Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Effects of Sex on Intra-Individual Variance in Urinary Solutes in Stone-Formers Collected from a Single Clinical Laboratory

Effects of Sex on Intra-Individual Variance in Urinary Solutes in Stone-Formers Collected from a Single Clinical Laboratory

  • Guy M. L. Perry, 
  • Steven J. Scheinman, 
  • John R. Asplin



Our work in a rodent model of urinary calcium suggests genetic and gender effects on increased residual variability in urine chemistries. Based on these findings, we hypothesized that sex would similarly be associated with residual variation in human urine solutes. Sex-related effects on residuals might affect the establishment of physiological baselines and error in medical assays.


We tested the effects of sex on residual variation in urine chemistry by estimating coefficients of variation (CV) for urinary solutes in paired sequential 24-h urines (≤72 hour interval) in 6,758 females and 9,024 males aged 16–80 submitted to a clinical laboratory.


Females had higher CVs than males for urinary phosphorus overall at the False Discovery Rate (P<0.01). There was no effect of sex on CV for calcium (P>0.3). Males had higher CVs for citrate (P<0.01) from ages 16–45 and females higher CVs for citrate (P<0.01) from ages 56–80, suggesting effects of an extant oestral cycle on residual variance.


Our findings indicate the effects of sex on residual variance of the excretion of urinary solutes including phosphorus and citrate; differences in CV by sex might reflect dietary lability, differences in the fidelity of reporting or genetic differentiation in renal solute consistency. Such an effect could complicate medical analysis by the addition of random error to phenotypic assays. Renal analysis might require explicit incorporation of heterogeneity among factorial effects, and for sex in particular.


‘Physiological’ traits [1] such as renal solute excretion frequently have high proportions of residual (error) variance: there is high inter- and intra-individual variation in urinary excretion of phosphate, urate, citrate, sodium, oxalate [2], [3], [4] and albumin [5]. Residual variance in oxalate excretion is higher among calcium oxalate stone formers than non-formers [6]; intra-individual variation (IIV) in 24-h calcium excretion is higher in children with nocturnal enuresis, sometimes associated with hypercalciuria, compared to controls [7], [8]. It has been argued from these phenomena that single 24-hour urine collections are insufficient for the evaluation of risk factors for kidney stones; a sample of 1142 individuals collected from a private urology practice and a stone research clinic found that in 68% of cases paired urine measurements for stone disease were sufficiently different by a value large enough to produce misdiagnosis if diagnosis had been based only on a single measurement [9].

There is an accumulating body of evidence that residual variability in phenotype may be under partial genetic control [10], [11], [12], [13]. There are examples of sexual modification of residual variance or genetic effects on residual variance, including Drosophila morphology [14], environmental resistance in fish [10] and rodent models of behavior [15]. Variation in serum lipids and apolipoprotein concentrations is higher in women over long (>6 month) intervals [16]. Environmental buffering of phenotype in mice also appears to be sex-specific [17]. Using a rodent model of hypercalciuria, the Genetic Hypercalciuric Stone-forming (GHS) rat, we have detected several sex-specific loci associated with absolute-transformed residuals in urinary calcium [13], a model of hypercalciuria and kidney stone disease. Sex-related differences in human urinary calcium and stone formation are well established [18], [19], and there is also substantial residual variance in renal solute excretion [2], [3], [5], [6], [7], [8], [9]. Clarifying sexual biases in residual variance could improve analysis and diagnosis of urinary solute patterns in renal disease.

Given sex and genetic effects on residual variance, we hypothesized that CVs in urinary solute excretion, including urinary calcium, would also differ between men and women in a sample of 15,782 kidney stone-formers (6,758 females and 9,024 males) collected by Litholink Inc. (Chicago, IL), a commercial assay company. A priori predictions of direction (higher CVs in males or females) were not possible since sex differences in residual variation of urinary calcium in our rat model indicated sex-by-strain interaction [13], so that only an effect of sex was predicted rather than direction. Sex-limited effects on random residuals might affect medical analysis and/or diagnosis; intra-individual variance (IIV) in renal traits may be of great diagnostic significance [9].

