition of exogenous calcium. While addition of stable calcium to low-volume samples produces a consistent sample size and uniform current output for the AMS measurement, this addition sacrifices sensitivity by decreasing the 41Ca:Ca ratio. The modified protocol minimizes the amount of 41Ca that needs to be administered to subjects and maximizes the amount of time after dosing during which samples have 41Ca:Ca ratios that are well above background. The latter feature is significant for studies that require long-term follow up of subjects, for example, for research into cancer survivor’s bone health.

The sets of samples compared in this study were each measured during the same time period during the same AMS run under similar foil and ion conditions. This approach minimized differences that might otherwise be attributable to differences in instrument parameters. Despite the fact that the samples prepared with the modified method had higher ion currents, values for both total interference rate (cps) for the split-urine samples and tails rate (cpm) for the between-study samples were significantly lower (p < 0.05). The decrease in tails rate would translate to a significant decrease in background. The mean K:Ca ratio for the standard protocol samples from the between-study comparison (3.6% w/w) was at least seven-fold higher than for all other samples in Table 3 (≤ 0.5% w/w). These higher K:Ca ratios would translate to greater interference from 41K, as documented by the AMS spectra, most likely contributing to the higher uncertainty values obtained for these standard protocol samples (Table 3, bottom).

Simplified, faster chemical processing, higher ion currents, and lower interference rates are key factors for translation of this technology to clinical applications. The modified protocol described in this report was recently implemented in a clinical trial whose aim is to gain insight into the acute (short-term) and prolonged (long-term) bone complications associated with radiation and chemotherapy in cancer survivors (SK Hui, data not shown). By extension, this detailed and focused technical report on sample methodology will help others to test the application of 41Ca in a clinical laboratory setting as a diagnostic tool to monitor patients’ bone health, with the ultimate goal to limit morbidity from fracture and improve quality of life among cancer survivors and the elderly.

Highlights.

  • ▪

    A method was optimized to precipitate calcium oxalate from large-volume urine samples

  • ▪

    The modified protocol resulted in AMS measurements with significantly higher ion current and decreased interference

  • ▪

    Result is lower effort and cost and higher sensitivity for bone remodeling studies

ACKNOWLEDGMENTS

We would like to acknowledge Mr. Tom Kubley, Purdue University, for organizing operation of the AMS during these measurements; Ms. Anna Kempa-Steczko, Purdue University, for preparing the oxalate precipitates; and Dr. Manju Sharma, University of Minnesota, for helping with urine chemistries. Research for MHB, JJM, and SPC was funded by GHSU start-up funds (Departments of Orthopedic Surgery and Medicine) and Phillip Morris USA, Inc. and Phillip Morris International. Research at Purdue University (GSJ, CMW, JE) was funded by Grant P50-AT00477 from the National Institutes of Health. Research at University of Minnesota (SKH) was funded by the National Institute of Child Health and Human Development (1K12-HD055887-01), National Institute of Arthritis and Musculoskeletal Disease (RO3 AR055333-02), and Cancer Center Support Grant P30 CA77398.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

