Alkaline Salts to Counteract Bone Resorption and Protein Wasting Induced by High Salt Intake: Results of a Randomized Controlled Trial

High sodium chloride (NaCl) intake can induce low-grade metabolic acidosis (LGMA) and may thus influence bone and protein metabolism.

We hypothesized that oral potassium bicarbonate (KHCO₃) supplementation may compensate for NaCl-induced, LGMA-associated bone resorption and protein losses. Eight healthy male subjects participated in a randomized trial with a crossover design. Each of two study campaigns consisted of 5 d of dietary and environmental adaptation followed by 10 d of intervention and 1.5 d of recovery. In one study campaign, 90 mmol KHCO₃/d were supplemented to counteract NaCl-induced LGMA, whereas the other campaign served as a control with only high NaCl intake.

When KHCO₃ was ingested during high NaCl intake, postprandial buffer capacity ([HCO₃⁻]) increased (P = 0.002). Concomitantly, urinary excretion of free potentially bioactive glucocorticoids [urinary free cortisol (UFF) and urinary free cortisone (UFE)] was reduced by 14% [∑(UFF,UFE); P = 0.024]. Urinary excretion of calcium and bone resorption marker N-terminal telopeptide of type I collagen was reduced by 12 and 8%, respectively (calcium, P = 0.047; N-terminal bone collagen telopeptide, P = 0.044). There was a trend of declining net protein catabolism when high NaCl was combined with KHCO₃ (P = 0.052).

We conclude that during high salt intake, the KHCO₃-induced postprandial shift to a more alkaline state reduces metabolic stress. This leads to decreased bone resorption and protein degradation, which in turn might initiate an anticatabolic state for the musculoskeletal system in the long run.

For decades, the potential risk of high sodium chloride (NaCl) intake, not only for salt-sensitive hypertension but also for osteoporosis, has been the object of investigation (1–4). We have recently shown that high NaCl intake is accompanied by low-grade metabolic acidosis (LGMA) (5–7), which in turn might contribute to the bone loss described by others (8–11). According to Frassetto et al. (12), a high NaCl intake has as much as 50–100% of the acidifying properties of the diet's net acid load, another potential source of diet-induced LGMA. The latter has been negatively associated with bone cortex area and bone mineral content during growth (13) and bone mineral density in pre- and perimenopausal women, and positively with markers of bone resorption in pre-, peri-, and postmenopausal women (14, 15).

In addition to the effect on bone turnover, the catabolic mechanisms of acidosis include stimulation of branched-chain amino acid breakdown (16) and proteolytic activation of the ubiquitin-proteasome system (17), as well as increased glucocorticoid (GC) activity (18), which may have an impact on whole-body protein metabolism. Consistent with these results, in another study we found exaggerated disuse-induced nitrogen losses with NaCl-induced LGMA (6).

On the other hand, counteracting LGMA induced by a high dietary acid load by dietary base precursors seems to maintain bone mass and protein content; a reduction of urinary calcium and bone resorption markers (19–21), an increase in bone mineral density (22), and a decline in urinary nitrogen (23, 24) have been reported. Therefore, we hypothesized that alkaline loading also may counteract NaCl-induced bone resorption and protein wasting.

Subjects and Methods

Study protocol

Eight healthy male volunteers (mean age, 26 yr, sd 4; mean body weight, 75 kg, sd 3.6; mean height, 178.6 cm, sd 5.7) participated in an interventional, randomized crossover trial examining the effects of potassium bicarbonate (KHCO₃) during high NaCl intake on bone metabolism and protein turnover, the “Salty Life 8 Study” (Fig. 1). For the whole duration of the study, subjects were confined to the metabolic ward of the German Aerospace Center (Cologne, Germany). Primary outcome measures were urinary markers of bone resorption; secondary outcome measures included markers of bone formation, acid-base status, whole-body protein turnover, urinary free cortisol (UFF), and urinary free cortisone (UFE). The protocol was approved by the ethics commission of the Aerztekammer Nordrhein (Duesseldorf, Germany) and conducted in accordance with the Declaration of Helsinki. The study consisted of two campaigns, each divided into 5 d of dietary and environmental adaptation, 10 d of intervention, and 1.5 d of medical recovery. During the adaptation periods, subjects received an average normal NaCl intake [2.6 mmol NaCl/kg body mass (BM)/d]. During the intervention periods, a very high NaCl diet (7.3 mmol NaCl/kg BM/d) was applied and was supplemented with 90 mmol KHCO₃/d in one study campaign. The very high amount of NaCl was considered to give evidence on changes in the outcome measures already within a rather short time frame. Baseline data were collected at the end of the adaptation period. The amount of time between the two study campaigns was 25 d. Subjects were randomized 1:1 to “high NaCl first” or “high NaCl + KHCO₃ first.” The two campaigns were identical with respect to environmental conditions (mean temperature, 22.7 sd 1.9 C; mean humidity, 63.2 ± 3.7%), study protocol, and diet. Physical exercise was prohibited. Subjects were not allowed to lie in bed or sleep for 16 h during the day.

Fig. 1.

A, Study design. B, Mean nutrient intake.

Subject selection

Before subjects were enrolled, an extensive medical and psychological screening process was applied to exclude any medical or psychological disorders that would influence the ability to accomplish the objectives of the study. All volunteers were nonsmokers, did not use any medication, and were exercising moderately less than four times a week before the study began. To avoid any systematic errors from vitamin D deficiency, the minimum 25-hydroxyvitamin D levels to achieve were 30 μg/liter. All volunteers gave written informed consent before starting the protocol.

Diet and KHCO₃ supplementation

Subjects received an individually tailored, weight-maintaining diet for the entire study. The resting metabolic rate of each subject was computed by indirect calorimetry on the first day in the ward (Deltatrac II MBH 200 Metabolic Monitor; Datex Ohmeda, Madison, WI). To account for low physical activity, 40% of the resting metabolic rate was added, plus 10% to account for diet-induced thermogenesis. Figure 1B shows the intake of all nutrients controlled. Vitamin D3 was supplemented in the amount of 400 IU, and endogenous production through exposure to UV exposition was prevented by access to artificial light exclusively. Caffeine, methylxanthine, and alcohol intake were prohibited.

Individual menus of each volunteer were predefined with the nutritional calculation software PRODI 4.2 (Nutri-Science GmbH, Hausach, Germany). Bread was homemade, and meals were provided by a manufacturer with accurate nutrition facts (Apetito AG, Rheine, Germany). Food with a sodium content greater than 20 mmol/100 g was chemically analyzed. Meals were prepared according to the method of weighed intake (25). During the entire study, the volunteers strictly adhered to the provided diet.

During the adaptation period, the acid load of the diet, estimated by the potential renal acid load (PRAL) model (26), was marginally alkaline (−9.9 ± 13.4 mEq/d). However, to study the antagonistic properties of KHCO₃ toward NaCl-induced LGMA (when diet per se is nonacidogenic), PRAL was clearly alkaline (−37.1 ± 18 mEq/d) during the intervention periods. This background alkalization was realized by application of a bicarbonate-rich mineral water.

KHCO₃ was supplied in the amount of 30 mmol three times a day with the main meals to allocate the alkaline load throughout the day. The NaCl content of the meals was 20–23% for breakfast, 30–33% for lunch, 25–28% for dinner, and less than 10% for defined snacks between the meals and before bedtime (2130 h), respectively.

Blood and urine sampling

Urine was collected void by void during the entire study period and pooled in the laboratory for 24-h collection periods. Samples were aliquoted and frozen at −20 C or −80 C until analysis. Whole blood was centrifuged (3000 rpm, 4 C) after coagulation, and serum was distributed into aliquots and immediately frozen at −20 C or −80 C. Arterialized capillary blood for blood gas analysis was collected after blood drawing in the fasting state. Additional blood gases were analyzed before (baseline) and 1, 1.5, 2, and 2.5 h after ingestion of a NaCl-rich meal with/without KHCO₃ on d 3 and 8 of the intervention periods. The samples were collected in an anticoagulant tube by trained personnel and were immediately analyzed by an automated blood gas system (ABL 5; Radiometer, Willich, Germany) for pH, and partial pressure of carbon dioxide and oxygen (pCO₂, pO₂). Bicarbonate concentration ([HCO₃⁻) and base excess (BE) were calculated according to the Henderson-Hasselbalch equation.

Analysis

Serum sodium and potassium concentrations were analyzed by flame photometry (Efox 5053; Eppendorf, Hamburg, Germany), and chloride concentration by an automated analyzer (COBAS INTEGRA 400; Roche Diagnostics, Mannheim, Germany). Urinary pH was measured with a pH-sensitive electrode (inoLab pH720; WTM, Weilheim, Germany). Net acid excretion (NAE) was analyzed according to Luthy et al. (41) and calculated as titratable acid (TA) + ammonia (NH₄⁺) − bicarbonate (HCO₃⁻). All titrations were carried out with an automated Titrator (DL50; Mettler Toledo, Giessen, Germany). Precision of the measurements was determined by analyzing daily standard solutions. The coefficient of variation was 1.73% for NH₄⁺ and 1.15% for HCO₃⁻ measurement. C- and N-terminal bone collagen telopeptides (UCTX, UNTX) in 24-h urine were used as markers of bone resorption, and serum levels of bone-specific alkaline phosphatase (bAP) and N-terminal propeptide type I (PINP) as markers of bone formation. All biomarkers of bone metabolism were analyzed by commercially available immunoassays (UCTX, urine crossLaps; Immunodiagnostic Systems, Bolton, UK; UNTX, Osteomark; Wampole Laboratories, Princeton, NJ; bAP, Ostase Irma; Immuntech, Veverska Bityska, Czech Republic; and PINP, PINP RIA, Orion Diagnostica, Espoo, Finland). Samples of each subject were analyzed in a batch to avoid interassay variation; intraassay variation was as follows: bAP, 5.9%; PINP, 2.2%; UCTX, 3.8%; and UNTX, 2.4%. Total urinary nitrogen was determined by highly sensitive chemiluminescence with an automated analyzer (Antec 7000V; CTC Analytics AG, Zwingen, Switzerland). The coefficient of variation of this method was 1.6% within a series of analyses performed on 1 d using freshly prepared calibrators. Nitrogen balance was estimated as nitrogen intake (protein/6.25) minus urinary nitrogen. Losses through skin and feces were not taken into account because they are usually low and are regarded as constant (27).

UFF and UFE were quantified by specific in-house RIAs established at the Steroid Laboratory of the Department of Pharmacology, University of Heidelberg (28). Before RIA recovery-corrected extraction and chromatographic purification were performed, intra- and interassay coefficients of variation were less than 10% and less than 15%, respectively. Normal values range from 10 to 60 μg/d (UFF) and from 20 to 140 μg/d (UFE).

We used phenylalanine and tyrosine stable isotopes to trace whole-body protein kinetics in the postabsorptive state according to Wolfe (29). After overnight fasting, 3 h of primed constant infusion of isotopes was carried out via forearm vein catheters: L[ring-D5]phenylalanine (prime, 252 μmol; infusion rate, 4.2 μmol/min), L[ring-D4]tyrosine (prime, 84 μmol), L[ring-D2]tyrosine (prime, 126 μmol; infusion rate, 2.1 μmol/min). Blood samples were taken from the contralateral wrist-forearm vein 150, 165, and 180 min after the constant infusion started to determine steady-state plasma enrichment of tracers. The forearm was heated beforehand to obtain arterialized venous blood. Tracers were purchased from Cambridge Isotope Laboratory (Andover, MA). Sterile solutions were shown to be pyrogen-free before use. Plasma enrichment of phenylalanine and tyrosine was determined using gas chromatography-mass spectrometry (HP 5890; Agilent Technologies, Santa Clara, CA). Whole-body rates of protein synthesis, degradation, and balance were calculated as referenced (30). Calculations of tracer and trace ratios include corrections for background enrichment.

Statistics

Anthropometric data are presented as mean ± sd, and all results are expressed as means ± sem. Statistical analyses were performed with the software STATISTICA (StatSoft, Hamburg, Germany). Comparability of baseline levels was evaluated with Student's paired t test for dependent samples. If not otherwise stated, baseline data were identical, so that effects of the intervention are considered to depend on the treatment. Data from postprandial blood gas analysis were baseline-adjusted, and the sets from d 3 and 8 were pooled. A two-way ANOVA with repeated measures was performed to compare data sets from the two interventions. An effect of KHCO₃ administration was considered significant when either the influence of KHCO₃ (“treatment”) or the interaction of KHCO₃ and time (“time*treatment”) was significant. If time-dependent differences between high NaCl and high NaCl + KHCO₃ were observed, a post hoc test (Newman-Keuls) was performed. If not otherwise stated, the text describes “treatment” effects. Differences in whole-body protein kinetics between the treatments were assessed by Student's paired t test for dependent samples. For data that did not achieve normal distribution (NAE, TA, HCO₃⁻), nonparametric testing was performed (Wilcoxon). Differences between means were considered statistically significant when P < 0.05.

Results

Bone metabolism

With KHCO₃ supplementation, calcium excretion was reduced by 12% (P = 0.047) and N-terminal bone collagen telopeptide (NTX) excretion by 8% (P = 0.044). However, creatinine clearance was stable (P = 0.070). No difference in C-terminal bone collagen telopeptide (CTX) excretion between the two interventions could be detected (P = 0.176) (Fig. 2). Markers of bone formation remained unchanged as well (PINP, P = 0.926; bAP, P = 0.953) (Table 1).

Fig. 2.

Urinary excretion of calcium (A) and bone resorption markers UNTX (B) and UCTX (C) for the respective study days of both intervention periods. †, ANOVA repeated measures, treatment effect.

Table 1.

Serum concentrations of bone formation markers bAP and PINP along with serum sodium, chloride, and potassium concentrations ([Na⁺], [Cl⁻], [K⁺]), pH, [HCO₃⁻], BE, and blood gases (pCO₂, pO₂) in capillary fasting morning blood for the respective study days of both intervention periods

	BL	Day 3	Day 6	Day 8	Day 10
bAP (μg/liter)
High NaCl	14.46 ± 0.77	13.74 ± 0.84	14.11 ± 0.89	13.91 ± 0.86	14.44 ± 1.03
High NaCl + KHCO3	14.14 ± 1.49	14.60 ± 1.44	14.17 ± 1.57	14.36 ± 1.66	14.80 ± 1.76
P value	0.691a	0.953b
PINP (mmol/liter)
High NaCl	69.22 ± 11.12	66.77 ± 9.78	67.53 ± 8.57	66.83 ± 10.26	67.87 ± 8.35
High NaCl + KHCO3	66.73 ± 5.99	66.97 ± 6.92	67.20 ± 6.77	67.37 ± 7.34	68.86 ± 7.96
P value	0.867a	0.926b
[Na+] (mmol/liter)
High NaCl	135.6 ± 0.5	136.3 ± 0.6	136.3 ± 0.6	136.6 ± 0.5	137.2 ± 0.3
High NaCl + KHCO3	136.2 ± 0.7	137.4 ± 0.7	134.9 ± 0.4	136.2 ± 0.4	136.1 ± 0.6
P value	0.437a	0.317b
[Cl−] (mmol/liter)
High NaCl	102.4 ± 0.9	104.4 ± 1.2	104.4 ± 1.1	104.5 ± 0.7	104.1 ± 1.1
High NaCl + KHCO3	102.3 ± 1.1	102.2 ± 3.1	103.4 ± 1.1	104.4 ± 0.9	104.2 ± 1.1
P value	0.946a	0.911b
[K+] (mmol/liter)
High NaCl	3.96 ± 0.11	3.82 ± 0.22	3.72 ± 0.13	3.83 ± 0.14	3.84 ± 0.21
High NaCl + KHCO3	3.99 ± 0.05	3.97 ± 0.29	3.93 ± 0.25	3.99 ± 0.10	3.89 ± 0.09
P value	0.664a	0.012b
pH
High NaCl	7.412 ± 0.01	7.411 ± 0.01	7.415 ± 0.01	7.426 ± 0.00	7.416 ± 0.01
High NaCl + KHCO3	7.414 ± 5 * 10−3	7.414 ± 5 * 10−3	7.42 ± 0.01	7.419 ± 5 * 10−3	7.411 ± 5 * 10−3
P value	0.691a	0.695b
[HCO3−] (mmol/liter)
High NaCl	25.56 ± 0.58	25 ± 0.47	25.69 ± 0.46	25.85 ± 0.46	25.85 ± 0.53
High NaCl + KHCO3	26.25 ± 0.62	25.5 ± 0.49	26.04 ± 0.49	26.13 ± 0.44	26.19 ± 0.65
P value	0.054a	0.043b
BE (mmol/liter)
High NaCl	0.88 ± 0.43	0.60 ± 0.42	1.29 ± 0.41	1.65 ± 0.39	1.35 ± 0.48
High NaCl + KHCO3	1.63 ± 0.58	1.13 ± 0.44	1.56 ± 0.44	1.63 ± 0.36	1.44 ± 0.57
P value	0.040a	0.102b
pCO2 (mm Hg)
High NaCl	40.19 ± 0.83	39.5 ± 0.57	40.63 ± 0.72	39.58 ± 0.64	40.48 ± 0.76
High NaCl + KHCO3	41.38 ± 0.86	40.4 ± 0.70	40.44 ± 0.91	40.69 ± 0.81	41.19 ± 0.92
P value	0.014a	0.092b
pO2 (mm Hg)
High NaCl	71.94 ± 1.5	67.75 ± 1.54	66.21 ± 1.94	66.59 ± 1.99	68.48 ± 2.07
High NaCl + KHCO3	70 ± 1.93	69.92 ± 2.81	67.56 ± 1.96	67.31 ± 1.79	67.56 ± 2.63
P value	0.532a	0.697b

Data are expressed as mean ± sem, unless otherwise specified. BL, baseline.

^a

Student's paired t test for dependent samples.

^b

ANOVA repeated measures, treatment effect.

Protein metabolism

When KHCO₃ was supplemented during high NaCl intake, the difference in rates of postabsorptive phenylalanine hydroxylation, used as a marker of net protein catabolism, was just below the stated level for statistical significance (high NaCl, 6.41 ± 0.36 μmol/min; high NaCl + KHCO₃, 5.78 ± 0.40 μmol/min; P = 0.052) (Fig. 3). Rates of postabsorptive proteolysis, quantified as rate of phenylalanine appearance, as well as rates of protein synthesis, were not influenced by KHCO₃ administration (rate of phenylalanine appearance, high NaCl, 60.84 ± 2.65 μmol/min; high NaCl + KHCO₃, 61.95 ± 2.38 μmol/min; P = 0.616; and protein synthesis, high NaCl, 54.43 ± 2.54 μmol/min; high NaCl + KHCO₃, 56.17 ± 2.23 μmol/min; P = 0.389). Nitrogen balance, an indicator for changes in whole-body protein content, remained unchanged when supplementing KHCO₃ (mean high NaCl, 1.94 ± 0.14 g/d; mean high NaCl + KHCO₃, 2.12 ± 0.08 g/d; P = 0.550).

Fig. 3.

Phenylalanine hydroxylation after 10 d of high NaCl ± KHCO₃. *, Student's paired t test for dependent samples.

Glucocorticoids

With KHCO₃ supplementation, excretion of potentially bioactive GCs [Σ(UFF, UFE)] in 24-h urine was reduced by 14% (high NaCl, 193.65 ± 30.46 μg/d; high NaCl + KHCO₃, 165.77 ± 16.54 μg/d; P = 0.024).

Electrolyte and acid-base status

During high NaCl intake, serum sodium and chloride concentration in the fasted state remained unchanged when KHCO₃ was supplemented, but as expected, the potassium concentration was higher (P = 0.012) (Table 1).

