Printer Friendly

Comparison of specific gravity and creatinine for normalizing urinary reproductive hormone concentrations.

Urine specimens have several advantages over serum, saliva, and blood spots for monitoring reproductive hormone patterns. Urine collection is noninvasive, poses minimal infectious disease risk to participants and researchers, and provides sufficient volume for multiple assays and future research. Urine specimens are ideally suited for large studies because they can be collected and stored by participants, and compliance is high (1-3). Urine also provides integrated hormone measures without the confounding effects of pulsatile secretion (4). Finally, urinary concentrations of reproductive steroid hormone metabolites are generally higher than the circulating serum concentrations, providing better quantification of the lower end of the physiologic scale in humans, which can be close to or below the limit of detection in serum RIAs (5). Urinary hormone results must, however, be adjusted to account for the concentration of the specimen, which depends on an individual's hydration status and time since last urination.

We examined the performance of two methods of adjustment: specific gravity (SG) [3] correction and creatinine (CR) correction. Urine SG is the ratio of the density of a urine specimen to the density of water (6). SG increases with solute concentration and is most accurately measured by refractometry (6). CR, a byproduct of muscle activity, is cleared from the bloodstream by the kidneys and excreted in urine (7). Urinary CR concentrations are determined by colorimetric assay (8), and analyte concentrations are usually reported as a ratio of the analyte concentration to CR concentration.

On the basis of an early finding that daily individual CR excretion was fairly consistent (9), urinary CR became a common method of assessing kidney function in clinical settings and correcting for analyte concentrations, including reproductive hormones, in urine (4). Use of CR to normalize urinary analyte concentrations can be problematic, however, because there is evidence that CR excretion is not consistent: numerous studies have found considerable inter- and intrasubject variability in CR values and dependence on sex, age, activity, and diet (10-18). Population variation in CR excretion may also exist, but to the best of our knowledge, this topic has not been investigated.

SG is an alternative method with several advantages over CR, although it is not widely used, perhaps in part because of a lack of data evaluating its performance relative to CR. In this study we compared SG and CR correction methods on urinary hormone metabolite concentrations from healthy US women, using serum hormone measurements as the standard. We then applied both methods to spot specimens from Bangladeshi women to evaluate their applicability in a nonindustrialized population.

Materials and Methods


A total of 799 daily urine and serum specimens were collected over one menstrual cycle from 30 US women in 1997-1998. Thirteen women 20-25 years of age and 17 women 40-45 years of age were recruited for a study on reproductive aging. Monetary compensation was provided, participants provided written informed consent, and all procedures were approved by the Institutional Review Board of the University of Washington. All participants were normally cycling, in good health, had a mean body mass (SD) index of 22.6 (2.36) kg/[m.sup.2] (range 18.9-27.7 kg/[m.sup.2]), and were not using medications or hormones. Blood specimens were obtained by venipuncture, beginning with the first day of menstrual bleeding and continuing until the first day of menstrual bleeding of the subsequent cycle. Serum specimens were immediately assayed, and all cycles were confirmed ovulatory by transvaginal ultrasound. Urine specimens were taken daily in the clinic, usually before 1200, at the same time as serum collection and immediately stored at -20 [degrees]C. Urine specimens remained frozen until thawing 2 years later for aliquoting, assay, and measurement of SG (19). The specimens underwent two to three more freeze-thaw cycles before CR assay in 2003.

For the Bangladeshi sample, 13 cycling women were selected randomly from a sample of women participating in a 9-month study of early pregnancy loss. All participants were married, noncontracepting residents in the nonintervention demographic surveillance region of the rural Matlab district in Bangladesh (20). No monetary compensation was provided, all participants provided written informed consent, and all procedures were approved by the Institutional Review Boards of The Pennsylvania State University and the International Centre for Diarrhoeal Disease and Research, Bangladesh (20). Spot urine specimens were collected by community healthcare workers every 3 or 4 days over the course of one menstrual cycle in 1993. The specimens were stored at the healthcare workers' homes in a cooler with ice packs for 1-3 days until they were transported to a field hospital and stored at 4 [degrees]C. One to three days later, the specimens were brought to room temperature, and SGs were taken. The specimens were preserved with 0.17 g/mL boric acid solution, stored at -20 [degrees]C, and transported to the US by frozen air freight. Specimens remained frozen until 1995, when they were thawed and assayed for CR and steroid hormone metabolites.