Materials and Methods

Sample collection

Paired 24-hour urines with a maximum interval of 72 hours were obtained as part of routine clinical testing for 7,372 female and 9,780 male kidney stone patients and submitted to a single reference laboratory (Litholink Corp) for urine chemistry measurement. Urines were collected at room temperature using an antimicrobial preservative. Individual 24-h urines were measured for all subjects for calcium (Ca; mg), citrate (Cit; mg), chloride (Cl; mmoles), creatinine (Cr; mg), potassium (K; mmoles), sodium (Na; mmoles), magnesium (Mg; mg), oxalate (Ox; mg), phosphorus (P; g), ammonium (NH4; mmoles), sulfate (SO4; meq), uric acid (UA; g) and urea nitrogen (UN; mg). Cr was collected for the purposes of measuring completeness of collection, and Na and UN for the purposes of controlling for dietary inputs of salt and protein. Where a measurement was missing for any urinary solute, that individual was removed from all analysis for that urinary trait. Weight (kg), age (yrs) and sex were collected for each individual but no additional clinical or background information was collected by Litholink Inc. Individuals known to be receiving treatment for stones at the time of collection were excluded from the study, although not all patients had a fully-known treatment history. Use of this data set was approved by the Western Institutional Review Board (Olympia, WA). Records were collected over the history of Litholink's operation. No experimentation was performed.

Differences in univariate coefficients of variance (CV) by sex

Coefficients of variation (CV) for paired 24-h measurements of urinary excretion rates were estimated for all individuals as . CVs for each trait were ranked in ascending order x1xn for independent analysis Ranked CVs were used to avoid complications from non-normal distributions in this nonparametric statistic while still permitting multivariate models to account for the effects of significant covariates. Global effects on CV for each urinary solute were estimated simultaneously with covariate selection using multivariate backwards regression in SAS [20] with elimination of nonsignificant variables at P <0.1, beginning with the starting model (Model 1):where was the coefficient of variation ranking for urinary variable y, was the model intercept, are the coefficient of () and incident vector for () the effect of sex on CVy, is the effect of weight (kg) on CVy, is the effect of higher-order CV-weight interactions on CVy, is the linear effect of age on CVy and the effect of higher-order curvature age-CVy relationships to account for changes in CV within or across life history categories, was the effect of ranked CV for differences in urinary volume between the two samples, was the effect of ranked on CVy, was the effect of ranked on CVy, was the effect of ranked and was error. Colinearity among the covariates (, and ) was tested in PROC REG prior to further analysis to prevent biased control of dietary and collection effects on urinary output by effective duplication of closely connected effects. Where variance inflation functions (VIF) exceeded 10.0 for a pair of linear and quadratic terms for the same covariate, the quadratic term was excluded from the model. The final complement of regression terms included via backwards regression normally ranged from three to five (Table S1). Multiple regression for linear covariates was preferred to the establishment of narrow factorial categories in order to reduce degrees of freedom lost on subcategorizations [21]; our single experiment involving age categories was based on explicit life-history contrasts (see below). Sex was coded as a binary variable with female indicated as 0 and males as 1, so that positive, significant values for indicated high ranked variance for males and negative values high ranked variance in females. All individuals with scores above the upper 30th percentile were excluded prior to analysis as possibly irreparably biased due to errors of collection. Ranked was included to account for collection variance for all traits in the remaining individuals, and and were included as covariates to account for dietary protein and salt on those traits [22]. Mean CV for each trait by sex was estimated using a least-squares (LS) model based on terms retained from backward stepwise regression, above, with sex coded as a fixed effect and all other terms as partial regressions [20].

The significance of sex effects on CVs were evaluated using the false discovery rate (FDR) [23]. Nominal P-values (Pnom) for sex effects on CV were ranked in ascending order from 1 to m (the total number of traits). Where the ranked Pi for hypothesis Hi for each trait i satisfied the relation Piαi/m, where α was the nominal a priori significance threshold ( = 0.05), we rejected the null hypothesis of no significant difference in CV between males and females for that test and all tests with lower Pi [23].