[HCO₃⁻] in the fasting morning blood was significantly higher when subjects took the KHCO₃ supplement during high NaCl intake (P = 0.043). However, because baseline levels before KHCO₃ administration already showed an upward trend relative to the control data (P = 0.054), this cannot be interpreted to depend only on the treatment. pH, BE, pO₂, and pCO₂ in the fasting state were not affected by KHCO₃ (Table 1).

However, supplementation of KHCO₃ induced postprandial changes in systemic acid-base status (Fig. 4): pH slightly increased over the postprandial period (P = 0.036, time*treatment effect). Yet, according to the post hoc test, there was a pure treatment trend only within the first hour after lunch ± supplement (high NaCl, −0.006 ± 0.002; high NaCl + KHCO₃, +0.005 ± 0.003; P = 0.054). Postprandial [HCO₃⁻] and BE were significantly higher for the whole duration of postprandial measurements after KHCO₃ ingestion ([HCO₃⁻], P = 0.002; BE, P < 0.001).

Fig. 4.

pH (A) and [HCO₃⁻] (B) in capillary blood. Postprandial changes from baseline (BL) after NaCl-rich meal ± KHCO₃. †, ANOVA repeated measures, treatment effect; ‡, ANOVA repeated measures, time*treatment effect; *, Newman-Keuls post hoc test, P = 0.032.

As expected, urinary acid-base status was alkaline when high NaCl intake was supplemented with KHCO₃: mean urinary pH was significantly higher during KHCO₃ administration (high NaCl, 6.63 ± 0.02; high NaCl + KHCO₃, 7.12 ± 0.02; P < 0.001). Consistently, net acid excretion was lower by an average of 75.31 mEq/d (high NaCl, 35.82 ± 3.14 mEq/d; high NaCl + KHCO₃, −39.50 ± 2.26 mEq/d; P = 0.012) (Table 2). This difference is almost equivalent to the calculated amount of potassium absorbed from the supplement (72.14 mEq/d).

Table 2.

Urinary NAE and excretion of TA, HCO₃⁻, and ammonium (NH₄⁺)

	Intervention, mean day 9/10	Baseline, mean day 4/5
NAE (mmol/liter)
High NaCl	35.82 ± 3.14	49.17 ± 1.06
High NaCl + KHCO3	−39.50 ± 2.26	48.93 ± 1.46
P value	0.012a	0.674a
TA (mmol/liter)
High NaCl	11.47 ± 0.85	16.81 ± 1.14
High NaCl + KHCO3	−2.91 ± 0.43	14.75 ± 0.86
P value	0.012a	0.017a
[HCO3−] (mmol/liter)
High NaCl	24.86 ± 1.41	14.79 ± 1.16
High NaCl + KHCO3	71.42 ± 1.71	12.56 ± 1.13
P value	0.012a	0.161a
[NH4+] (mmol/liter)
High NaCl	49.21 ± 2.63	47.14 ± 1.79
High NaCl + KHCO3	34.83 ± 1.42	46.74 ± 1.97
P value	<0.001b	0.759b

Data are expressed as mean ± sem, unless specified otherwise.

^a

Wilcoxon test.

^b

Student's paired t test for dependent samples.

Discussion

In the present study, supplementation of 3 × 30 mmol KHCO₃/d for 10 d decreased NaCl-induced calciuria by 12%. An effect on bone turnover was visible in one of the measured bone resorption markers (NTX), reduced by 8%. Net protein catabolism, quantified as postabsorptive phenylalanine hydroxylation, tended to decrease.

As previously shown, a high NaCl intake shifts the systemic equilibrium of buffer substance and pH slightly in the acid direction (5, 6), whereas alkaline mineral salts reduce net endogenous acid production by directly providing alkali equivalents or after cellular metabolism (31). Simultaneous changes in bone and protein turnover are often attributed to respective shifts in acid-base status (5, 6, 4, 12). Accordingly, we investigated the neutralizing effect of NaCl and KHCO₃ on each other. After NaCl-rich meals were supplemented with KHCO₃, metabolic components of acid-base status were increased. However, these immediate alkalizing properties of KHCO₃ were no longer clearly detected in the fasting morning blood samples. Given that a very high NaCl intake induced persistent LGMA with changes still apparent in the fasting morning blood (5, 6), supplementation of 3 × 30 mmol KHCO₃ seemed insufficient to consistently counterbalance those changes. One might argue that the relation of more than 500 mmol NaCl:90 mmol KHCO₃ was unbalanced to compensate for the effects of the high NaCl intake. However, our data show that the observed limited acid-base balancing properties of KHCO₃ with high NaCl intake arise from renal excretion of the alkali load: under KHCO₃ supplementation, NAE was lower in an amount almost equivalent to the absorbed base precursors. We conclude that with a diet very high in NaCl but basically alkaline (low in PRAL), there was no physiological need for additional alkalization.

Yet, the alkalizing properties of KHCO₃ in this study may have even been restricted by the particular metabolic consequences of a high NaCl load, as findings from Cogan et al. (32) demonstrate: a net acid-producing diet with low NaCl intake was supplemented with 2 mmol NaHCO₃/kg BM/d; these conditions generated an alkalization of plasma and urine. However, the systemic [HCO₃⁻] and pH increase were again counteracted by addition of 2 mmol NaCl/kg BM/d to the diet; urinary alkalization remained unchanged. The authors conclude that NaCl loading reduces the renal set point for [HCO₃⁻] reabsorption; the higher NaCl intake led to adjustment of the [HCO₃⁻] set point to a concentration of 25 mmol/liter, which is 3 mmol/liter below the set point during preagricultural extremely low NaCl intakes. Correspondingly, in our subjects who consumed a high-NaCl diet, even marked alkalization of diet and urine led only to moderate short-term increases in systemic [HCO₃⁻] at an average set point of about 26 mmol/liter. These findings are in agreement with the Stewart theory (33), which implies that the reduction of [HCO₃⁻] serves as electroneutral compensation of increased serum chloride concentration.

However, our findings demonstrate that no continuous reduction in circulating [HCO₃⁻] and free protons is required to affect such functional physiological measures as GC activity. The strong reduction in daily acid load and respective short-term increases in systemic pH and [HCO₃⁻] were sufficient to positively influence GC status. Earlier studies had already provided evidence that the slight hyperglucocorticoidism of LGMA can be compensated by alkaline minerals (34). This hormonal effect may have contributed to the observed mild “anticatabolism,” as described below.

With very high NaCl intake, we found a preserving effect of KHCO₃ on calcium metabolism and bone resorption, which is basically in line with data from previous studies in which alkaline minerals were applied to counterbalance PRAL-induced LGMA. Ceglia et al. (23) found that urinary calcium decreased by 9% when they supplemented a high-protein diet (1.5 g/kg BM/d) with 90 mmol KHCO₃/d for 10 d. Dawson-Hughes et al. (19) supplemented an average American diet with 67.5 mmol KHCO₃/d for 3 months and achieved a 13.4% reduction of urinary NTX relative to the placebo group. This influence of clinically slight alterations in acid-base status on osteoclastic function could be mediated by soluble adenylyl cyclase acting as a [HCO₃⁻] sensor (35) or, as described above, by changes in GC activity. Both an inhibition and a knockdown of the former showed in vitro an inability of HCO₃⁻ to block osteoclast growth and differentiation; the findings were confirmed in intact bone (35). Even when added to a nonacidogenic diet with a PRAL of −27 mEq/d, potassium citrate can at least acutely reduce urinary calcium and NTX (36), indicating that bone-related effects of alkalizing agents are not necessarily restricted to prevailing states with higher PRAL.

In the context of NaCl-induced bone loss, data from a randomized, placebo-controlled trial give evidence that an oral alkali load can prevent the characteristic increase in calcium excretion and bone resorption. The rise in urinary calcium and NTX with increasing sodium intake from low to moderately high (87 to 225 mmol/d) was inhibited by adding 90 mmol potassium citrate to the diet (21). However, in our study the hypothesized bone-preserving properties of KHCO₃ did not suffice to compensate for the full metabolic consequences of a very high NaCl intake as found in earlier studies; Frings-Meuthen et al. (5) reported a 33% rise in UNTX when sodium intake was increased from a moderately high level (2.8 mmol Na/kg BM/d) to a very high level (7.7 mmol Na/kg BM/d). In the present study, this was limited to no more than 8%. At the same time, urinary CTX was unchanged.

These findings show that NaCl-induced alteration of acid-base status is just one of several factors contributing to bone catabolism when NaCl intake is very high. Other mechanisms, such as hormonal or osteoimmunological effects as well as ion-exchange processes at the osteoclast-bone interface, are reasonable but, to our knowledge, are not yet clarified in depth (37, 38).

NaCl-induced increases in bone catabolic processes are accompanied by reductions in whole-body protein content (6). In the present study, we observed a strong trend of reduced phenylalanine hydroxylation. Because markers of protein turnover were defined as a secondary outcome measure of the study, the effect size (0.54) turned out to be too low to prove potential effects with n = 8. In light of this result, we interpret our findings as a suggestion that mild protein-preserving effects of KHCO₃ occur during high NaCl intake, but this hint must be confirmed in a larger sample size. With respect to whole-body protein content, favorable effects of KHCO₃ have been reported with PRAL-induced LGMA; nitrogen losses with a protein-rich diet were reduced within 10 d of KHCO₃ administration (90 mmol/d) (23), and nitrogen losses observed during ingestion of a “Western diet” were reduced within 18 d of supplementation (60–120 mmol KHCO₃/d) (24). In our study, nitrogen balance was not affected within 10 d of KHCO₃ supplementation, although net protein catabolism showed a downward trend, and GC activity was reduced. However, hydroxylation of phenylalanine was used as an early marker for changes in whole-body protein turnover, whereas nitrogen balance—even under strictly controlled study conditions—identifies changes in protein content with a precision of 0.1% (1.93 g for BM of 75 kg). Accordingly, further long-term investigations are needed to examine the relevance of the detected changes to protein content, muscle mass, and function.

In summary, we observed anticatabolic effects of KHCO₃ with NaCl-induced LGMA. Still, oral administration of 3 × 30 mmol KHCO₃/d did not compensate for the entire metabolic alteration induced by a very high NaCl intake as documented previously (6). Whether the observed changes in turnover are applicable to maintenance of bone and muscle mass as well as function needs to be clarified in long-term observations. However, a NaCl intake of that magnitude will be obtained only in a few populations (39) or under extreme environmental conditions, such as during space flight (40). It is more likely that alkalizing diets reduce catabolic effects from a rather usual, moderately high NaCl intake, as indicated by Sellmeyer et al. (21).

Acknowledgments

We thank the volunteers who gave time and effort to ensure the success of the project; the staff of the space physiology department for their collaboration in conducting the study; F. May, who was in charge of the medical care of the volunteers; the team of G. Kluge and H. Soll, who carried out medical and psychological screenings of the volunteers; G. Kraus, J. Hjorth-Müller, I. Schrage, and M. Sturma for excellent biochemical analysis; the team of K. Ruberg for preparation of the iv solutions; and J. Krauhs for revision of the manuscript.

J.B. was supported by a scholarship from the Wernher von Braun Foundation. The research was funded by the DLR space program.

Disclosure Summary: The authors have no conflicts of interest.

Abbreviations

 
• bAP

  Bone-specific alkaline phosphatase

 
• BE

  base excess

 
• BM

  body mass

 
• CTX

  C-terminal bone collagen telopeptide

 
• GC

  glucocorticoid

 
• [HCO₃⁻]

  bicarbonate concentration

 
• KHCO₃

  potassium bicarbonate

 
• LGMA

  low-grade metabolic acidosis

 
• NaCl

  sodium chloride

 
• NAE

  net acid excretion

 
• NTX

  N-terminal bone collagen telopeptide

 
• PINP

  N-terminal propeptide type I

 
• PRAL

  potential renal acid load

 
• TA

  titratable acid

 
• UCTX

  urinary CTX

 
• UFE

  urinary free cortisone

 
• UFF

  urinary free cortisol

 
• UNTX

  urinary NTX.