All serum specimens were assayed by RIA for estradiol (E2) and by immunofluorometric assay for luteinizing hormone (LH), but serum progesterone (P4) was measured only in the luteal phase. The RIA for E2 (ICN Biomedicals) cross-reacts 20% with estrone, 1.5% with estriol, and <1% with all other steroids. The inter- and intraassay CVs were 16% and 7%, respectively. The interand intraassay CVs for the LH immunofluorometric assay (Delphia) were 2.8% and 4.7%, respectively. The RIA for P4 (Diagnostic Systems Laboratories) cross-reacts <5% with all other steroids, and the inter- and intraassay CVs were 13% and 11%, respectively.

Urine specimens were analyzed by competitive enzyme immunoassays (EIAs) for urinary steroid hormone metabolites. All US and Bangladeshi urine specimens were assayed for pregnanediol glucuronide (PDG) and estrone conjugate (E1Cs). The PDG and E1C EIAs have been described elsewhere (19). Briefly, the PDG assay uses the monoclonal antibody Q330 (Quidel Corporation) and reference calibrator 5[beta]-pregnane-3[alpha],20[alpha]-diol glucuronide (Sigma; cat. no. P3635). The inter- and intraassay CVs for the PDG assay were 10.3% and 9.2%, respectively (19). E1Cs were measured with the 155-B3 monoclonal capture antibody (Dr. Fortune Kohen, Weizmann Institute, Rehovet, Israel) and estrone-[beta]-d-glucuronide reference calibrator (Sigma; cat. no. E1752). The inter- and intraassay CVs were 10.9% and 7.3% (19). The absorbance for the EIAs was measured with a Dynatech MR7000 Plate Reader (test wavelength, 405 nm; reference wavelength, 570 nm). All specimens were run in duplicate on microtiter plates, and hormone concentrations were estimated from absorbance by use of a four-parameter logistic model in Biolinx 1.0 Software (Dynex Laboratories, Inc).

Urinary CR was measured by reaction with sodium hydroxide and picric acid in the method described by Jaffe (8) with calibrators purchased from Sigma (cat. no. 925-11). The absorbance of specimens was read with a Dynatech MR7000 Plate Reader (test wavelength, 490 nm; reference wavelength, 630 nm), and CR concentrations were estimated with a four-parameter logistic model in Biolinx 1.0 Software. The inter- and intraassay CVs for the CR assay were 1.6% and 14%, respectively.

The CR correction formula applied to each sample was as follows:

CR-corrected concentration (nmol/nmol) = raw hormone concentration (nmol/volume)/CR (nmol/volume)

SG measurements were taken with a hand-held urine SG refractometer (Atago Uricon-PN; NSA Precision Cells, Inc). The SG of Bangladeshi urine specimens was measured before freezing, whereas the SG for the US specimens was measured after the first freeze-thaw cycle. The correction formula applied to each hormone result was as follows:

SG-corrected concentration = raw hormone concentration x ([] - 1.0)/([SG.sub.sample] - 1.0)

where [] is a population mean SG (20). The target SG used was 1.020 (21) for US specimens and 1.015 (20) for Bangladeshi specimens.


Pearson correlations between serum hormone values and raw, SG-corrected, and CR-corrected urinary hormone results were calculated with an optimum lag day. Individual cycles were aligned by day of midcycle serum LH peak (day 0), and a mean hormone value for each cycle day was calculated (n = 34 paired urine/serum cycle days from 30 cycles). The effect of time between appearance of hormones in serum and clearance into the urine was evaluated by correlating serum results with urine results lagged by -1, 0, 1, 2, and 3 days behind serum. The optimum lag day was selected on the basis of the highest correlation with serum results.

The correlations of serum E2 with urinary E1Cs and serum P4 with urinary PDG were weighted by the number of specimens available for each cycle day. The 95% confidence intervals for the difference between SG-corrected urine vs serum, CR-corrected urine vs serum, and uncorrected urine vs serum mean correlations were obtained by use of bootstraps. The hypothesis that two correlations were the same was rejected when the confidence interval did not include zero.