REFERENCES

  1. Bhattacharyya MH. Cadmium osteotoxicity in experimental animals: mechanisms and relationship to human exposures. Toxicol. Appl. Pharmacol. 2009;238:258–265. doi: 10.1016/j.taap.2009.05.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Brown K, Dingley KH, Turteltaub KW. Accelerator mass spectrometry for biomedical research. Methods in Enzymology (Biological Mass Spectrometry) 2005;402:423–443. doi: 10.1016/S0076-6879(05)02014-8. [DOI] [PubMed] [Google Scholar]
  3. Cheong JMK, Martin BR, Jackson GS, Elmore D, McCabe GP, Nolan JR, Barnes S, Peacock M, Weaver CM. Soy isoflavones do not affect bone resorption in postmenopausal women: a dose-response study using a novel approach with 41Ca. J. Clin. Endo. Metab. 2007;92:577–582. doi: 10.1210/jc.2006-0369. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Denk E, Hillegonds D, Vogel J, Synal A, Geppert C, Wendt K, Fattinger K, Hennessy C, Berglund M, Hurrell RF, Walczyk T. Labeling the human skeleton with 41Ca to assess changes in bone calcium metabolism. Anal. Bioanal. Chem. 2006;386:1587–1602. doi: 10.1007/s00216-006-0795-5. [DOI] [PubMed] [Google Scholar]
  5. Denk E, Hillegonds D, Hurrell RF, Vogel J, Fattinger K, Häuselmann HJ, Kraenzlin M, Walczyk T. Evaluation of 41calcium as a new approach to assess changes in bone metabolism: effect of a bisphosphonate intervention in postmenopausal women with low bone mass. J. Bone Miner. Res. 2007;22:1518–1525. doi: 10.1359/jbmr.070617. [DOI] [PubMed] [Google Scholar]
  6. Ebert-McNeill A, Clark SP, Miller JJ, Birdsall P, Chandar M, Wu L, Cerny EA, Hall PH, Johnson MH, Isales C, Chutkan N, Bhattacharyya MH. Cadmium intake and systemic exposure in postmenopausal women and age-matched men who smoke cigarettes. Toxicol Sci. 2012;130(1):191–204. doi: 10.1093/toxsci/kfs226. [DOI] [PubMed] [Google Scholar]
  7. Elmore D, Bhattacharyya MH, Sacco-Gibson N, Peterson DP. Calcium-41 as a long-term biological tracer for bone resorption. Nucl. Instr. and Meth. B. 1990;52:531–535. [Google Scholar]
  8. Elmore D, Conard N, Kubik PW, Fabryka-Martin J. Computer controlled isotope ratio measurements and data analysis. Nucl. Instrum. Methods Phys. Res. B. 1984;5(2):233–237. [Google Scholar]
  9. Fitzgerald RL, Hillegonds DJ, Burton DW, Griffin TL, Mullaney S, Vogel JS, Deftos LJ, Herold DA. 41Ca and accelerator mass spectrometry to monitor calcium metabolism in end stage renal disease patients. Clinical Chemistry. 2005;51(11):2095–2102. doi: 10.1373/clinchem.2005.049650. [DOI] [PubMed] [Google Scholar]
  10. Freeman SPHT, Serfass RE, King JC, Southon JR, Fang Y, Woodhouse LR, Bench GS, McAninch JE. Biological sample preparation and 41Ca AMS measurement at LLNL. Nucl. Instr. and Meth. B. 1995;99:557–561. [Google Scholar]
  11. Freeman SPHT, King JC, Vieira NE, Woodhouse LR, Yergey AL. Human calcium metabolism including bone resorption measured with 41Ca tracer. Nucl. Inst. and Meth. in Phys. Res. B. 1997;123:266–270. [Google Scholar]
  12. Jackson GS, Weaver C, Elmore D. Use of accelerator mass spectrometry for studies in nutrition. Nutrition Research Reviews. 2001;14:317–334. doi: 10.1079/NRR200129. [DOI] [PubMed] [Google Scholar]
  13. Johnson RR, Berkovits D, Boaretto E, Gelbart Z, Ghelberg S, Meirav O, Paul M, Prior J, Sossi V, Venczel E. Calcium resorption from bone in a human studied by 41Ca tracing. Nucl. Inst. and Meth. in Phys. Res. B. 1994;92:483–488. [Google Scholar]
  14. Knies DL, Elmore D. The PRIME Lab gas ionization detector. Nucl. Instr. & Meth. B. 1994;92:134–137. [Google Scholar]
  15. Lee WH, Wastney ME, Jackson GS, Martin BR, Weaver CM. Interpretation of 41Ca data using compartmental modeling in post-menopausal women. Anal. Bioanal. Chem. 2011;399:1613–1622. doi: 10.1007/s00216-010-4454-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Lin Y, Hillegonds DJ, Gertz ER, Van Loan MD, Vogel JS. Protocol for assessing bone health in humans by tracing long-lived 41Ca isotope in urine, serum, and saliva samples. Anal. Biochem. 2004;332:193–195. doi: 10.1016/j.ab.2004.05.019. [DOI] [PubMed] [Google Scholar]
  17. McCrudden FH. The quantitative separation of calcium and magnesium in the presence of phosphates and small amounts of iron devised especially for the analysis of foods, urine and feces. J. Biol. Chem. 1910;7:83–100. [Google Scholar]
  18. Shohl AT, Pedley FG. A rapid and accurate method for calcium in urine. J. Biol. Chem. 1921;49:537–544. [Google Scholar]
  19. Trudeau DL, Freier EF. Determination of Calcium in Urine and Serum by Atomic Absorption Spectrophotometry (AAS) Clin. Chem. 1967;13:101–114. [PubMed] [Google Scholar]
  20. Washburn ML, Shear MJ. Composition of bone XIII. Direct gravimetric determination of Ca Mg, and PO4. J. Biol. Chem. 1932;99:21–41. [Google Scholar]
  21. Weaver CM, Martin BR, Jackson GS, McCabe GP, Nolan JR, McCabe LD, Barnes S, Reinwald S, Boris ME, Peacock M. Antiresorptive effects of phytoestrogen supplements compared with estradiol or risedronate in postmenopausal women using 41Ca methodology. J. Clin. Endo. Metab. 2009;94:3798–3805. doi: 10.1210/jc.2009-0332. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES