Metabolic Alkalemia in Hypercalciuria Stone Formers: Does It Matter?

Most nephrolithiasis cases are due to idiopathic hypercalciuria (∼50%), whose oxalate and phosphate calcium salts may precipitate in Randall’s plaque [1, 2]. Therefore, reducing urinary calcium excretion is a strategy to prevent kidney stone formation in these cases. As thiazide increases renal calcium reabsorption [3, 4], this drug has been indicated as the pharmacological choice to prevent nephrolithiasis crises [5‒7]. However, this therapeutic approach may change in the next years due to the recent results reported by the NOSTONE trial. This double-blind trial did not show thiazide efficiency at different doses to prevent the recurrence of kidney stones when compared to the placebo group [8], though some methodological limitations compromise this conclusion. For example, the study did not stratify patients based on pre-existing hypercalciuria, which is a significant factor in determining the treatment’s response. This lack of patient stratification could have obscured potential benefits in subgroups with specific urinary calcium profiles.

It is well known that thiazide inhibits Na-Cl reabsorption in the distal nephron segment, leading to significant losses of these ions in the urine. In consequence, metabolic alkalemia may occur due to the depletion of chloride in the body [9, 10]. Additionally, thiazide may cause other side effects, like hypokalemia and hypomagnesemia due to the increment of these ion losses in the urine [1‒4]. Furthermore, metabolic disorders detected by increases in uric acid, glucose, and triglycerides in the serum may also occur, as a result of chronic use of thiazides [11‒13].

Besides thiazide prescription, some patients with renal calcium stones also receive potassium citrate supplementation to prevent their recurrence of kidney stones. Since citrate is an alkali organic anion, the urine pH rises [2, 14, 15], which can facilitate calcium reabsorption in the distal nephron via alkaline-sensitive TRPV5 channels [16]. Another important point concerns the fact that citrate binds calcium forming a soluble complex in the urine [17]. Besides, citrate can also reduce the excretion of calcium in the urine because a smaller amount of it is filtered by the kidneys due to less bone turnover and, therefore, a smaller amount of quantity is absorbed in the small intestine [18]. On the other hand, the use of thiazides as a powerful approach to prevent kidney stone recurrence can decrease citrate excretion, as seen in the NONSTONE trial, thus resulting in no evident differences among patients receiving hydrochlorothiazide (HCTZ) and those receiving placebo, regarding relative supersaturation ratios and hypercalciuria levels.

For the reasons described above, we would like to verify how often metabolic alkalemia may occur in hypercalciuric nephrolithiasis outpatients. Moreover, we also aimed to certify whether the prescription of thiazide supplemented or not with potassium citrate was prevalent in these patients. Additionally, we would like to check the clinical characteristics, electrolyte homeostasis, and metabolic parameters of idiopathic hypercalciuric patients who did not develop metabolic alkalemia.

We reviewed the electronic medical records of hypercalciuric nephrolithiasis outpatients aged 18 years old or older followed up from January 2013 to July 2021 in our institution. We defined as hypercalciuric patients in whose charts we found the values of urinary calcium excretion above 200 mg/day [19] at least one time during the study period.

We diagnosed the patients with metabolic alkalemia when, in their medical records, the venous blood gas tests had shown pH ≥7.46 and bicarbonate concentration >26 mEq/L. If we diagnosed metabolic alkalemia two or more times, we would choose the results of laboratory data in which the value of bicarbonate concentration was the highest.

We compared the data collected from the electronic medical records of patients with and without metabolic alkalemia. For a better analysis, we divided the patients into six groups: (i) thiazide with metabolic alkalemia, patients only treated with thiazide that developed metabolic alkalemia; (ii) thiazide plus citrate with metabolic alkalemia, patients treated with thiazide and supplemented with potassium citrate that developed metabolic alkalemia; (iii) thiazide without metabolic alkalemia, patients only treated with thiazide that did not develop metabolic alkalemia; (iv) thiazide plus citrate without metabolic alkalemia, patients treated with thiazide and supplemented with potassium citrate that did not develop metabolic alkalemia; (v) no thiazide and no citrate without metabolic alkalemia, patients without thiazide and citrate in their treatment that did not develop metabolic alkalemia; (vi) citrate without metabolic alkalemia; patients only treated with citrate that did not develop metabolic alkalemia.

We considered the patients without metabolic alkalemia as control groups. So, we only enrolled patients with idiopathic hypercalciuria in these groups and we established 40 months as the time to collect their data because this was the mean duration time of the thiazide treatment in patients who had developed metabolic alkalemia. As the patients were routinely followed up every 8–12 months, we excluded the patients’ records without three medical appointments because the mean treatment length was 40 months. So, we could verify in three successive followed-up medical appointments whether the patients had been really following or not the treatment prescribed.