We used a linear mixed-effects model to test for differences in mean CR or SG by population (US vs Bangladeshi). Mixed-effects models allow the significance of a fixed effect (in this case, the mean CR or SG difference between the US and Bangladeshi populations) to be tested while accounting for within-subject repeated measures by modeling a random effect for each individual as follows:

y = a + bx + [c.sub.i] + [e.sub.ji]

where y is either a CR or SG value; x is 0 for US and 1 for Bangladesh; a is the estimated CR or SG for the US sample; b is the estimated difference between CR or SG between Bangladesh and the US; ci is a random effect estimated for the ith individual; and [e.sub.ji] is the residual error associated with the jth specimen from the ith individual. We also used linear mixed-effects models to compare between- and within-subject (residual) variability for each population in separate models by estimating the SD for the ci term, which represents between-subject variability, and the [e.sub.ji] term, the within-subject variability.


For both assays, the highest correlation between urine and serum values occurred on the same lag day for uncorrected, SG-corrected, and CR-corrected values (Table 1). PDG and E1C results lagged 1 day behind their respective serum values. When optimum lag day was used, weighted correlations between serum and SG-corrected urinary hormone values were high for both assays, ranging from 0.94 to 0.97 (Table 2). Correlations were also high for CR-corrected (0.93-0.98) and uncorrected urine results (0.92-0.93; Table 2). There was no significant difference in the correlations between SG-corrected urine vs serum values and CR-corrected urine vs serum for either hormone (Table 3). In the progesterone assays, there was a significant difference between uncorrected urine vs serum correlations and both SG-corrected urine vs serum and CR-corrected urine vs serum correlations (Table 3).

All US specimens had a usable CR result, whereas 37% of Bangladeshi specimens had CR values below the assay limit of detection (0.002 nmol/L) even when assayed undiluted. In the linear mixed-effects analyses, the US sample had significantly higher CR concentrations than the Bangladeshi sample (mean US, 11.58 nmol/L; mean Bangladeshi, 1.58 nmol/L; P <0.001). Fig. 1 shows the distribution of CR values for each US and Bangladeshi participant and illustrates the large difference in CR values between the two populations. US specimen interand intrasubject SD were 4.8 and 0.16 nmol/L, respectively. In Bangladeshi specimens, these measures were 1.14 and 0.27 nmol/L, showing greater intrasubject variation in CR values. The ratio of within- to between-subject SD for CR in the US sample was almost seven times smaller than the Bangladeshi sample (0.033 for US vs 0.24 for Bangladesh).

When hormone profiles of individual Bangladeshi women were corrected with CR, extremely low CR values produced spuriously high corrected hormone results (Fig. 2). Very low CR values in the Bangladeshi sample caused unreasonably poor concordance between uncorrected and CR-corrected hormone results (Fig. 3A). For example, if low (<0.1 nmol/L) CR specimens were removed, the correlation between CR-corrected and uncorrected PDG results increased (Fig. 3B). However, all of these specimens with unusable CR results gave an acceptable correlation with uncorrected PDG values when adjusted by SG (Fig. 3C).

The distribution of SG values for each US and Bangladeshi participant is shown in Fig. 4. Unlike for CR, the distributions of SG values for the two populations were similar. The mean SG values in the US (1.014) and Bangladesh (1.013) specimens did not differ significantly (P = 0.50). The ratio of within- to between-subject SD in SG values was 1.52 for the US sample and 1.22 for the Bangladeshi sample, indicating a similar amount of variation in this measure for the two populations.


Our results show that SG performs as well as CR in correcting steroid hormone metabolite results for urine concentration, in agreement with other studies comparing SG and CR normalization of urinary analyte concentrations (22, 23). We demonstrated that SG adjustment is applicable in specimens with CR concentrations that are not sufficiently high to correct metabolite concentrations. We have also provided evidence of a population concentration difference in urinary CR values and concurrent population concentration similarities in SG measurements. Currently, CR correction is widely used to adjust for urine analyte concentration in spot and first-morning specimens, although several studies have concluded that it offers no advantage over unadjusted results (24-27). Our results did indicate a small but significant improvement in PDG correlation with serum P4 values when adjusted by CR and SG. Although CR and SG perform similarly, SG offers some advantages, particularly for large-scale research on reproductive hormones.

SG is inexpensive; a handheld refractometer and transfer pipettes are the only equipment needed, whereas the CR assay is a microtiter plate-based assay. The CR assay uses picric acid, a harsh physical irritant that is explosive when dry, so special use and storage conditions are necessary. Finally, although multiple samples can be assayed simultaneously for CR, it does not offer a time advantage over SG. An equal number of samples can be measured on the refractometer in the time needed to complete the CR assay.