Sex differences in urinary CV by major age categories in female life-history

In order to test the effects of post-menarche vs. post-menopausal female physiology on CVs for urinary traits, we divided the population into two classes: 1) a rough ‘menarche-menopause’ (MM) category of 16–45 years and 2) a rough ‘post-menopausal’ (PM) group of 56–80 years. These age categories were selected to coincide with the approximate mean age of inception of menopause in Western women [24] with a ten-year interval window to limit overlap of approximate menstrual- and menopausal-age individuals outside their respective categories. The effects of sex and other factors on urinary CVs was then tested using backwards regression (Pthr<0.1) in Model 1 as above, within each of categories 1) and 2). Significance thresholds were estimated separately within each group at the FDR as above.

We next tested for interactive effects of age-group with sex on residual variance again using backwards stepwise regression for joint covariate selection and analysis, starting with the saturated general linear model (Model 2)where was the coefficient of variation ranking for urinary variable y, was the model intercept, is the effect of sex i on the CV for trait y, is the effect of age group j, is the effect of interaction between sex i and age group j on CVy and all other terms were as per Model 1 (above). As above, from three to five terms were normally included in the final analysis via backwards regression (Table S1). The significance of sex×age class interaction was corrected using the FDR [23]. Covariates were removed as required by VIF conditioning as in Model 1. Least-squares estimates of means and standard errors by sex and age-class were determined in a general linear model in PROC GLM [20] to estimate mean CV ranks while avoiding contamination from other dependent variables.


Means and CVs

After excluding individuals in the upper 30th percentile of , a total of 9,024 males and 6,758 females remained for analysis. CVs were highest overall for Na and Cl, and lowest for Cr although the range of individual CVs was highest for Ca (Table 1).

Table 1. Mean coefficients of variation () and minimum and maximum CVs for urinary (Ca; mg), citrate (Cit; mg), chloride (Cl; mmoles), creatinine (Cr; mg), potassium (K; mmoles), sodium (Na; mmoles), magnesium (Mg; mg), oxalate (Ox; mg), phosphorus (P; g), ammonium (NH4; mmoles), sulfate (SO4; meq), uric acid (UA; g) and urea nitrogen (UN; mg) within a complete population of 9,024 male and 6,758 female kidney stone patients aged 16–80, and in individuals aged 16–45 and 56–80.

Differences in univariate variability by sex

Females had significantly higher than males across the entire population (Model 1; P<0.01) at the FDR (Table 2; Figure 1). Females had higher CVs for Mg and Na, and males for Cl and Ox at nominal (α0.05) a priori thresholds, which were not significant after FDR correction (Table 2; Figure 1). Female CV was typically from 8–10% higher than in males (). We detected no evidence of sex effects on using Model 1 at the FDR threshold or at nominal a priori levels of significance (P>0.3).

Figure 1. Mean ranked coefficients of variation (CV) for chloride (Cl), magnesium (Mg), sodium (Na), oxalate (Ox) and phosphorus (P) in a complete population of 6,758 females and 9,024 males submitted to Litholink, Inc.

significant at nominal (Cl, Mg, Na and Ox) and α0.05 False Discovery Rate thresholds selected using backwards stepwise regression (Pacceptance = 0.1).

Table 2. Effects of gender on ranked coefficients of variation (CV) from paired 24-h urinary measurements of urinary ammonium (NH3), calcium (Ca), chloride (Cl), citrate (Cit), potassium (K), magnesium (Mg), sodium (Na), oxalate (Ox), phosphorus (P), sulfate (SO4), uric acid (UA) and urine urea nitrogen (UN) adjusted for creatinine, urinary volume, age and weight in the complete population of 6,758 females and 9,024 males aged 16–80, and in individuals aged 16–45 and 56–80 in a stepwise backward regression (Pacceptance≤0.1).

Changes in CVs by major life-history category

A total of 4,128 males and 4,219 females from ages 16–45 and 3,578 male and 1,772 females were available. We detected marked changes in the pattern of male and female CVs in in the subdivided Model 1 for the examination of sex effects on variation at ages 16–45 and 56–80. Males had significantly higher from the ages of 16–45 years compared to females (P<0.01) and females significantly higher than males from 56–80 years at the FDR (P<0.01) (Table 2; Figure 2). From 16–45 years, males had significantly higher (P<0.01) at the nominal significance threshold but not after multiple correction.