When chlortalidone (CTD) was the thiazide-type used, we multiplied its dose by factor 2 to calculate the amount of thiazide taken per day because CTD is considered twice as potent as HCTZ [13, 20]. We analyzed clinical and demographic characteristics, as well as medication and available laboratory data. We estimated glomerular filtration rate (eGFR) using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation [21] and we stratified renal functions using the chronic kidney disease nomenclature determined by KDIGO [22].

We expressed data as mean ± SE, and we used the Kolmogorov-Smirnov test for normality verification. We applied the unpaired Student’s t test or the Mann-Whitney test for comparisons between two groups when normality or non-normality had been indicated, respectively. We verified the differences among groups by employing one-way analysis of variance (ANOVA) and Tukey’s multiple comparisons as a post hoc test when the Kolmogorov-Smirnov test indicated normality. In the cases of non-normality, we employed the Kruskal-Wallis test followed by Dunn’s multiple comparisons as a post hoc test. We also evaluated whether continuous variables showed correlations between distinct categories using Pearson’s correlation analysis and non-parametric Spearman correlation when normality or non-normality had been indicated, respectively. The statistical significance was set at p ≤ 0.05 and we performed all statistical analysis by using the Graph Pad Prism 8.0 (La Jolla, CA).

We analyzed 395 electronic medical records of hypercalciuric nephrolithiasis outpatients treated in our institution. We diagnosed metabolic alkalemia in only 17 (4.3%) cases. In all of them, we found thiazide in their prescription, except for 1 patient who received furosemide treatment due to chronic renal disease, as illustrated in Figure 1. We also verified that the majority of patients were women either in groups with and without metabolic alkalemia (Table 1).

Besides metabolic alkalemia, we have also verified lower potassium, chloride, ionized calcium, and magnesium serum concentration values in these patients than in those without a disorder in their acid-base equilibrium. Furthermore, patients with metabolic alkalemia had higher urine pH and lower urinary citrate excretion than the ones without it (Table 1).

As illustrated in Figure 1, the great majority of hypercalciuric nephrolithiasis outpatients of our institution enrolled in this study had thiazide in their prescription (84%). We verified that thiazide was only prescribed to prevent nephrolithiasis crises in them. We did not find any case in which thiazide had been used as an antihypertensive drug.

With regard to age and blood pressure, the patients without metabolic alkalemia treated only with citrate (Group VI) were younger and had lower mean blood pressure than those only treated with thiazide (Group III). Furthermore, the mean blood pressure in patients only treated with citrate (Group VI) was also lower than in patients treated with thiazide plus citrate who developed metabolic alkalemia (Group II), as demonstrated in Figure 2. We also observed that the patients only treated with citrate (Group VI) were those using less antihypertensive medication and they were also younger than the other ones (Tables 2, 3).

We detected mild metabolic alkalemia either in patients supplemented or not with potassium citrate, except for one case in Group I and another one in Group II (Fig. 3a, b). Even though many patients in Groups III and IV showed increases in bicarbonate concentration, these changes were probably secondary to the increase in venous blood pCO₂ because their blood pH was in the normal range or below 7.35 (Fig. 3a–c). Thus, these patients exhibited metabolic alkalosis but not metabolic alkalemia.

Since the patients supplemented or not with potassium citrate had similar characteristics regarding their clinical and demographic situation (Table 2), we decided to cluster Groups I and II and Groups III and IV. Thus, we increased the sample of patients who had developed metabolic alkalemia, allowing a better statistical analysis, and we compared their data to the control Group, formed by assembling Groups III and IV.

Following the analysis, we investigated whether the amount of thiazide taken per day could be distinct between patients with and without metabolic alkalemia. In fact, the daily thiazide intake was significantly higher in patients with metabolic alkalemia than in those without it, as shown in Table 1. However, these statistical differences disappeared when we restricted the analysis to thiazide-treated patients. The values were 47.66 ± 6.06 and 40.45 ± 1.71 mg of thiazide taken per day in patients who developed or not metabolic alkalemia, respectively (p < 0.18 determined by the Mann-Whitney test). Moreover, we did not find any correlation between the amount of thiazide taken per day and the eGFR in thiazide-treated hypercalciuric nephrolithiasis outpatients who developed metabolic alkalemia, but the eGFR in these patients decreased significantly with age as often occurs in the general population (Table 4).