SG refractometers measure urine density via the ratio of light refraction between air and a urine specimen; refraction increases with solute concentration of the specimen (28). Urine density, however, varies with the total mass of solutes, which depends not only on the number of particles present, but also on their molecular weight. Therefore, SG is affected more by the presence of heavy molecules such as glucose, albumin, phosphates, sulfates, radiocontrast media, and heavy metals than it is by low-molecular-weight substances such as sodium, chloride, and urea (6, 29, 30). Thus, SG correction may not be an appropriate method for individuals with diabetes mellitus and nephrotic syndrome, which cause high concentrations of glucose and protein in urine, increasing SG and underestimating urine analyte concentrations (30). Diabetes, starvation, and dehydration produce ketones from fat metabolism. The ketones are excreted in urine and erroneously lower SG readings because they are less dense than water (28). Several authors have also indicated that urine solute concentrations given by SG correction may be inaccurate if the urine specimen is very dilute or very concentrated (31, 32). However, a minimum acceptable SG of 1.010 (16, 33) seems too high; 31% (247 of 799) of our specimens from healthy US women had SG values lower than this, and our averaged SG-corrected urine vs serum correlations were very high. SG readings are also affected by temperature fluctuations, which make specimens expand or contract, altering their density (16); therefore, readings should be taken at a consistent temperature.


CR correction of urinary hormone values is problematic primarily because CR excretion exhibits inter- and intrasubject variations and is influenced by time of day, age, sex, diet, body mass, and activity level. This variation could complicate interpretations of analytes reported in ratio to CR. Because CR is a byproduct of muscle use, its production is expected to vary with body composition and activity. Edwards and Whyte (34) reported a correlation between lean body mass and urinary CR of 0.65 from a group of 31 men and women. Bleiler and Schedl (35) found the same measure to be 0.47-0.48 in 11 women and 0.53-0.55 in 51 men and reported correlations between urinary CR and weight and body surface area in 24-h specimens. Muscularity also contributes to observed sex differences in CR concentrations. Men produce more CR than women and have a higher clearance rate (17). Kesteloot and Joossens (18) reported a mean CR clearance of ~101 mL/min in 2075 men and 86.9 mL/min in 1933 women. This and other studies (36, 37) also showed a decrease in CR clearance with age, leading to increased serum and decreased urinary concentrations at older ages, making CR correction particularly questionable for adjusting urinary hormone values in research concerning aging, such as the transition to menopause.


Diets with substantial amounts of particular kinds of meat, such as beef, can also affect urinary CR concentrations. Meat contains creatine, the precursor of CR, which can accumulate in the body and lead to a gradual increase in CR excretion (7). Cooking meat converts creatine to CR, which is quickly excreted and can cause considerable short-term CR increases in the hours after ingestion (10, 15).

Population differences in activity, nutritional status, and body composition may account for the differences observed between the Bangladeshi and US samples, and they challenge the use of CR as a correction method for this population. Most women in Bangladesh suffer from chronic undernutrition and infectious disease and have limited access to healthcare (38). They also have very low body mass indexes: the mean (SD) body mass index for a large random sample of nonpregnant women 15-45 years of age in 1992 in Matlab, Bangladesh was 18.8 (1.9) kg/[m.sup.2] (38). Thus, the low and variable CR concentrations in our Bangladeshi samples may reflect these factors. It should be noted that the effects of differences in specimen treatment conditions between the Bangladeshi and US samples on detected CR was not tested and cannot be ruled out. However, several studies have shown that CR is very stable in refrigerated and frozen samples (39, 40), and we did not find any effect of specimen treatment conditions on urinary concentrations of E1Cs and PDG in a test of a range of treatment conditions (19).


Additionally, although the original use of CR was to check the completeness of 24-h urine specimens, many studies have shown considerable intraindividual variation in daily CR excretion. Alessio et al. (16), using four consecutive 24-h urine collections, observed variation ranging between 9.2% and 79.4% in the extreme values of 16 individuals. Similarly, Greenblatt et al. (14) found a range of 63-244% in 24-h collections from eight individuals. In a summary of studies on 24-h CR excretion, Curtis and Fogel (12) showed that some individuals had relatively consistent daily CR excretion with individual CV around 5%, whereas others had highly variable excretion with individual CV exceeding 20%.