Figure 2. Mean ranked coefficients of variation (CV) for citrate (Cit), chloride (Cl), potassium (K), magnesium (Mg) and phosphorus (P) in 4097 females (F) and 4014 males (M) in an approximate ‘menarche-menstrual’ (MM) period (ages 16–45) and in 1715 females and 3479 males in an approximate ‘postmenopausal’ group (PM) (ages 56–80).

Significance of differences in CV by trait and age group (MM vs. PM) indicated as P<0.05*, P<0.01**.

Reminiscent of Model 1, females had higher from 56–80 years and marginally higher from 16–45 at the nominal significance threshold. Females also had nominally higher and than males in the 56–80 year age group while was nominally higher in males. None of these associations was significant after multiple correction (PFDR>0.05); was marginally higher in females at the FDR in the 56–80 age category (P<0.1) (Table 2; Figure 2).

Using partial regression covariates identified in Model 1, we detected significant effects of sex×age class interaction on at the FDR threshold (P<0.001) between the two age groups (16–45 years and 56–80 years). There were no significant sex×age group effect for any other urinary metabolite after multiple correction (PFDR>0.1) (Table 3).

Table 3. Effects of sex-by-age category (16–45 years vs. 56–80 years) interaction on ranked coefficients of variation (CV = σ/μ) calculated from original paired 24-h urinary solutes (urinary ammonium (NH3), calcium (Ca), chloride (Cl), citrate (Cit), potassium (K), magnesium (Mg), sodium (Na), oxalate (Ox), phosphorus (P), sulfate (SO4), uric acid (UA), urine urea nitrogen (UN)) collected from 6,758 female and 9,024 male stone-formers aged 16–80 at the False Discovery Rate [Verhoeven, 2005 #22836].


Females had higher coefficients of variation for urinary phosphorus over the entire age distribution and for citrate from 56–80 years, while males had higher CV for citrate from 16–45 years of age. We did not detect male-female differences for CVs in urinary calcium despite predictions from our rodent model [13]. Other studies have found diurnal [25], inter- and/or intra-individual variation in urine solutes including calcium and oxalate [2], [3], [4], [5], [7], [8], [26] exhibit high variability, including in stone-formers [6], but sex effects on residual variation in urinary solutes are a novel finding.

Since our study did not employ dietary controls or journaling, our findings might result from day-to-day male-female dietary variability, male∶female genetic or endocrine differences, or differences in the fidelity of urine collections; sex-by-strain and sex-by-genotype effects in our rodent model were detected on controlled, constant diets [13]. Sulfate is encountered in many food additives [27] and while we included urinary nitrogen as a dietary control variable for protein intake (see [22]) there are numerous dietary sources of phosphorus such as dark sodas and preserved fast foods [28] besides meat and dairy products [29] which might create day-to-day variability in urinary phosphorus. Our results may reflect changing output from transient dietary preference, largely in females, or integral differences in residual variance between the sexes in urinary solutes. We cannot be certain, therefore, that human residual variance in urinary solutes represents the same phenomenon we document in our rodent model. No ethnic variables were available; our findings thus may represent a ‘generalized’ trend within stone-formers.