When we checked the electrolyte’s homeostasis in thiazide-treated hypercalciuric nephrolithiasis outpatients, we verified lower values of serum potassium concentration in patients with metabolic alkalemia than in those without it, p < 0.0001 (Fig. 4a). We diagnosed hypokalemia in 37.5% of patients with metabolic alkalemia (6 patients out of 16), and in 1.2% of patients without metabolic alkalemia (2 patients out of 161). We also observed that CTD was the thiazide-type used in almost all the hypokalemic cases, as illustrated in Figure 4a. However, we did not verify any correlation either between serum potassium and the amount of thiazide taken daily (Fig. 4b; Table 4) or between serum potassium and blood pH (Fig. 4c; Table 4) in the patients who developed metabolic alkalemia.

Regarding other electrolytes’ homeostasis in thiazide-treated hypercalciuric nephrolithiasis outpatients, we detected that the concentrations of chloride, magnesium and ionized calcium were significantly lower in the serum of these patients with metabolic alkalemia than in those without it (Fig. 5a–c). As demonstrated in Table 4, we did not find any correlation between these S electrolytes’ concentration and the amount of thiazide taken daily. However, we verified a positive correlation between the blood pH values and S chloride and S ionized calcium. On the other hand, we did not detect any correlation between S bicarbonate and S chloride or S ionized calcium. We also would like to remark that we found S chloride registered in only 50% of electronic medical records of thiazide-treated hypercalciuric nephrolithiasis outpatients who developed metabolic alkalemia, and only 4 patients out of 8 of these cases had S chloride values below 98 mEq/L. Regarding S ionized calcium, we tried to find correlations between this parameter and distinct categories that could control calcium homeostasis, like the serum concentrations of parathormone and 25-hydroxyvitamin D, but we did not find any (Table 4).

We have really evaluated hypercalciuric nephrolithiasis outpatients who responded well to the thiazide treatment. As shown in Table 2, the urinary calcium excretion values have been reduced to under 200 mg per day at data-collecting time in all groups. The variation in their urinary calcium excretion expressed in mg per day was 96.60 ± 46.69 in Group I, 198.00 ± 30.07 in Group II, 106.00 ± 11.84 in Group III, 116.20 ± 9.95 in Group IV, p < 0.19, determined by the Kruskal-Wallis test. On the other hand, we did not find any increment in urinary citrate excretion values in the hypercalciuric nephrolithiasis outpatients who received potassium citrate supplementation either in those that developed or not metabolic alkalemia (Table 2).

We also applied tests to verify whether urinary calcium excretion in thiazide-treated hypercalciuric nephrolithiasis outpatients who developed metabolic alkalemia would be correlated with the amount of both thiazide and vitamin D taken per day or other parameters. As shown in Table 4, we did not detect any significant correlation. With regard to the urine output and the specific urine gravity, all groups showed similarities in their values, as demonstrated in Tables 1 and 2.

Nevertheless, we noted that the urine pH values were higher in the hypercalciuric nephrolithiasis outpatients with metabolic alkalemia than in those without an acid-base disorder (Table 1). As illustrated in Figure 6a, urine pH values remained higher in patients with metabolic alkalemia when compared to patients without it, even when we restricted the analysis to thiazide-treated hypercalciuric nephrolithiasis outpatients. Furthermore, we found a positive correlation between urine pH and urinary citrate excretion in these patients who developed metabolic alkalemia (Fig. 6b; Table 4).

It is well known that thiazides can induce metabolic disorders as side effects. However, we did not find any statistical differences concerning S glucose, S triglycerides and S uric acid between thiazide-treated patients who developed or not metabolic alkalemia (Fig. 7a–c). Moreover, we did not find any correlation of these metabolic parameters concerning the amount of thiazide prescribed or S potassium levels (Table 4). Although the occurrence of gout in patients treated with thiazide can be developed, we did not find any case in our study.

Our study showed that metabolic alkalemia may be present in hypercalciuric nephrolithiasis outpatients, but it was a rare clinical disorder since we only diagnosed it in 4.3% of the cases. We also verified that thiazide prescription was prevalent in these cases.

In the title of the present study, we asked whether the presence of metabolic alkalemia in hypercalciuria stone formers mattered. We think we can answer this question by saying, “Yes, it does” since we found a high prevalence of hypokalemia in these patients. We would also like to point out that the majority of the patients with hypokalemic metabolic alkalemia were those treated with CTD. In addition, we found lower concentration levels of chloride, ionic calcium and magnesium in the serum of the thiazide-treated patients with metabolic alkalemia than those without this kind of acid-base disorder. Thus, it matters to diagnose metabolic alkalemia in hypercalciuric thiazide-treated patients and follow them up due to the possibility of electrolyte disorders being present.