CR correction is particularly questionable when applied to spot samples because CR excretion over short intervals also shows considerable variation. Verstergaard and Leverett (11) showed that subsequent 2-h interval samples varied by >100%, and several studies have reported that spot-sample CR variation is several times higher than variation for 24-h values (12, 13). The inapplicability of CR correction to spot specimens presents a major problem for large prospective studies on reproductive hormones. Frequent collections are needed to accurately characterize hormonal patterns, and daily 24-h urine collections are impractical, difficult to obtain, and would present storage and shipping problems.

In conclusion, this study supports the use of SG as an alternative to CR for urinary hormone concentration correction. Both methods show high correlation with serum hormone values. However, SG correction offers several practical advantages over CR. Given these advantages, we recommend SG as an alternative to CR for adjusting urinary steroid hormone metabolite concentrations, particularly in populations with very low or highly variable urinary CR concentrations.

This research was supported by the NSF (DBS-9218734 and DBS-9600690), NIA R01 AG15141, NIA R01 AG14579, the Mellon Foundation, the Hill Foundation, and the American Institute for Bangladesh Studies.

Received February 2, 2004; accepted February 25, 2004.

Previously published online at DOI: 10.1373/clinchem.2004.032292


(1.) Collins WP, Collins PO, Kilpatrick MJ, Manning PA, Pike JM, Tyler JP. The concentrations of urinary oestrone-3-glucuronide, LH and pregnanediol-3-glucuronide as indices of ovarian function. Acta Endocrinol 1979;90:336-48.

(2.) Kesner JS, Wright DM, Schrader SM, Chin NW, Kreig EF Jr. Methods of monitoring menstrual function in field studies: efficacy of methods. Reprod Toxicol 1992;6:385-400.

(3.) Lasley BL, Mobed K, Gold EB. The use of urinary hormonal assessments in human studies. Ann N Y Acad Sci 1994;709: 229-311.

(4.) Munro CJ, Stabenfeldt GH, Cragun JR, Addiego LA, Overstreet JW, Lasley BL. Relationship of serum estradiol and progesterone concentrations to the excretion profiles of their major urinary metabolites as measured by enzyme immunoassay and radioimmunoassay. Clin Chem 1991;37:838-44.

(5.) Shideler SE, DeVane GW, Kaira PS, Benirschke K, Lasley BL. Ovarian-pituitary hormone interactions during the perimenopause. Maturitas 1989;11:331-9.

(6.) Osborne CA. Urine specific gravity, refractive index, osmolality: which would you choose? DVM (The Newsmagazine of Veterinary Medicine) 1998;29:5S-8S.

(7.) Boeniger MF, Lowry LK, Rosenberg J. Interpretation of urine results used to assess chemical exposure with emphasis on creatinine adjustments: a review. Am Ind Hyg Assoc J 1993;54: 615-27.

(8.) Taussky HH. A microcolorimetric determination of creatinine in urine by the Jaffe reaction. J Biol Chem 1954;208:853-61.

(9.) Shaffer PA. The excretion of kreatin and kreatinine in health and disease. Am J Physiol 1908;10:1-10.

(10.) Camara AA, Arn KD, Reimer A, Newburgh LH. The twenty-four hourly endogenous creatinine clearance as a clinical measure of the functional state of the kidneys. J Lab Clin Med 1951;37:74363.

(11.) Vestergaard P, Leverett R. Constancy of urinary creatinine excretion. J Biol Chem 1958;51:211-8.

(12.) Curtis G, Fogel M. Creatinine excretion: diurnal variation and variability of whole and part-day measures. Psychosom Med 1970;32:337-50.

(13.) Cockcroft DW, Gault MH. Predication of creatinine clearance from serum creatinine. Nephron 1976;16:31-41.

(14.) Greenblatt DJ, Ransil BJ, Harmatz JS, Smith TW, Duhme DW. Variability of 24-hour urinary creatinine excretion by normal subjects. J Clin Pharmacol 1976;16:321-8.

(15.) Jacobsen FK, Christensen CK, Mogensen CE, Andreasen F, Heilskov NS. Pronounced increase in serum creatinine concentration after eating cooked meat. Br Med J 1979;1:1049-50.