Whether resulting from dietary variability or endogenous endocrine or genetic effects, IIV in urinary solutes may have significant diagnostic consequences: Parks et al [9] found that single measurements were insufficient for the analysis of urinary risk factors for kidney stones, with daily assays varying by a standard deviation (σ) or more. Our rodent work suggests that IIV can bias estimation of ‘classical’ genetic effects on means: a region of chromosome 1 associated with CV in urinary calcium in our rodent model (‘hypercalciuria coefficient of variation 1’) obscured the detection of a nearby locus with effects on mean urinary calcium (‘hypercalcuria 1’) [30]. We have also detected sex-by-genotype and sex-by-strain effects on residual variance in urinary calcium using genetic mapping in rodents, suggesting sex linkage of these effects [13] and other results of ours indicate a heritability of 7% for in humans and 60% in rats (Perry et al unpubs). Sex might similarly affect the measurement of effects on renal solutes, whether as a heritable effect or a general male-female difference in dietary consistency. Differences in reporting accuracy might be excluded as a possible cause for our findings, since sex effects on CV were only detected for two urinary solutes. Notably, females did have significantly higher CVs for urinary phosphorus. Phosphorus and calcium are physiologically related [31]; dietary phosphorus affects urinary calcium and calcium balance [32], [33]. Phosphorus was not assayed in our rodent work, but it is possible that sex-based differences in urinary calcium might be dependent on countercurrent IIV in urinary phosphorus.

Male CVs for urinary citrate were higher than in females in the menarche-menopause group, and females higher at post-menopausal ages. Both female humans and rats excrete more citrate than males overall [34], [35], [36], which may confer some protection against nephrolithiasis in females [37]. Assuming that our sample was representative within our defined age categories, our results might indicate periodic endogenous or dietary changes in relative citrate protection between the sexes depending on age. Female stone incidence is higher than in males at greater ages in many populations [38].


Exogenous and endogenous sex hormones have major main effects on renal physiology [39]; estrogen decreases urinary oxalate, urinary calcium and crystal deposition while testosterone promotes stone formation by suppressing renal osteopontin and increasing urinary oxalate [40]. Men experience more stones and produce more urinary phosphate, calcium, uric acid, magnesium, oxalate and creatinine and less citrate than women of the same age [39], [41], [42]. Our findings indicate that males and females also have differences in IIV for citrate and phosphorus with changing IIV for citrate by age and sex age, either as a function of dietary differences or from endogenous effects associated with sex. These differences might affect the reliability of individual phenotype in genetic mapping assays, alter estimates of basal physiology, or even cause transient declines in protective metabolites.

Supporting Information

Table S1.

Final models for partial regression terms in stepwise backward regression (Pacceptance≤0.1) of ranked coefficients of variation (CV) for 24-h urinary ammonium (NH3), calcium (Ca), chloride (Cl), citrate (Cit), potassium (K), magnesium (Mg), sodium (Na), oxalate (Ox), phosphorus (P), sulfate (SO4), uric acid (UA) and urine urea nitrogen (UN) in 9,024 male and 6,758 female stone-formers aged 16–80, and in age categories 16–45 years and 56–80 years. Partial regression terms were removed at nominal P-value (Pnom)>0.1.



We acknowledge the advice of Dr. Stephen V. Faraone (Department of Psychiatry, SUNY Upstate Medical University), Dr. Sriram S. Narsipur (Department of Medicine, SUNY Upstate Medical University) and Dr. Steven J. Knohl (Department of Medicine, SUNY Upstate Medical University) in the writing of this manuscript and in technical suggestions underlying our analysis. Dr. John Asplin is the Stone Medical Director at Litholink, Inc (Chicago, IL).

Author Contributions

Conceived and designed the experiments: GMLP JRA. Performed the experiments: JRA. Analyzed the data: GMLP. Contributed reagents/materials/analysis tools: JRA. Wrote the paper: GMLP SJS JRA.