Besides thiazide prescription, some patients have also been supplemented with potassium citrate to reduce calcium precipitation in the urine. Nevertheless, we did not find any difference in the results among the cases, whether they were supplemented with an alkali source or not. Thus, we assembled them into two groups: (a) patients with metabolic alkalemia (Groups I and II), (b) patients without metabolic alkalemia (Groups III and IV). In addition, we also verified the characteristics of hypercalciuric patients without thiazide treatment. However, we would like to point out that it is likely that these findings indicate a bias because the patients only treated with citrate (Group VI) were those using less antihypertensive medication, besides their low age.

We recognize that our study is limited by the small number of patients who developed metabolic alkalemia, which prevents us from making statements either about the safe use of thiazides or potassium citrate supplementation. However, this kind of acid-base disorder in thiazide-treated hypercalciuric nephrolithiasis outpatients had not been reported in the literature, yet.

Moreover, the present study does not allow us to corroborate or not the NOSTONE trial findings [8] because the number of patients who did not receive thiazide treatment was also small. Additionally, the aim of our study was not to investigate whether thiazide was a good choice to treat hypercalciuric kidney stone recurrence or not. However, it will be a good proposition for further investigations.

The main mechanism in the development of metabolic alkalemia is an increase in H⁺ secretion by intercalated cells that are expressed in the collecting duct. This process may be linked or not to chloride losses [10].

In the past, chloride losses were considered responsible for generating a kind of metabolic alkalosis called contraction alkalosis [9, 23]. Since chloride concentration regulates the extracellular volume, the losses of this anion result in the contraction of this compartment. Consequently, the bicarbonate concentration in the serum increases, leading to renal bicarbonate retention in the body [23]. Currently, this explanation is considered a consequence and no longer the cause of metabolic alkalosis [9].

In fact, chloride depletion is responsible for increments in sodium delivery to the collecting duct because the cotransporters Na⁺, K⁺-2Cl⁻ (NKCC2), and Na⁺-Cl⁻ (NCC) that are expressed, respectively, in the lumen cells of the thick ascending limb of the Henle’s loop and the distal tubule, do not work well. Therefore, a large amount of sodium reaches the luminal side of the collecting duct, which leads to its large reabsorption through ENaC that is expressed in the lumen of principal cells. This, in turn, generates high potential difference in the lumen of the collecting duct, favoring the secretion of K⁺ and H⁺ [9, 10, 24].

In the present study, we failed to demonstrate whether metabolic alkalemia was due to chloride depletion induced by thiazide because we found S chloride values registered in only 50% of the electronic medical records of patients who developed metabolic alkalemia. Moreover, we did not find any evaluation about chloride urinary excretion [10]. Hypokalemia is an electrolyte disorder often diagnosed in metabolic alkalemia that is mainly caused by two mechanisms: (1) K⁺ and H⁺ shifts between extra and intracellular compartments in order to promptly restore blood pH within normal range [25]; (2) potassium losses in the urine due to increases of this cation secretion by collecting duct cells, as explained above.

In the present study, we also remarked that most cases of hypokalemia have been treated with chlorthalidone. This thiazide-type has a longer half-life than HCTZ, which justifies the greater occurrence of severe side effects [13, 20]. As for the association between hypokalemia and the amount of thiazide taken daily, we failed to demonstrate it, probably due to the small sample size of patients who developed metabolic alkalosis. Furthermore, investigating the dose dependency of HCTZ in the NOSTONE trial, it revealed varying incidences of hypokalemia across different dosage groups. This observation highlights the need for careful monitoring and dosage adjustment to mitigate electrolyte imbalances.

Moreover, potassium depletion can worsen metabolic alkalemia because renal potassium reabsorption will be stimulated by increases in H⁺, K⁺-ATPase activity. This pump is expressed in the luminal side of intercalated cells in the collecting duct. For each molecule of K⁺ reabsorbed, one molecule of H⁺ is secreted [10, 24].

Another possibility to justify the development of metabolic alkalemia would be the citrate potassium supplementation. Many patients with calcium kidney stones receive this prescription to increase the solubility of calcium ions in the urine by rising urinary pH [14, 15], and this alkali source added to the body reduces the calcium concentration in the urine [16‒18]. Despite our limited group of patients with metabolic alkalemia, we have also observed a positive correlation between urinary citrate excretion and increment in urinary pH.