(16.) Alessio L, Berlin A, Dell'Orto A, Toffoletto F. Reliability of urinary creatinine as a parameter used to adjust values of urinary biological indicators. Int Arch Occup Environ Health 1985;55:99-106.

(17.) James GD, Sealey JE, Alderman M, Ljungman S, Mueller FB, Pecker MS, et al. A longitudinal study of urinary creatinine and creatinine clearance in normal subjects: race, sex, and age differences. Am J Hypertens 1988;1:124-31.

(18.) Kesteloot H, Joossens JV. On the determinants of the creatinine clearance: population study. J Hum Hypertens 1996;10:245-9.

(19.) O'Connor KA, Brindle E, Holman DJ, Klein NA, Soules MR, Campbell KL, et al. Urinary estrone conjugate and pregnanediol 3-glucuronide enzyme immunoassays for population research. Clin Chem 2003;49:1139-48.

(20.) Holman DJ. Total fecundability and fetal loss in rural Bangladesh [PhD Thesis], University Park, PA: Pennsylvania State University, 1996:321pp.

(21.) Goldberger BA, Loewenthal B, Darwin WD, Cone EJ. Intrasubject variation of creatinine and specific gravity measurements in consecutive urine specimens of heroin users. Clin Chem 1995; 41:116-7.

(22.) Berlin A, Alessio L, Sesana G, Dell'Orto A, Ghezzi I. Problems concerning the usefulness of adjustment of urinary cadmium from creatinine and specific gravity. Int Arch Occup Environ Health 1985;55:107-11.

(23.) Haddow JE, Knight GJ, Palomaki GE, Neveux LM, Chilmonczyk BA. Replacing creatinine measurements with specific gravity values to adjust urine cotinine concentrations. Clin Chem 1994;40:562-4.

(24.) Denari JH, Farinati Z, Figueroa Casas PR, Oliva A. Determination of ovarian function using first morning urine steroid assays. Obstet Gynecol 1981;58:5-9.

(25.) Hakim RB, Gray RH, Zacur HA. Is there a need for creatinine adjustment of urinary steroid hormone levels in studies of early fetal loss? Clin Chim Acta 1994;230:209-14.

(26.) Stanczyk FZ, Miyakawa I, Goebelsmann U. Direct radioimmunoassay of urinary estrogen and pregnanediol glucuronides during the menstrual cycle. Am J Obstet Gynecol 1980;137:443-50.

(27.) Thompson SG, Barlow RD, Wald NJ, Van Vunakis H. How should urinary cotinine concentrations be adjusted for urinary creatinine concentration? Clin Chim Acta 1990;187:289-96.

(28.) George JW. The usefulness and limitations of hand-held refractometers in veterinary laboratory medicine: an historical and technical review. Vet Clin Pathol 2001;30:201-10.

(29.) Parikh CR. Screening for microalbuminuria simplified by urine specific gravity. Am J Nephrol 2002;22:315-9.

(30.) Voinescu GC. The relationship between urine osmolality and specific gravity. Am J Med Sci 2002;323:39-42.

(31.) Elkins HB, Pagnotto LD, Smith HL. Concentration adjustments in urinalysis. Am Ind Hyg Assoc J 1974;35:559-65.

(32.) Sauer MV, Paulson RJ. Utility and predictive value of a rapid measurement of urinary pregnanediol glucuronide by enzyme immunoassay in an infertility practice. Fertil Steril 1991;56: 823-6.

(33.) Trevisan A. Concentration adjustment of spot samples in analysis Clinical Chemistry 50, No. 5, 2004 931 of urinary xenobiotic metabolites. Am J Ind Med 1990;17:63742.

(34.) Edwards OM, Whyte HM. Creatinine excretion and body composition. Clin Sci 1959;18:361-6.

(35.) Bleiler RE, Schedl HP. Creatinine excretion: variability and relationship to diet and body size. J Lab Clin Med 1962;59:945-53.

(36.) Carrieri M, Trevisan A, Battista Bartolucci G. Adjustment to concentration-dilution of spot urine samples: correlation between specific gravity and creatinine. Int Arch Occup Environ Health 2001;74:63-7.

(37.) Hall Moran V, Leathard HL, Coley J. Urinary hormone levels during the natural menstrual cycle: the effect of age. J Endocrinol 2001;170:157-64.