  1. 1. Roff D (1997) Evolutionary quantitative genetics. Toronto, Canada: International Thompson Publishing. 493 p.
  2. 2. Ryall R, Marshall V (1990) Idiopathic urolithiasis. In: Wickham J, Buck A, editors. Renal tract Stone: metabolic basis and clinical management. 1st ed. Churchill Livingstone: Avon. pp. 307–331.
  3. 3. Anandaram P, De Bolla A, Hudson P, Davies G, Majumdar P, et al. (2006) Problems in the metabolic evaluation of renal stone disease: audit of intra-individual variation in urine metabolites. Urological Research 34: 249–254.
  4. 4. Marengo SR, Romani AM (2008) Oxalate in renal stone disease: the terminal metabolite that just won't go away. Nat Clin Pract Nephrol 4: 368–377.
  5. 5. Jerums G, Allen T, Gilbert R, Hammond J, Cooper M, et al. (1995) Natural history of early diabetic nephropathy: what are the effects of therapeutic intervention? Journal of Diabetes and Its Complications 9: 308–314.
  6. 6. Holmes R, Assimos D, Goodman H (1998) Genetic and dietary influences on urinary oxalate excretion. Urological Research 26: 195–200.
  7. 7. Valenti G, Laera A, Gouraud S, Pace G, Aceto G, et al. (2002) Low-calcium diet in hypercalciuric enuretic children restores AQP2 excretion and improves clinical symptoms. American Journal of Physiology: Renal Physiology 283: F895–F903.
  8. 8. Kamperis K, Hagstroem S, Rittig S, Djurhuis J (2006) Urinary calcium excretion in healthy children and children with primary monosymptomatic nocturnal enuresis. Journal of Urology 176: 770–773.
  9. 9. Parks J, Goldfisher E, Asplin J, Coe F (2002) A single 24-hour urine collection is inadequate for the medical evaluation of nephrolithiasis. Journal of Urology 167: 1607–1612.
  10. 10. Perry G, Ferguson M, Danzmann R (2003) Effects of genetic sex and genomic background on epistasis in rainbow trout (Oncorhynchus mykiss). Genetica 119: 35–50.
  11. 11. Hill WG (2010) Genetic analysis of environmental variation. Genetical Research (Cambridge) 92: 381–395.
  12. 12. Ibáñez-Escriche N, Moreno A, Nieto B, Piqueras P, Salgado C, et al. (2008) Genetic parameters related to environmental variability of weight traits in a selection experiment for weight gain in mice; signs of correlated canalised response. Genet Sel Evol 40: 279–293.
  13. 13. Perry G, Bushinsky D, Reid R, Scheinman S (2012) Sex modifies genetic effects on residual variance in urinary calcium excretion in rat (Rattus norvegicus). Genetics 192: 1003–1013.
  14. 14. Mackay T, Lyman R (2005) Drosophila bristles and the nature of quantitative genetic variation. Philosophical Transactions of the Royal Society B Biological Sciences 360: 1513–1527.
  15. 15. Perry G, Sagvolden T, Faraone S (2010) Intraindividual variability (IIV) in a rat model of ADHD - the Spontaneously Hypertensive Rat. Behavioral and Brain Functions 6: 56.
  16. 16. Reilly S, Ferrell R, Kottke B, Kamboh M, Sing C (1991) The gender-specific apolipoprotein E genotype influence on the distribution of lipids and apolipoproteins in the population of Rochester, MN. I. Pleiotropic effects on means and variances. American Journal of Human Genetics 49: 1155–1166.
  17. 17. Fraser HB, Schadt EE (2010) The quantitative genetics of phenotypic robustness. PLoS ONE 5: e8635.
  18. 18. Coe F (1988) Prevention and treatment of kidney stones. Bethesda, MD: US Public Health Service.
  19. 19. Curhan G, Willett W, Rimm E, Stampfer M (1997) Family history and risk of kidney stones. Journal of the American Society of Neprology 8: 1568–1573.
  20. 20. SAS (2000). 8.1 ed. Cary, NC: SAS Institute.
  21. 21. Steel R, Torrie J (1980) Principles and Procedures of Statistics. New York: McGraw-Hill Book Co.
  22. 22. Trinchieri A, Mandressi A, Luongo P, Longo G, Pisani E (1991) The influence of diet on urinary risk factors for stones in healthy subjects and idiopathic renal calcium stone formers. Br J Urol 67: 230–236.
  23. 23. Verhoeven K, Simonsen K, McIntyre L (2005) Implementing false discovery control: increasing your power. Oikos 108: 643–647.
  24. 24. Kato I, Toniolo P, Akhmedkhanov A, Koenig K, Shore R, et al. (1998) Prospective study of factors influencing the onset of natural menopause. Journal of Clinical Epidemiology 51: 1271–1276.
  25. 25. Ahlstrand C, Larsson L, Tiselius H (1984) Variations in urine composition during the day in patients with calcium oxalate stone disease. Journal of Urology 131: 77–81.
  26. 26. Fenton T, Eliasziw M, Lyon A, Tough S, Brown J, et al. (2009) Low 5-year stability of within-patient ion excretion and urine pH in fasting-morning-urine specimens. Nutritional Research 29: 320–326.
  27. 27. Magee EA, Curno R, Edmond LM, Cummings JH (2004) Contribution of dietary protein and inorganic sulfur to urinary sulfate: toward a biomarker of inorganic sulfur intake. Am J Clin Nutr 80: 137–142.
  28. 28. Shutto Y, Shimada M, Kitajima M, Yamabe H, Razzaque M (2011) Lack of awareness among future medical professionals about the risk of consuming hidden phosphate-containing processed food and drinks. PLoS ONE 6: e29105.
  29. 29. Houterman S, van Faassen A, Ocke M, Habets L, van Dieijen-Visser M, et al. (1997) Is urinary sulfate a biomarker for the intake of animal protein and meat? Cancer Letters 114: 295–296.
  30. 30. Perry G, Nehrke K, Bushinsky D, Nandi J, Lewandowski K, et al.. (2011) Heritable residual variation for renal physiology? A QTL for the coefficient of variation in urinary calcium. In: Cuppen C, Hubner N, Kwitek A, Gould M, editors; Dec 7–10 2011; Cold Spring Harbor, NY. Cold Spring Harbor Laboratory. pp. 97.
  31. 31. Liu ES, Carpenter TO, Gundberg CM, Simpson CA, Insogna KL (2011) Calcitonin administration in X-linked hypophosphatemia. N Engl J Med 364: 1678–1680.
  32. 32. Hegsted M, Schuette S, Zemel M, Linkswiler H (1981) Urinary calcium and calcium balance in young men as affected by level of protein and phosphorus intake. Journal of Nutrition 111: 553–663.
  33. 33. Yuen D, Draper H (1983) Long-term effects of excess protein and phosphorus on bone homeostasis in adult mice. Journal of Nutrition 113: 1374–1380.
  34. 34. Horl W, Echsel E, Hohenegger M (1985) The key role of sex dependency on kidney citrate metabolism in the rat. Research in Experimental Medicine (Berlin) 185: 69–75.
  35. 35. Rudman D, Kutner M, Redd S, Waters W, Gerron G, et al. (1982) Hypocitraturia in calcium nephrolithiasis. Journal of Clinical Endocrinology and Metabolism 55: 1052–1057.
  36. 36. Parks J, Coe F (1986) A urinary calcium-citrate index for the evaluation of nephrolithiasis. Kidney International 30: 85–90.
  37. 37. Hamm LL, Hering-Smith KS (2002) Pathophysiology of hypocitraturic nephrolithiasis. Endocrinol Metab Clin North Am 31: 885–893, viii.
  38. 38. Romero V, Akpinar H, Assimos D (2010) Kidney stones: a global picture of prevalence, incidence, and associated risk factors. Reviews in Urology 12: e86–e96.
  39. 39. Psihogios NG, Gazi IF, Elisaf MS, Seferiadis KI, Bairaktari ET (2008) Gender-related and age-related urinalysis of healthy subjects by NMR-based metabonomics. NMR Biomed 21: 195–207.
  40. 40. Sims R, Rehrer N, Bell M, Cotter J (2008) Endogenous and exogenous female sex hormones and renal electrolyte handling: effects of an acute sodium load on plasma volume at rest. Journal of Applied Physiology 105: 121–127.
  41. 41. Trinchieri A, Mandressi A, Luongo P, Rovera F, Longo G (1992) Urinary excretion of citrate, glycosaminoglycans, magnesium and zinc in relation to age and sex in normal subjects and in patients who form calcium stones. Scand J Urol Nephrol 26: 379–386.
  42. 42. Lancina Martin JA, Rodriguez-Rivera Garcia J, Novas Castro S, Rodriguez Gomez I, Fernandez Rosado E, et al. (2002) [Metabolic risk factors in calcium urolithiasis according to gender and age of the patients]. Actas Urol Esp 26: 111–120.