As we previously stated, we might expect aggravation of the metabolic alkalemia in patients supplemented with potassium citrate due to the alkali addition to the body. However, our results were similar in both groups, supplemented and non-supplemented with citrate potassium, either in patients with or without metabolic alkalemia. These outcomes agree with a study that analyzed, in a combining way, the serum and urine parameters of patients submitted to a long-term combined treatment with thiazide and potassium citrate. They showed evidences that potassium citrate supplementation helped maintain normokalemia without the development of metabolic alkalemia [15].

Nowadays, vitamin D supplementation is widely used in the Western world to get a better immune response. As it plays an important role in calcium absorption in the small intestine, a link between vitamin D serum concentration and an increase in kidney stone formers due to the increment in calciuria would be expected [26]. Another possibility would be the development of calcium-alkali syndrome [27], which can occur in patients treated with thiazide associated with vitamin D supplementation. For this reason, we tried to find whether the concentration of 25-hydroxyvitamin D or the amount of vitamin D taken per day would be correlated with categories linked to calcium homeostasis. Again, we recognize the limitation of our data due to the small number of patients with metabolic alkalemia.

Regarding other electrolyte disorders, alkalemia per se can reduce ionic calcium and magnesium concentration in the serum [21]. These cations have a high serum albumin affinity that increases when the blood pH rises, leading to a decrease in their serum concentrations, consequently. In our study, patients with metabolic alkalemia showed lower serum calcium and magnesium concentration than those without it. Nevertheless, we did not diagnose hypocalcemia because thiazide probably enhanced renal calcium reabsorption [4] and the bone tissue regulates calcium homeostasis [6]. However, thiazide downregulates the Mg²⁺-channel expression in the distal tubular cells raising magnesium urinary excretion. These magnesium urinary losses can result in hypomagnesemia [4].

Besides metabolic abnormalities, hypokalemia associated with hyperuricemia could play an important role in endothelial dysfunction, worsening the metabolic syndrome [11, 13]. In our study, we did not find any correlation between metabolic parameters and the amount of thiazide prescribed or serum potassium levels, as described in the literature. Again, we need to consider the sample size of our study, especially regarding the patients who developed metabolic alkalemia.

Lastly, we did not detect any correlation between the amount of thiazide taken per day and eGFR, too. These findings are contrary to some studies in humans that suggest an association between chronic thiazide use and the development of end-stage renal disease, despite the beneficial effect of this medication in lowering blood pressure [12]. The explanation to this negative side effect is based on the activation of the renin-angiotensin-aldosterone system due to extracellular volume depletion. Hence, renal vasoconstriction and other negative factors, like fibrosis, are developed, resulting in renal injuries [28, 29].

Although the development of metabolic alkalemia was uncommon in hypercalciuric kidney stone formers, we can suggest that the venous blood gas test and the determination of the concentrations of serum electrolytes matter should be included in the schedule of laboratory tests during their outpatient follow-up. Moreover, the prescription of thiazides, associated or not with potassium citrate supplementation, seems to be safe in these cases.

The authors would like to thank the Laboratório de Investigação Médica – HC-FMUSP and the Fundação Faculdade de Medicina (FFM).

This study followed ethical standards established by the World Medical Association Declaration of Helsinki. The Research Ethics Committee of our Institution (Comissão de Ética para Análise de Projetos de Pesquisa – CAPPesq, da Diretoria Clínica do Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo) approved this study on February 2, 2017, CAPPesq No. 1.905.977. Written informed consent was not obtained from individual patients because we collected the data from electronic medical records of nephrolithiasis outpatients. This decision is in accordance with the opinion of the Ethics Committee of our Institution (Comissão de Ética para Análise de Projetos de Pesquisa – CAPPesq, da Diretoria Clínica do Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo), which approved this study on February 2nd, 2017, CAPPesq No. 1.905.977. This study does not have a trial ID because it is a cohort retrospective study.

The authors have no conflicts of interest to declare.

Renato V.M. Starek received a grant from the Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP, No. 2017/03029-0) because he was an undergraduate student of Faculdade de Medicina da Universidade de São Paulo.

Renato V.M. Starek, Samirah A. Gomes, and Claudia M.B. Helou contributed equally to the conception of the study, collection, analysis, and interpretation of data, as well as for the initial writing of the article; have also read and approved the manuscript in its present form; and have agreed to its submission to the Kidney and Blood Pressure Research. Samirah A. Gomes and Claudia M.B. Helou were responsible for reviewing all data analysis and reviewing the article for publication.

The data that support the findings of the current study are not publicly available because they contain information that could compromise the privacy of research participants but are available from the corresponding author (C.M.B.H.) and after permission of the Ethics Committee of the Hospital das Clínicas da Faculdade de Medicina da USP – CAPPesq.