Men are known to be at greater risk of urolithiasis and cardiovascular and renal diseases than women. Previous studies suggest that greater urine concentration is associated with acceleration of progression of chronic kidney disease (CKD), increased urinary albumin excretion, and delayed renal sodium excretion. The present review addresses possible sex-related differences in urine volume and osmolality (Uosm) that could participate in this male risk predominance. Because of the scarcity of information, we reanalyzed 24-h urine data collected previously by different investigators for other purposes. In nine studies concerning healthy subjects (6 studies) or patients with CKD or diabetes mellitus, Uosm (or another index of urine concentration based on the urine/plasma creatinine concentration ratio) was 21–39% higher (i.e., about a 150 mosm/kgH2O difference) in men than in women. Urine volume was not statistically different. Thus, the larger osmolar load of men (related to their higher food intake) is excreted in a more concentrated urine with no difference in urine volume. This sex difference was not influenced by the level of sodium excretion and was still present in CKD patients. Sex differences in thirst threshold, AVP level, and other regulatory mediators may all contribute to the higher male Uosm. Because of the previously demonstrated adverse effects of vasopressin and/or high urine concentrating activity, the greater tendency of men to concentrate urine could participate in their greater susceptibility to urolithiasis and hypertension and to the faster progression towards end-stage renal failure.
- urine volume
there is a sex-related difference in susceptibility to renal and cardiovascular diseases that includes a higher male prevalence of urolithiasis (24, 46), hypertension, and chronic kidney disease (CKD), as well as a faster rate of progression of CKD (40, 48, 51, 66). A similar sex difference is found in rats (13, 47, 58). Several experimental studies have revealed adverse effects of vasopressin through its V2 receptor-mediated effects and/or the resulting urine concentration on progression of CKD (17, 18), albuminuria of diabetes mellitus (DM) (11), and hypertension (28). It is thus interesting to evaluate whether sex differences in urine concentration could contribute to the greater male susceptibility to vascular and renal diseases. The amount of information in this field is surprisingly small. Urine osmolality (Uosm) is rarely measured in clinical investigations and urine volume, even when used for calculation of 24-h excretions, is rarely reported. This prompted us to gather and reanalyze data obtained for other purposes in previous clinical investigations to characterize possible sex differences in urine concentration that could play a role in the male predominance of lithiasis, hypertension, and progression of CKD.
Direct and Indirect Evaluation of Urine Concentration
Studies A to G concerned healthy subjects of various ages (7, 15, 32, 36, 38, 50, 65) while studies R, S, and T included patients with CKD or DM (4, 5, 36) (see Table 1). In all studies (except study G), 24-h urine was collected in subjects of both sexes that were in steady state on an ad libitum diet and fluid intake. Investigators from the initial studies provided demographic information (see Table 2) along with measurements of 24-h urine volume and sodium and potassium concentrations (UNa and UK, respectively). In some studies, plasma and urine creatinine concentration (Pcreat and Ucreat, respectively), urine urea concentration (Uurea), or urine osmolality (Uosm) were also available.
The 24-h urine volume was measured in all studies (except study G) but Uosm was available only in studies A, B, G, and S. In the others, the level of urine concentration was evaluated in two different ways. When UNa, UK, and Uurea were available (studies C, D, and R), an estimated value of Uosm (eUosm) was calculated according to the formula: eUosm = (UNa+UK)*2 + Uurea. Highly significant and similar correlation coefficients were found between Uosm and eUosm for men and women in studies A+B (combined because they both concern French subjects of similar age) (r = 0.95 and 0.97, respectively), and in study S (0.92 and 0.92, respectively) (P < 0.001 for each). The slopes of the regression lines were about 8–12% lower than unity (with no significant sex difference), which indicates that eUosm understimates slightly true Uosm, likely because the formula for calculation of eUosm does not take into account the minority solutes. In the remaining three studies (E, F, and T), Pcreat and Ucreat allowed an indirect evaluation of the kidney’s tendency to concentrate urine. Because water, but not creatinine, is progressively reabsorbed along the successive nephron segments, tubular fluid creatinine concentration increases above the plasma value. Thus, the ratio Ucreat/Pcreat provides a relative index of urine concentration (UCI). In the two studies in which both Uosm and UCI were available, UCI was linearly and positively correlated with Uosm (r = 0.86 and 0.89 in healthy men and women of study A and 0.82 and 0.82 in CKD patients of study S). The equations of the regression lines did not differ significantly between men and women in either study. Tubular secretion of creatinine could increase UCI and lead to an overestimation of the kidney’s tendency to concentrate urine, but this secretion becomes significant only when plasma creatinine is elevated (39), as in CKD or during infusion of exogenous creatinine (as used in older studies) (31, 42). A higher renal secretion of creatinine in males could induce a bias because testosterone has been shown to stimulate this secretion in rats (31). But, among the large number of more recent investigations that have explored 24-h endogenous creatinine clearance and a more reliable marker of glomerular filtration rate (GFR), none has mentioned a greater discordance between the two variables in males than females (33, 35, 49).
Urine Concentration in Men and Women
The age and body mass index were very close in men and women in all studies (Table 2). Food intake was likely higher in men than in women (men-to-women ratio for 24-h osmolar excretion was > 1). In every case, whether involving healthy subjects or patients with DM or CKD, Uosm, eUosm, or UCI was higher in men than in women by 15–39% (significant except in studies with n < 20 per sex). The male/female ratios obtained with UCI fell in the same range as those obtained with Uosm or eUosm (Table 2). In 5 of 9 studies, 24-h urine volume was almost similar in men and women (within ± 5%), and in the others, it was 10–14% lower in men. To our knowledge, only one prior study reported urine volume and osmolality in men and women separately (21). It shows similar differences as those observed here (Uosm = 678 ± 39 and 493 ± 34 mosm/kgH2O in men and women, respectively, P < 0.05; urine volume = 1.37 ± 0.09 and 1.54 ± 0.09 liters/day, not significant; male-to-female ratio = 1.38 for Uosm and 0.88 for volume). Rats and dogs also show male-to-female ratios of Uosm of similar magnitude as humans (1.16 to 1.24) (27, 34, 44).
It is interesting to note the wide interindividual variability in 24-h urine concentration (Fig. 1) and volume (0.5 to 5.4 liter/24 h in all studies together, not shown) among healthy subjects, probably related to a parallel variability in thirst, fluid intake, and plasma vasopressin (PAVP) (67). Within the 24-h cycle, Uosm is highest at night and it increases after a protein-rich meal (30). However, the sex difference in Uosm does not result from a proportionately higher protein intake in men, because we verified that the proportion of urea in the urine did not differ in the two sexes. Differences in sodium intake are not involved since selective changes in dietary sodium over a relatively wide range do not alter Uosm in either sex, as shown in study D (Table 3). Two-way ANOVA showed a significant sex difference for Uosm and volume (P < 0.002 for each) but no influence of sodium itself and no interaction. Urine concentrating ability declines with age and linear regression of Uosm vs. age in subjects of study D shows a decline (P < 0.0003) by ∼50 mosm/kgH2O per decade, maintaining a significant sex difference across ages (P = 0.005) (Fig. 2A).
Subjects with CKD or DM exhibited a sex difference in Uosm similar to that in healthy subjects (Table 1). Fig. 2B illustrates the decline in Uosm as a function of the degree of renal failure in study S. Two-way ANOVA revealed differences for sex and CKD class (P < 0.0001 for each) with a significant interaction (P = 0.001). Uosm declined progressively in men and more abruptly in women. In the same study, a sample of early morning urine was about 50–100 mosm/kgH2O higher than the 24-h average, even in CKD stages III and IV, showing that the concentrating activity of the kidney is still increasing at night throughout most of the disease progression.
Possible Causes of the Sex Difference in Uosm
It is unlikely that sex hormones directly influence urine concentration since a higher Uosm and similar urine volume were already present in boys compared with girls before puberty, and urine concentration did not vary with age in either sex from 4 to 15 yr (26). Further, the difference in Uosm remained significant in women after the age of menopause in studies D (Fig. 2A) and C (all above 50 yr). Finally, Uosm was not altered 5 wk after gonadectomy in male and female rats (27).
Because men excrete a higher osmolar load through an increase in urine concentration rather than in urine volume, it may be assumed that their thirst/vasopressin system has higher thresholds than those of women, and that they drink proportionally less. However, information is lacking to document these differences. One study reported a higher water intake in female vs. male rats (62). Some studies (1, 22, 54), but not all (21), reported higher values for plasma and/or urinary vasopressin in men than in women, a difference also observed in rats (54). Vasopressin secretion seems to be more sensitive to osmotic stimuli (e.g., hypertonic saline infusion) in males than in females in rats and humans (43, 56).
However, a higher plasma vasopressin is not sufficient to explain the sex difference in urine concentration because study G showed that the difference in Uosm was not abolished during maximal stimulation of the urinary concentrating mechanism, at least in an acute situation. Six healthy subjects (3 of each sex) were infused with a high dose of dDAVP (1-desamino[8-d-arginine vasopressin], a selective V2 receptor agonist of vasopressin) (7, 15). Uosm rose from 835 ± 150 to 955 ± 10 mosm/kgH2O in men and from 540 ± 140 to 775 ± 20 mosm/kgH2O in women. This observation suggests that the higher urine concentration in men than women involves downstream events, probably in the kidney itself, although clearly more data is required. Animal studies support a sex difference in vasopressin actions (55). The male kidney is more sensitive to vasopressin because the antidiuretic response to exogenous hormone was greater in male than female rats (61, 64). In addition, papillary collecting duct cells from male rats exhibit more V2 receptors and a greater vasopressin-induced cAMP accumulation than those from females (64). A higher male sensitivity to vasopressin in humans is also suggested by the higher Uosm observed in men than women who exhibited similar PAVP (21).
Because prostaglandins or a high medullary blood flow are known to interfere with the antidiuretic effect of vasopressin and the ability to concentrate urine (14, 25, 37), known sex differences in the production of prostaglandins (45, 60, 63) and in medullary or papillary blood flow (27) could also play a role in the greater male urine concentration.
Possible Consequences of the Sex Difference in Uosm
The higher tendency of men to concentrate urine compared with women may participate in their higher susceptibility to several diseases or their more severe rate of progression, including urolithiasis, CKD, and some forms of hypertension.
Urolithiasis is 2–3 times more frequent in men than in women (20, 23, 57), but to our knowledge, no study has considered the possibility that a difference in urine concentration could contribute to this sex difference. The higher Uosm in men will obviously favor the occurrence of supersaturation which is responsible for crystallization of poorly soluble compounds (24). In study C, almost half of the men but only 5% of the women had 24-h urine exceeding 600 mosm/kgH2O (Fig. 1A), illustrating the greater risk in men. Even if the concentration of poorly soluble solutes does not reach supersaturation threshold in the pooled 24-h urine, it may well exceed it during transient episodes of higher concentration occurring after protein-rich meals, at night, or during intense physical exercise, especially in summer, a season during which men show a remarkable decrease in urine volume (46).
Vasopressin and/or the resulting rise in urine concentration, influence renal function in different ways. Chronic stimulation of urine concentrating activity by dDAVP in normal rats increases GFR (16) and urinary albumin excretion (10) and induces a hypertrophy of the kidney that resembles that induced by a high protein intake (6, 8). In rats with experimental CKD, detrimental effects of V2 receptor stimulation or beneficial effects of their inhibition were reported on proteinuria, glomerulosclerosis, and tubulointerstitial injury (17, 18, 59). V2 receptor activation also participates in the rise in albuminuria observed in rat models of DM and salt-sensitive hypertension (10–12, 28). Because of the sex difference in Uosm, all of these effects may be more pronounced in males vs. females. The antidiuretic action of vasopressin probably adds its influence to that of the renin-angiotensin system (3, 10, 48, 51, 66) when the intensity of both systems is increased in response to their respective stimuli.
Another potentially adverse effect of vasopressin could result from the direct stimulation of sodium reabsorption in the collecting duct mediated by the activation of the epithelial sodium channel ENaC (41, 52). This effect is most likely responsible for the diminished ability of healthy humans to excrete sodium (7, 19). It is detectable only above a certain threshold of ∼500–600 mosm/kgH2O (2, 7, 9) and thus, should affect males more than females. Interestingly, after an infusion of hypertonic saline (3% NaCl), sodium excretion increased less in men than in women, and only men showed an increase in systolic and pulse pressures, suggesting a hypertensive shift in the pressure-natriuresis curve associated with an increase in extracellular fluid volume (56). These observations suggest that vasopressin could play a role in salt-sensitive hypertension as shown in rats (28).
In summary, the present investigation shows, in several independent studies performed in France and in the United States, that men concentrate urine more than women in free living conditions on their usual diet and spontaneous fluid intake. This difference does not seem to depend on direct effects of sex hormones and is not influenced by the level of sodium intake. It is still present during aging and in CKD. Whether this difference plays a role in the greater prevalence of urolithiasis and hypertension in men and in their faster CKD progression remains to be evaluated. Hopefully, this review will stimulate further studies addressing this issue and including simultaneous measurements of thirst, fluid intake, PAVP, Uosm, and TcH2O in men and women. Newly developed vasopressin V2 receptor antagonists (29, 53) will also represent useful tools for evaluating in humans the possible adverse consequences of the concentrating activity of the kidney, with special attention to sex differences.
NOTE ADDED IN PROOF
An ethnic difference in urine concentration has been recently reported in study E in a separate paper. Black subjects were shown to concentrate urine significantly more than whites (see Bankir L et al. Ethnic differences in urine concentration: possible relationship to blood pressure. Clin J Amer Soc Nephrol. In press). The sex-related difference reported here (Table 2) was of similar magnitude in both ethnic groups. Interestingly, a sex difference in black and in white subjects was also visible in a study of Cowley et al (21).
Part of the results included in this paper have been presented at the 37th Annual Meeting of the American Society of Nephrology in 2004 and published in abstract form (J Am Soc Nephrol 16: 560A, 2004).
We are particularly indebted to the following investigators who kindly provided us with their data and allowed us to use them for the analyses performed in this paper: Aoumeur Hadj-Aïssa, Dépt. de Néphrologie, Hôpital Edouard Herriot, Lyon, France (study A); Marie-Marcelle Trinh-Trang-Tan, Institut National de la Santé et de la Recherche Médicale, Unité 665, INTS, Paris, France (study B); Paul R. Conlin, Dept. of Medecine, Brigham and Women’s Hospital, Boston, MA (study D); Myron H. Weinberger, Dept. of Medicine, Indiana University School of Medicine, Indianapolis, IN (study E); Susie Q. Lew, Department of Medecine, Renal Div., George Washington Univer. Medical Center, Washington D.C. (studies F and T); Daniel G. Bichet, Research Center and Nephrology Service, Hopital du Sacre-Coeur, Montreal, QC, Canada (study G); Bernard Bauduceau, Service d’Endocrinologie, Hôpital Begin, Saint-Mandé, France (study R); Paul Jungers, Service de Néphrologie, Hôpital Necker, Paris, France (study S).
The authors thank Paul R. Conlin (Brigham and Women’s Hospital, Boston, MA) for critical reading of the manuscript and Paul Jungers and Michel Daudon (Hôpital Necker, Paris, France) for valuable discussions.
- Copyright © 2007 the American Physiological Society