(38.) Holman DJ, O'Connor KA. Bangladeshis. In: Ember M, Ember C, eds. Encyclopedia of medical anthropology: health and illness in the world's cultures. New York: Plenum Press, 2003:579-90.

(39.) d'Eril GM, Valenti G, Pastore R, Pankopf S. More on stability of albumin, n-acetylglucosaminidase, and creatinine in urine samples. Clin Chem 1994;40:339-40.

(40.) Riboli E, Haley NJ, De Waard F, Saracci R. Validity of urinary biomarkers of exposure to tobacco smoke following prolonged storage. Int J Epidemiol 1995;24:354-8.

(3) Nonstandard abbreviations: SG, specific gravity; CR, creatinine; E2, estradiol; LH, luteinizing hormone; P4, progesterone; EIA, enzyme immunoassay; PDG, pregnanediol glucuronide; and E1C, estrone conjugate.

Rebecca C. Miller, [1] Eleanor Brindle, [1] Darryl J. Holman, [1] Jane Shofer, [1] Nancy A. Klein, [2] Michael R. Soules, [2] and Kathleen A. O'Connor [1]*

[1] Department of Anthropology and Center for Studies in Demography and Ecology, and [2] Department of Obstetrics and Gynecology, University of Washington, Seattle, WA.

* Address correspondence to this author at: Department of Anthropology, Box 353100, University of Washington, Seattle, WA 98195. Fax 206-543-3285; e-mail
Table 1. Pearson correlations between serum and lagged urinary
hormone values in US women. (a)

 Urine concentrations

 Uncorrected CR-corrected

Lag E1C vs E2 PDG vs P4 E1C vs E2 PDG vs P4

Urine 1 day before 0.47 0.79 0.45 0.85
None 0.72 0.91 0.45 0.85
Urine 1 day after 0.91 0.95 0.91 0.98
Urine 2 days after 0.84 0.91 0.8 0.92
Urine 3 days after 0.61 0.57 0.61 0.53

 Urine concentrations


Lag E1C vs E2 PDG vs P4

Urine 1 day before 0.51 0.83
None 0.76 0.95
Urine 1 day after 0.93 0.98
Urine 2 days after 0.84 0.93
Urine 3 days after 0.61 0.55

(a) For E1C vs E2, n = 34 mean paired urine and serum cycle days from
30 US women; for PDG vs P4, n = 17 mean paired urine and serum cycle
days from 30 US women.

Table 2. Weighted correlations between serum and urinary hormone
values in US women. (a)

 Urine concentration

Urine/Serum Uncorrected SG-corrected CR-corrected

E1C vs E2 0.92 0.93 0.94
PDG vs P4 0.92 0.98 0.97

(a) Correlations incorporate a 1-day lag of urine behind serum. For E1C
vs E2, n = 34 mean paired urine and serum cycle days from 30 US women;
for PDG vs P4, n = 17 mean paired urine and serum cycle days from 30 US

Table 3. Mean differences (95% confidence intervals) between
correlations of serum with urinary hormone values, corrected
by CR or SG and without correction, from US specimens. (a)

 CR-corrected vs SG-corrected vs
 uncorrected uncorrected

E1C vs E2 0.013 (-0.032 to 0.079) 0.019 (-0.007 to 0.058)
PDG vs P4 0.063 (0.011-0.197) (b) 0.056 (0.017-0.179) (b)

 SG-corrected vs CR-corrected

E1C vs E2 0.006 (-0.024 to 0.043)
PDG vs P4 -0.007 (-0.036 to 0.015)

(a) Correlations incorporate a 1-day lag of urine behind serum.
For E1C vs E2, n = 34 mean paired urine and serum cycle days
from 30 US women; for PDG vs P4, n = 17 mean paired urine and
serum cycle days from 30 US women.

(b) Significantly different (P <0.05).
COPYRIGHT 2004 American Association for Clinical Chemistry, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2004 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:General Clinical Chemistry
Author:Miller, Rebecca C.; Brindle, Eleanor; Holman, Darryl J.; Shofer, Jane; Klein, Nancy A.; Soules, Mich
Publication:Clinical Chemistry
Date:May 1, 2004
Previous Article:Plasma fluorescence scanning and fecal porphyrin analysis for the diagnosis of variegate porphyria: precise determination of sensitivity and...
Next Article:False-negative urine protein electrophoresis by semiautomated gel electrophoresis.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters