Introduction
Chronic metabolic acidosis (CMA) is a manifestation of retained acid from nonrespiratory sources. It is associated primarily with advanced CKD (1). However, recent studies have widened the definition of MA, and demonstrated new populations at risk of this abnormality. In patients with early CKD, acid may be retained before changes in serum electrolytes (2), and mild anion gap (AG) elevations may also be present (3). Our studies (4,5) demonstrated that in patients with normal kidney function (eGFR ≥90 ml/min per 1.73 m2), greater obesity is a risk factor for AG elevations and MA (4,5). This finding implicates obesity alongside reduced kidney function as a risk factor for CMA. Acid retention has far-reaching adverse effects on human health, including peripheral insulin resistance (6), kidney disease development and progression (7,8), connective tissue degradation (9,10), and impaired physical functioning (11).
However, the exact source of this abnormality in obesity is uncertain: body mass index (BMI) does not provide specific information about pathology. It is more likely that pathologic fat mass associated with greater body mass, or metabolic dysfunction related to obesity, mediates this disturbance. A prior cross-sectional study showed an association of greater waist circumference (WC) with lower serum bicarbonate (12); however, this analysis was limited to patients with CKD (12). A second cross-sectional study showed an association of greater insulin resistance with lower serum bicarbonate and higher AG (13). This study did not investigate metabolic disease states more generally; neither did it include adjustment for dietary acid (14).
Given the implications of acid retention on human health, and the high prevalence of metabolic syndrome (MetS) in the United States and worldwide, this study was designed to explore whether elements of metabolic disease are risk factors for MA and anion elevations. In total, 20 years of available data from the continuous National Health and Nutrition Examination Survey (NHANES) were compiled to test the following hypotheses: (1) greater WC, and a greater number of features of the MetS are associated with lower serum bicarbonate, higher AG, and AGMA; and (2) associations of BMI with these outcomes are attenuated after accounting for metabolic disease.
Methods
Study Population
Continuous NHANES is an annual, nationally representative survey of the United States noninstitutionalized civilian population. At the time of our analyses, data from 20 NHANES survey years (1999–2018) (n=101,316) were available. Inclusion criteria were age ≥18 years and nonpregnant (initial n=54,131). Exclusions were self-reported pulmonary disease (n=9555), liver disease (n=2271), congestive heart failure (n=1730), or cancer (n=4816); history of dialysis or an eGFR <15 ml/min per 1.73 m2 (n=1654); or incomplete anthropometric, laboratory, or dietary data (n=7783). A total of 31,163 participants were included as the main cohort. A subgroup of 12,860 participants provided fasting laboratory measurements for determining the presence of MetS.
Data Collection
Demographics and past medical history were self-reported. BMI was calculated as weight in kilograms divided by height in meters squared. Diabetes and hypertension diagnoses were defined by participant report of a physician diagnosis, use of medications, or an A1c >=6.5%, or BP >140/90 mm Hg. Coronary artery disease was defined by participant report of a physician diagnosis. A 24-hour dietary recall was obtained by study personnel using a computerized questionnaire, and dietary macro- and micronutrient content were quantified using a standardized process.
Formulae and Definitions
eGFR was calculated using the Chronic Kidney Disease Epidemiology Collaboration equation (15). The urine albumin to creatinine ratio (mg/g) was calculated as urine albumin (mg/dl)/urine creatinine (g/dl). Dietary acid was calculated as estimated net endogenous acid production (mEq/L) = ([54.5 × protein (g/d)]/potassium [mEq/d]) − 10.2 (16). Insulin resistance was calculated using the homeostasis model assessment for insulin resistance = (fasting plasma insulin [mU/L] × fasting plasma glucose [mmol/L])/22.5 (17). Insulin r esistance was defined as the highest quartile of homeostasis model assessment for insulin resistance. MetS components were defined according to the National Cholesterol Education Program panel definitions (18): WC >102 cm for men, or >88 cm for women; triglycerides ≥150 mg/dl; HDL <40 mg/dl for men, or <50 mg/dl for women; BP ≥130/85 mm Hg; and fasting glucose ≥110 mg/dl.
MA was defined as serum bicarbonate ≤23 mEq/L (4,5). Elevated AG was defined as >95th percentile for the total population (5). This cutoff was used previously by us (5) because its high value is more suggestive of pathology rather than minor electrolyte variations. The formulae for AG (in mEq/L) were: traditional AG = serum sodium − (serum bicarbonate + serum chloride); corrected AG = traditional AG + (4.4 − serum albumin [g/dL]) × 2.5 (19); and full AG = corrected AG + serum potassium (mEq/L) + ionized calcium (mEq/L) − serum phosphate (mEq/L) (3,19). Contributions of calcium and phosphate to the full AG formula were: ionized calcium (mEq/L) = 0.5 × (total calcium [mg/dl] × (0.8 [4 − serum albumin (g/dl)]/2); and serum phosphate (mEq/L) = 0.323 × serum phosphate (mg/dl) × 1.8 (3). Compared with traditional, the corrected and full AG calculations have been shown to be more sensitive disease risk indicators (3,20). For these analyses the full AG definition (3,19) was designated as the primary definition. AGMA was defined as the simultaneous presence of MA and elevated full AG.
Laboratory Methods
Glucose was measured by hexokinase enzymatic reaction with a Roche Cobas Mira (Roche Diagnostics, 9115 Hague Road, Indianapolis, IN 46250) in 2003–2004; a Roche/Hitachi 911 in 2005–2006; a Roche Modular P chemistry analyzer in 2007–2008; and Roche Cobas C311 analyzer in 2015–2018. Insulin was measured by radioimmunossay in 1999–2002; immunoenzymometric assay using a Tosah AIA-PACK in 2003–2004 and 2013–2018; by Merocodia insulin ELISA in 2005–2009; and by Elecsys 2010 insulin chemiluminescent immunoassay in 2010–2012. Triglycerides were measured by lipase/peroxidase enzymatic reaction with a Hitachi 717 and 912 in 2005–2006; and a Roche Modular P Chemistry Analyzer in 2007–2014. HDL was measured by heparin manganese method with a Hitachi 717 and Hitachi 912 in 1999–2006; by enzymatic method with a Roche Modular P chemistry analyzer in 2003–2012, and by Roche Modular P and Roche Cobas 6000 chemistry analyzers in 2013–2018.
Serum sodium, potassium, bicarbonate, chloride, calcium, phosphate, and creatinine were analyzed using a Hitachi Model 704 multichannel analyzer (Boehringer Mannheim Diagnostics, Indianapolis, IN) for years 1999–2002 and 2017–2018; and on a Beckman Synchron LX20 and Beckman UniCel DxC800 Synchron for years 2003–2016. Serum sodium, potassium, and chloride were measured by ion selective electrode.
Serum bicarbonate was measured as total CO2 by the phosphoenolpyruvate method. Serum phosphate was measured by reaction with ammonium molybdate. Serum calcium was measured by o-cresolphthalein complexone reaction for 1999–2002, and by ion selective electrode for 2003–2018. Serum albumin was measured by the Bromcresol purple dye method. Serum creatinine was measured by a modified kinetic Jaffe reaction for 1999–2016, and by the creatinase enzymatic method for 2017–2018. Urine creatinine was measured by Jaffe reaction on a Beckman CX3 for 1999–2006; and by enzymatic method on a Roche/Hitachi Modular P Chemistry Analyzer for 2007–2018.
Statistical Analysis
Observation weights on the basis of NHANES methodology (21) were applied in each analysis. Multivariable linear regression models were used to analyze continuous outcomes. Multivariable logistic regression models were used to examine categorical outcomes. Covariates selected a priori were age, sex, race/ethnicity, insurance type, household income, eGFR, diabetes mellitus, hypertension, coronary artery disease, and estimated net endogenous acid production (16). Two-tailed hypothesis tests with an alpha of 0.05 were used. Data were analyzed using Stata 13.1 (StataCorp, College Station, TX, USA).
Sensitivity and Subgroup Analyses
To understand populations at risk and mechanisms, we excluded patients with insulin resistance for one analysis. Because medications for BP (12) and diabetes (22) can affect serum bicarbonate levels, we examined a subgroup of patients without either diagnosis. To exclude kidney disease as a contributor, we examined a subgroup of participants with an eGFR ≥90 ml/min per 1.73 m2 and without albuminuria (urine albumin to creatinine ratio ≥30 mg/g). Because obesity may predispose to respiratory abnormalities affecting acid-base status (23), we examined associations of MetS features with outcomes after excluding a BMI ≥30 kg/m2.
This study was determined to be exempt from Northwell Institutional Review Board review due to the use of deidentified, publicly available data.
Results
Participant Characteristics
Characteristics of the WC (main) and MetS (fasting) cohorts are presented in Table 1 and Supplemental Table 1. In the respective cohorts, 49% and 49% of participants were women. The mean ages were 44.1 (SE 0.2) years and 43.8 (SE 0.3) years. CKD was more prevalent with greater WC (Table 1). Diabetes mellitus, hypertension, and coronary artery disease were also more prevalent with higher WC. Serum bicarbonate was lower, and AG values higher with greater WC (Table 1). Similar distributions of characteristics were observed across MetS features (Supplemental Table 1).
Table 1. -
Characteristics of 30,163 participants of National Health and Nutrition Examination Survey 1999–2018 by sextiles of waist circumference
| Characteristic |
Waist Circumference, cm |
| <81.2 |
81.2–89 |
89.1–95.6 |
95.7–102.5 |
102.6–111.5 |
>111.5 |
| Number |
5049 |
4968 |
4915 |
5219 |
5095 |
4917 |
| Age, yr |
35.6 (0.4) |
40.7 (0.3) |
44.4 (0.3) |
47.5 (0.3) |
48.6 (0.4) |
47.3 (0.3) |
| Women, % |
68.0 (0.9) |
56.1 (1) |
47.5 (1) |
41.4 (1) |
38.7 (0.9) |
40.6 (0.9) |
|
Race and ethnicity, %
|
| Mexican American |
6.5 (0.5) |
9.0 (0.7) |
10.9 (0.8) |
11.0 (0.9) |
11.1 (0.9) |
8.7 (0.8) |
| Non-Hispanic Black |
11.9 (0.6) |
9.3 (0.6) |
9.9 (0.6) |
9.4 (0.7) |
10.3 (0.7) |
13.8 (0.9) |
| Non-Hispanic White |
64.7 (1) |
66.0 (1) |
64.6 (1) |
67.6 (1) |
70.0 (1) |
70.0 (1) |
| Other Hispanic |
5.1 (0.5) |
6.6 (0.7) |
6.5 (0.6) |
6.2 (0.6) |
4.7 (0.5) |
4.0 (0.4) |
| Other race / multiracial |
11.8 (0.7) |
9.1 (0.6) |
8.0 (0.6) |
5.8 (0.5) |
4.0 (0.4) |
3.9 (0.4) |
|
Insurance, %
|
| Private |
60.2 (2) |
57.4 (2) |
55.8 (2) |
55.6 (1) |
52.9 (2) |
55.7 (2) |
| Medicare |
4.4 (0.5) |
7.2 (0.6) |
9.0 (0.7) |
11.6 (0.7) |
12.4 (0.8) |
10.5 (0.7) |
| Medicaid |
4.0 (0.5) |
4.1 (0.5) |
3.3 (0.5) |
3.7 (0.5) |
4.5 (0.5) |
5.3 (0.6) |
| Other government insurance |
9.3 (0.9) |
8.6 (0.9) |
11.4 (1) |
10.8 (1) |
11.8 (1) |
12.5 (1) |
| No insurance or missing |
22.0 (1) |
22.6 (1) |
20.6 (1) |
18.3 (1) |
18.3 (1) |
16.0 (1) |
|
Household income, %
|
| <$25,000 |
19.9 (1) |
21.0 (1) |
19.3 (1) |
19.0 (1) |
19.3 (1) |
19.4 (1) |
| $25,000 to <$45,000 |
21.7 (1) |
20.2 (1) |
23.4 (1) |
21.4 (1) |
23.3 (1) |
25.7 (1) |
| $45,000 to <$75,000 |
21.1 (1) |
23.1 (1) |
21.7 (1) |
27.1 (1) |
25.4 (2) |
24.2 (1) |
| ≥$75,000 |
37.2 (2) |
35.7 (2) |
35.7 (2) |
32.4 (2) |
32.1 (1) |
30.6 (2) |
| NEAP (mEq/d) |
57.9 (0.6) |
58.6 (0.6) |
57.9 (0.5) |
57.6 (0.5) |
59.8 (0.5) |
63.4 (0.6) |
|
eGFRa (ml/min per 1.73 m2), %
|
| ≥120 |
24.9 (0.9) |
15.7 (0.7) |
12.4 (0.8) |
9.0 (0.6) |
9.3 (0.7) |
11.8 (0.6) |
| 90–119 |
52.4 (1) |
55.4 (1) |
52.2 (1) |
50.6 (1) |
47.2 (1) |
50.2 (1) |
| 60–89 |
21.1 (1) |
25.8 (1) |
31.3 (1) |
35.1 (1) |
37.9 (1) |
32.9 (1) |
| 15–59 |
1.6 (0.2) |
3.1 (0.3) |
4.0 (0.4) |
5.3 (0.4) |
5.6 (0.4) |
5.1 (0.4) |
| Hypertension, % |
8.9 (0.6) |
15.2 (0.7) |
19.6 (0.8) |
26.5 (0.8) |
34.8 (1) |
45.2 (1) |
| Diabetes mellitus, % |
0.8 (0.1) |
2.6 (0.2) |
4.7 (0.4) |
6.1 (0.4) |
9.6 (0.6) |
15.8 (0.7) |
| Coronary artery disease, % |
0.8 (0.1) |
2.3 (0.3) |
2.6 (0.2) |
4.0 (0.4) |
5.0 (0.4) |
5.4 (0.4) |
| Serum bicarbonate, mEq/L |
25.0 (0.1) |
25.0 (0.1) |
25.0 (0.1) |
24.9 (0.1) |
24.7 (0.1) |
24.4 (0.1) |
| Traditional anion gap, mEq/L |
11.0 (0.1) |
11.0 (0.1) |
11.1 (0.1) |
11.1 (0.1) |
11.3 (0.1) |
11.4 (0.1) |
| Corrected anion gap, mEq/L |
9.9 (0.1) |
10.0 (0.1) |
10.1 (0.1) |
10.3 (0.1) |
10.5 (0.1) |
10.9 (0.1) |
| Full anion gap, mEq/L |
13.9 (0.1) |
14.1 (0.1) |
14.2 (0.1) |
14.4 (0.1) |
14.6 (0.1) |
15.1 (0.1) |
| Elevated traditional anion gap, % |
4.4 (0.7) |
4.4 (0.6) |
5.1 (0.6) |
5.3 (0.6) |
5.5 (0.7) |
6.6 (0.7) |
| Elevated corrected anion gap, % |
3.3 (0.4) |
3.9 (0.5) |
4.3 (0.5) |
4.7 (0.5) |
5.6 (0.7) |
8.2 (1) |
| Elevated full anion gap, % |
3.0 (0.4) |
4.0 (0.5) |
4.3 (0.5) |
5.1 (0.6) |
5.6 (0.7) |
8.1 (1) |
| Metabolic acidosis (≤23 mEq/L), % |
25.0 (1) |
23.2 (1) |
22.5 (1) |
26.2 (1) |
28.2 (1) |
34.3 (1) |
| Metabolic acidosis (<22 mEq/L), % |
6.0 (0.5) |
5.4 (0.7) |
6.1 (0.7) |
6.7 (0.6) |
8.1 (0.6) |
10.1 (0.7) |
| Corrected anion gap metabolic acidosis, % |
2.3 (0.4) |
2.3 (0.3) |
2.6 (0.3) |
3.0 (0.4) |
3.4 (0.4) |
5.3 (0.8) |
| Full anion gap metabolic acidosis, % |
1.9 (0.4) |
2.2 (0.4) |
2.6 (0.4) |
3.2 (0.5) |
3.6 (0.5) |
5.2 (0.8) |
Continuous variables reported as mean (standard error) and categorical variables as percent (standard error). NEAP, net endogenous acid production; mEq, milliequivalent.
aCalculated using CKD-EPI, Chronic Kidney Disease Epidemiology Collaboration creatinine equation.
Associations with Serum Bicarbonate and AG
Greater WC, MetS features, and BMI were associated with progressively lower serum bicarbonate and higher AG (Table 2). Compared with the lowest WC sextile, those in the highest had 1.07 (95% confidence interval [95% CI], 1.22–0.90) mEq/L lower serum bicarbonate, and 0.93 (95% CI, 0.75–1.10) mEq/L higher full AG (Table 3). When examined across WC deciles, a dose-dependent relationship was evident (Figure 1, A and B). Findings for MetS features were similar. Compared with participants with zero features, those with three, four, and five MetS features had serum bicarbonate values 1.06 (95% CI, 1.26–0.86), 1.43 (95% CI, 1.71–1.16), and 1.71 (95% CI, 2.41–1.01) mEq/L lower; and full AG values 0.82 (95% CI, 0.60–1.04), 1.16 (95% CI, 0.91–1.41), and 1.41 (95% CI, 0.61–2.22) mEq/L higher (Table 2). Associations of BMI, WC, and MetS features with corrected AG were similar (Supplemental Table 2). Strong associations with traditional AG were evident only for MetS (Supplemental Table 2).
Table 2. -
Associations of metabolic variables with serum bicarbonate and anion gap
| Predictor Variable |
Serum Bicarbonate, mEq/L |
Full Anion Gap, mEq/L |
| Delta |
95% Confidence Interval |
P Value |
Delta |
95% Confidence Interval |
P Value |
|
Waist circumference, cm
|
| <81.2 |
Ref. |
– |
– |
Ref. |
– |
– |
| 81.2–89 |
−0.16 |
−0.28 to −0.05 |
<0.01 |
0.11 |
−0.02 to 0.23 |
0.10 |
| 89.1–95.6 |
−0.33 |
−0.48 to −0.18 |
<0.001 |
0.18 |
0.04 to 0.32 |
0.02 |
| 95.7–102.5 |
−0.60 |
−0.73 to −0.46 |
<0.001 |
0.38 |
0.23 to 0.54 |
<0.001 |
| 102.6–111.5 |
−0.79 |
−0.92 to −0.65 |
<0.001 |
0.56 |
0.40 to 0.72 |
<0.001 |
| >111.5 |
−1.07 |
−1.22 to −0.90 |
<0.001 |
0.93 |
0.75 to 1.10 |
<0.001 |
|
Metabolic syndrome features
|
| 0 |
Ref. |
– |
– |
Ref. |
– |
– |
| 1 |
−0.51 |
−0.63 to −0.39 |
<0.001 |
0.34 |
0.20 to 0.47 |
<0.001 |
| 2 |
−0.87 |
−1.02 to −0.73 |
<0.001 |
0.65 |
0.50 to 0.80 |
<0.001 |
| 3 |
−1.06 |
−1.26 to −0.86 |
<0.001 |
0.82 |
0.60 to 1.04 |
<0.001 |
| 4 |
−1.43 |
−1.71 to −1.16 |
<0.001 |
1.16 |
0.91 to 1.41 |
<0.001 |
| 5 |
−1.71 |
−2.41 to −1.01 |
<0.001 |
1.41 |
0.61 to 2.22 |
<0.01 |
|
Body mass index, kg/m2
|
| <18.5 |
0.21 |
−0.15 to 0.56 |
0.25 |
0.17 |
−0.12 to 0.46 |
0.25 |
| 18.5 to <25 |
Ref. |
– |
– |
Ref. |
– |
– |
| 25 to <30 |
−0.41 |
−0.51 to −0.31 |
<0.001 |
0.24 |
0.14 to 0.34 |
<0.001 |
| 30 to <35 |
−0.72 |
−0.84 to −0.59 |
<0.001 |
0.53 |
0.39 to 0.67 |
<0.001 |
| 35 to <40 |
−0.82 |
−0.96 to −0.67 |
<0.001 |
0.79 |
0.59 to 0.98 |
<0.001 |
| >40 |
−0.88 |
−1.08 to −0.67 |
<0.001 |
1.12 |
0.89 to 1.35 |
<0.001 |
Multivariable model adjusted for age, sex, race and ethnicity, eGFR, estimated net endogenous acid production, baseline income, insurance status, hypertension, diabetes, and coronary artery disease. mEq, milliequivalent; Ref., reference.
Table 3. -
Associations of metabolic variables with metabolic acidosis
| Predictor Variable |
Metabolic Acidosis |
Anion Gap Metabolic Acidosis |
| Odds Ratio |
95% Confidence Interval |
P Value |
Odds Ratio |
95% Confidence Interval |
P Value |
|
Waist circumference, cm
|
| <81.2 |
Ref. |
– |
– |
Ref. |
– |
– |
| 81.2–89 |
1.04 |
0.91 to 1.19 |
0.52 |
1.17 |
0.81 to 1.69 |
0.40 |
| 89.1–95.6 |
1.12 |
0.96 to 1.30 |
0.16 |
1.41 |
0.93 to 2.13 |
0.11 |
| 95.7–102.5 |
1.51 |
1.30 to 1.75 |
<0.001 |
1.81 |
1.19 to 2.75 |
<0.01 |
| 102.6–111.5 |
1.72 |
1.49 to 1.99 |
<0.001 |
2.01 |
1.44 to 2.80 |
<0.001 |
| >111.5 |
2.26 |
1.96 to 2.62 |
<0.001 |
2.89 |
1.97 to 4.21 |
<0.001 |
|
Metabolic syndrome features
|
| 0 |
Ref. |
– |
– |
Ref. |
– |
– |
| 1 |
1.58 |
1.30 to 1.78 |
<0.001 |
1.18 |
0.69 to 2.01 |
0.54 |
| 2 |
2.08 |
1.69 to 2.32 |
<0.001 |
1.86 |
1.07 to 3.22 |
0.02 |
| 3 |
2.52 |
1.95 to 2.94 |
<0.001 |
2.23 |
1.20 to 4.15 |
0.01 |
| 4 |
3.05 |
2.16 to 3.82 |
<0.001 |
2.86 |
1.48 to 5.52 |
<0.01 |
| 5 |
2.26 |
1.01 to 5.06 |
0.04 |
a |
– |
– |
|
Body mass index, kg/m2
|
| <18.5 |
1.05 |
0.81 to 1.37 |
0.69 |
1.01 |
0.52 to 1.97 |
0.98 |
| 18.5 to <25 |
Ref. |
– |
– |
Ref. |
– |
– |
| 25 to <30 |
1.38 |
1.23 to 1.55 |
<0.001 |
1.27 |
1.07 to 1.51 |
<0.01 |
| 30 to <35 |
1.80 |
1.58 to 2.05 |
<0.001 |
1.63 |
1.30 to 2.04 |
<0.001 |
| 35 to <40 |
2.01 |
1.76 to 2.29 |
<0.001 |
2.14 |
1.61 to 2.85 |
<0.001 |
| >40 |
2.12 |
1.78 to 2.53 |
<0.001 |
2.01 |
1.38 to 2.93 |
<0.001 |
Multivariable model adjusted for age, sex, race and ethnicity, eGFR, estimated net endogenous acid production, baseline income, insurance status, hypertension, diabetes, and coronary artery disease. Ref., reference.
aNot calculable due to small sample.
;)
Figure 1.: Associations of waist circumference with serum bicarbonate and anion gap. Serum bicarbonate (A) and full anion gap (B) values by waist circumference deciles among the total cohort and the subgroup with an eGFR >90 ml/min per 1.73 m2 and without albuminuria. Model adjusted for age, sex, race/ethnicity, income, insurance status, eGFR, estimated net endogenous acid production, hypertension, diabetes, and coronary artery disease. Bars denote 95% confidence intervals. @ P<0.05; # P<0.01; *P<0.001.
Associations with Elevated AG
Graded positive associations were observed between each metabolic variable and odds of elevated AG (Supplemental Table 3). Compared with participants in the lowest WC sextile, those in the highest had odds ratios (ORs) of 1.50 (95% CI, 1.11 to 2.01), 2.29 (95% CI, 1.68 to 3.13], and 2.45 (95% CI, 1.81 to 3.31) for elevated traditional, corrected, and full AG. Similar trends were observed for MetS features, and for BMI. Compared with participants with zero MetS features, those with three and four had ORs for elevated corrected and full AG of 2.32 (95% CI, 1.27 to 4.26) and 2.59 (95% CI, 1.50 to 4.46) (Supplemental Table 3).
Associations with Metabolic Acidosis
Greater WC and MetS features were each associated with progressively greater odds of MA and AGMA (Table 3). Compared with the lowest WC sextile, those in the highest had ORs of 2.26 (95% CI, 1.96 to 2.62) and 2.89 (95% CI, 1.97 4.21) for MA and AGMA. When ORs for MA and AGMA were compared across WC deciles, a graded association was seen (Figure 2, A–D). Findings for MetS features were similar (Table 3). Compared with those with zero features, those with three, four, and five features had ORs for MA of 2.52 (95% CI, 1.95 to 2.94), 3.05 (95% CI, 2.16 to 3.82), and 2.26 (95% CI, 1.01 to 5.06). Those with three and four features had ORs for AGMA of 2.23 (95% CI, 1.20 to 4.15) and 2.86 (95% CI, 1.48 to 5.52). Higher BMI was also associated with a graded increase in odds of these outcomes (Table 3).
Figure 2.: Associations of waist circumference with metabolic acidosis and anion gap metabolic acidosis. Odds ratios for metabolic acidosis (A) and (B) and anion gap metabolic acidosis (C) and (D) by waist circumference decile among the total cohort (A) and (C), and the subgroup with an eGFR >90 ml/min per 1.73 m2 and without albuminuria (B) and (D). Model adjusted for age, sex, race/ethnicity, income, insurance status, eGFR, estimated net endogenous acid production, hypertension, diabetes, and coronary artery disease. Bars denote 95% confidence intervals. # P<0.01; *P<0.001.
Associations with BMI After Adjustment for Metabolic Disease
Because we hypothesized that metabolic disease, and not greater body mass per se, mediated these findings, we examined associations of BMI and each outcome after adjustment for either WC or MetS features. In all instances, the association of BMI with outcomes was either attenuated or absent after adjustment (Supplemental Table 4).
Subgroup Analyses
When insulin resistance was included as a covariate in the model, associations of WC with AG were greatly attenuated, whereas strong associations with serum bicarbonate and MA remained (Supplemental Table 5). After excluding participants with insulin resistance, there was a similar attenuation of association of WC with AG, but associations with bicarbonate remained (Supplemental Table 6). In contrast, associations of several individual MetS features with both AG and bicarbonate remained after excluding insulin resistance (Supplemental Table 6). When participants with diabetes and hypertension were excluded, WC was still strongly associated with lower bicarbonate, but not AG (Supplemental Table 7). In contrast, MetS features remained associated with both outcomes in this subgroup (Supplemental Table 7). When participants with an eGFR <90 ml/min per 1.73 m2 or albuminuria were excluded, associations of WC with outcomes were largely unaffected (Figure 1, A and B, and Figure 2, B and D).
Sensitivity Analyses
When the definition of MA was changed to a serum bicarbonate <22 mEq/L, findings were similar for WC, MetS features, and BMI (Table 4). Notably, for MetS features the point estimates were raised (Table 4) compared with analyses using the less conservative definition (Table 3). Those with three and four versus zero MetS features had ORs of 3.56 (95% CI, 2.53 to 5.02) and 5.44 (95% CI, 3.66 to 8.08). When the subgroup of individuals without evidence of CKD (eGFR ≥90 ml/min per 1.73 m2 and without albuminuria) were analyzed, estimates were similar (Table 4). After exclusion of participants with obesity (BMI ≥30 kg/m2), associations of MetS features with outcomes were maintained (Supplemental Table 8). Finally, to further define the association of the MA and AG outcomes, we examined associations of WC with non-AGMA, defined as serum bicarbonate <22 mEq/L and AG ≤ 14.4 mEq/L (the population mean), after excluding eGFR <90 ml/min per 1.73 m2 or albuminuria. This was compared with AGMA also defined using serum bicarbonate <22 mEq/L. This analysis revealed a clear association of greater WC with AGMA, but not with non-AGMA (Supplemental Table 9).
Table 4. -
Association
a of metabolic variables with metabolic acidosis (serum bicarbonate <22 mEq/L)
| Predictor Variable |
Full Cohorts (Waist Circumference Cohort n=31,163; Metabolic Syndrome Cohort n=12,860) |
Excluding CKDb
(Waist Circumference Cohort n=12,765;
Metabolic Syndrome Cohort n=5143) |
| Odds Ratio |
95% Confidence Interval |
P Value |
Odds Ratio |
95% Confidence Interval |
P Value |
|
Waist circumference, cm
|
| <81.2 |
Ref. |
– |
– |
Ref. |
– |
– |
| 81.2–89 |
1.05 |
0.80 to 1.38 |
0.72 |
1.00 |
0.69 to 1.43 |
0.98 |
| 89.1–95.6 |
1.36 |
1.03 to 1.80 |
0.03 |
1.35 |
0.93 to 1.97 |
<0.01 |
| 95.7–102.5 |
1.65 |
1.28 to 2.13 |
<0.001 |
1.86 |
1.31 to 2.66 |
<0.001 |
| 102.6–111.5 |
2.07 |
1.64 to 2.61 |
<0.001 |
2.41 |
1.75 to 3.31 |
<0.001 |
| >111.5 |
2.48 |
1.95 to 3.14 |
<0.001 |
2.57 |
1.88 to 3.52 |
<0.001 |
|
Metabolic syndrome features
|
| 0 |
Ref. |
– |
– |
Ref. |
– |
– |
| 1 |
1.78 |
1.35 to 2.35 |
<0.001 |
1.82 |
1.27 to 2.63 |
<0.01 |
| 2 |
2.95 |
2.35 to 3.71 |
<0.001 |
3.11 |
2.26 to 4.27 |
<0.01 |
| 3 |
3.56 |
2.53 to 5.02 |
<0.001 |
4.21 |
2.57 to 6.90 |
<0.001 |
| 4 |
5.44 |
3.66 to 8.08 |
<0.001 |
7.04 |
3.61 to 13.72 |
<0.001 |
| 5 |
4.23 |
1.15 to 19.17 |
0.03 |
3.03 |
0.33 to 27.73 |
0.08 |
|
Body mass index, kg/m2
|
| <18.5 |
1.24 |
0.77 to 1.98 |
0.37 |
0.92 |
0.39 to 2.19 |
0.85 |
| 18.5 to <25 |
Ref. |
– |
– |
Ref. |
– |
– |
| 25 to <30 |
1.49 |
1.23 to 1.80 |
<0.001 |
1.59 |
1.22 to 2.07 |
<0.001 |
| 30 to <35 |
1.91 |
1.55 to 2.36 |
<0.001 |
2.18 |
1.64 to 2.88 |
<0.001 |
| 35 to <40 |
2.19 |
1.72 to 2.80 |
<0.001 |
2.66 |
1.89 to 3.73 |
<0.001 |
| >40 |
2.54 |
1.92 to 3.35 |
<0.001 |
2.59 |
1.71 to 3.90 |
<0.001 |
Ref, reference.
aMultivariable model adjusted for age, sex, race and ethnicity, eGFR, estimated net endogenous acid production, baseline income, insurance status, hypertension, and coronary artery disease.
bExcluding participants with eGFR <90 ml/min per 1.73 m3 or albuminuria (albumin/creatinine ratio ≥30 mg/g).
Discussion
This analysis of a large and representative US population demonstrates strong associations of metabolic disease, including abdominal obesity and the components of MetS, with MA and AG elevations. These abnormalities are not expected in a population largely without CKD (24). However, in our analysis, exclusion of participants with any decline in eGFR (<90 ml/min per 1.73 m2) or albuminuria (albumin to creatinine ratio ≥30 mg/g) preserved results. These findings confirm our primary hypothesis that metabolic disease, and not greater body mass per se, is closely related to this pathology. It is notable that after excluding participants with insulin resistance, associations of MetS features with outcomes were still strong, and exclusion of diabetes and hypertension did not affect findings. This suggests metabolic disease states are broadly characterized by MA and anion elevations. This abnormality is not unique to obesity, diabetes, hyperlipemia, or insulin resistance.
We demonstrated in several analyses an association of low bicarbonate with high AG values as the combined AGMA outcome. This outcome was prevalent among those with the highest WC and higher numbers of MetS features. This association suggests an endogenous source of MA and unmeasured anions in those with advanced metabolic disease. It is widely shown that patients with MetS features have higher fasting and postprandial (25) blood levels of the lactate anion (2526–27). Overtly high lactate and associated MA (28) are classically seen with exercise (28) and severe illness (29), conditions of mismatched and impaired oxidative capacity. However, lactate levels are mildly but chronically (25) elevated in obesity (25), nonalcoholic fatty liver disease (NAFLD) (30), insulin resistance (31), diabetes (25), hypertension (26), and hyperlipemia (27). Impaired oxidative capacity is also seen in these disease states in skeletal muscle, liver, and adipose tissue (3233–34). Although the lactate anion is a base equivalent (35), its formation in response to energy demand is accompanied by MA if oxidative capacity is insufficient (28,36). For acid-base homeostasis to prevail, oxidative capacity must improve, and organic acid must be cleared primarily by the liver (35,36). Patients with NAFLD and nonalcoholic steatohepatitis, comorbidities of MetS (37), have evidence of organic acid accumulation in liver tissue (38). More specifically, a recent study showed that lactate levels in NAFLD are higher in liver than in serum (38). This suggests portal vein lactate concentrations are elevated, and the liver’s capacity to clear it may be insufficient to prevent systemic exposure. The authors presented evidence that acetylation of lactate dehydrogenase B in hepatocytes may be responsible for this change in clearance. Adipose tissue produces lactate (31), and rates of glycolysis and lactate production are especially high in visceral fat (39), which drains into the portal system. As such, it may be that organic acid is overproduced and insufficiently cleared in patients with MetS.
Current evidence suggests that acid retention may be possible with normal kidney function. Wesson’s study using a rat model demonstrated that non-CKD and CKD rats both retain acid in response to a 7-day acid load, and this was measurable in kidney tissue (40). In noncontrolled and observational studies in humans, there is indirect evidence of acid retention in the presence of normal eGFR. Base supplementation given to healthy older adults on acidogenic diets with normal eGFR, and normal bicarbonate, reduces markers of bone and muscle injury (9,10). Among a US population, estimated net endogenous acid production was associated with lower bicarbonate levels in adults without CKD (14). However, to our knowledge, there have been no controlled studies of acid retention in patients with metabolic diseases and normal kidney function.
Endogenous acid from the diet may also affect serum bicarbonate and AG in certain patient groups (14,20). Although we adjusted for this factor, residual confounding is likely because we relied on dietary recall data, and indirect measures of dietary acid (16). Impaired absorption of gastrointestinal alkali has been described in obesity (41) and diabetes (42), which would reduce serum bicarbonate. Finally, impaired acid excretion by the kidneys may explain our findings for serum bicarbonate. Tubule dysfunction manifesting as impaired ammoniagenesis, or wasting of organic base equivalents in urine, might occur in metabolic disease states before GFR decline. Uric acid stone formers have both insulin resistance and evidence of impaired ammoniagenesis (41), and similar abnormalities are seen in obesity (41) and diabetes (42). However, diet-controlled studies have shown that patients with obesity and diabetes have higher, not lower, total daily net acid excretion, despite generally lower ammonium excretion (41,42).
Current evidence suggests that ventilatory changes related to obesity are an unlikely explanation for our findings. Greater body mass and abdominal obesity are risk factors for hypo- rather than hyperventilation, as in obesity hypoventilation syndrome (23). In this condition, carbon dioxide is retained and ultimately higher, not lower, serum bicarbonate levels develop (23). Additionally, a prior study that collected arterial blood gases in patients with class 3 obesity and without hypoventilation syndrome showed that arterial carbon dioxide is positively associated with BMI and WC (43). This suggests subclinical hypoventilation may develop in obesity, and mild renal bicarbonate retention, rather than loss, would be expected in response (44). Additionally, we demonstrated that associations of MetS with outcomes remained after excluding obesity.
The limitations of this study include its cross-sectional design. These findings show association, but do not demonstrate causation. A major limitation is the absence of comprehensive acid-base analyses. Without arterial blood gases we cannot determine whether these findings are mediated by respiratory disease, or if arterial pH is truly low. Our findings may represent a fully compensated defect, and not acidemia. Also, although we hypothesize that endogenous organic acid may be a contributor to our findings, measurements of lactate were not available for this dataset. Finally, although we adjusted for dietary acid, this was on the basis of indirect data. A diet-controlled study with urine measurements assuring dietary compliance would be needed to definitively rule out residual confounding by dietary acid as an explanation for our findings. Furthermore, controlled studies that better account for organic acids in the urine may reveal a potential role of the kidneys in our findings. The strengths of our study include the use of a large representative dataset and the consistency of our findings across subgroup and sensitivity analyses.
These results suggest that comprehensive acid-base analyses would be informative in patients with metabolic disease, both to determine tissue pH and to better characterize disease mechanisms. Truly reduced pH within tissue, for example within muscle (40), would certainly urge greater consideration of the role of MA in cardiometabolic disease pathophysiology. Acid retention causes kidney function decline (7,8), is implicated in hypertension (45,46), and associates with greater mortality (3). If acid were retained in this patient group, both supplemental and dietary alkali might be studied as preventive measures.
Disclosures
J. Kane reports having an advisory or leadership role with the Eli Lilly obesity advisory board (will be serving from August 2022). M.K. Abramowitz reports having consultancy agreements with Tricida; and reports having an ownership interest in Aethlon Medical, Inc. All remaining authors have nothing to disclose.
Funding
None.
Author Contributions
M. Abramowitz, J. Kane, and D. Lambert conceptualized the study; D. Lambert and A. Slaton were responsible for the data curation; D. Lambert and A. Slaton were responsible for the formal analysis; M. Abramowitz, J. Kane, and D. Lambert were responsible for the methodology; D. Lambert wrote the original draft; M. Abramowitz, J. Kane, D. Lambert, and A. Slaton reviewed and edited the manuscript.
Supplemental Material
This article contains the following supplemental material online at http://kidney360.asnjournals.org/lookup/suppl/doi:10.34067/KID.0002402022/-/DCSupplemental.
Supplemental Table 1. Characteristics of 12,860 participants of NHANES 1999–2018 by number of metabolic syndrome features.
Supplemental Table 2. Associations of metabolic variables with traditional and corrected anion gap.
Supplemental Table 3. Associations of metabolic variables with elevated anion gap.
Supplemental Table 4. Associations of body mass index with outcomes after adjustment for metabolic variables.
Supplemental Table 5. Associations of waist circumference with outcomes after adjustment for insulin resistance.
Supplemental Table 6. Associations of metabolic variables with outcomes, excluding participants with insulin resistance (HOMA-IR).
Supplemental Table 7. Associations of metabolic variables with outcomes, excluding participants with diabetes and hypertension.
Supplemental Table 8. Association of metabolic syndrome features with outcomes, excluding participants with obesity.
Supplemental Table 9. Associations of waist circumference with nonanion gap metabolic acidosis and anion gap metabolic acidosis after excluding CKD.
References
1. Kraut JA, Madias NE: Metabolic acidosis of CKD: An update. Am J Kidney Dis 67: 307–317, 2016
https://doi.org/10.1053/j.ajkd.2015.08.028
2. Wesson DE, Simoni J, Broglio K, Sheather S: Acid retention accompanies reduced GFR in humans and increases plasma levels of endothelin and aldosterone. Am J Physiol Renal Physiol 300: F830–F837, 2011
https://doi.org/10.1152/ajprenal.00587.2010
3. Abramowitz MK, Hostetter TH, Melamed ML: The serum anion gap is altered in early kidney disease and associates with mortality. Kidney Int 82: 701–709, 2012
https://doi.org/10.1038/ki.2012.196
4. Lambert DC, Abramowitz MK: Obesity and the risk of low bicarbonate: A cohort study. Kidney Med 3: 498–506.e1, 2021
https://doi.org/10.1016/j.xkme.2021.02.006
5. Lambert DC, Abramowitz MK: Obesity, anion accumulation, and anion gap metabolic acidosis: A cohort study. Kidney360 2: 1706–1715, 2021
https://doi.org/10.34067/KID.0003562021
6. DeFronzo RA, Beckles AD: Glucose intolerance following chronic metabolic acidosis in man. Am J Physiol 236: E328–E334, 1979
https://doi.org/10.1152/ajpendo.1979.236.4.E328
7. Shah SN, Abramowitz M, Hostetter TH, Melamed ML: Serum bicarbonate levels and the progression of kidney disease: A cohort study. Am J Kidney Dis 54: 270–277, 2009
https://doi.org/10.1053/j.ajkd.2009.02.014
8. Raphael KL, Wei G, Baird BC, Greene T, Beddhu S: Higher serum bicarbonate levels within the normal range are associated with better survival and renal outcomes in African Americans. Kidney Int 79: 356–362, 2011
https://doi.org/10.1038/ki.2010.388
9. Sebastian A, Harris ST, Ottaway JH, Todd KM, Morris Jr RC: Improved mineral balance and skeletal metabolism in postmenopausal women treated with potassium bicarbonate. N Engl J Med 330: 1776–1781, 1994
https://doi.org/10.1056/NEJM199406233302502
10. Frassetto L, Morris Jr RC, Sebastian A: Potassium bicarbonate reduces urinary nitrogen excretion in postmenopausal women. J Clin Endocrinol Metab 82: 254–259, 1997
https://doi.org/10.1210/jcem.82.1.3663
11. Yenchek R, Ix JH, Rifkin DE, Shlipak MG, Sarnak MJ, Garcia M, Patel KV, Satterfield S, Harris TB, Newman AB, Fried LF, Health A; Health, Aging, and Body Composition Study: Association of serum bicarbonate with incident functional limitation in older adults. Clin J Am Soc Nephrol 9: 2111–2116, 2014
https://doi.org/10.2215/CJN.05480614
12. Raphael KL, Zhang Y, Ying J, Greene T: Prevalence of and risk factors for reduced serum bicarbonate in chronic kidney disease. Nephrology (Carlton) 19: 648–654, 2014
https://doi.org/10.1111/nep.12315
13. Farwell WR, Taylor EN: Serum bicarbonate, anion gap and insulin resistance in the National Health and Nutrition Examination Survey. Diabet Med 25: 798–804, 2008
https://doi.org/10.1111/j.1464-5491.2008.02471.x
14. Amodu A, Abramowitz MK: Dietary acid, age, and serum bicarbonate levels among adults in the United States. Clin J Am Soc Nephrol 8: 2034–2042, 2013
https://doi.org/10.2215/CJN.03600413
15. Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro 3rd AF, Feldman HI, Kusek JW, Eggers P, Van Lente F, Greene T, Coresh J; CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration): A new equation to estimate glomerular filtration rate. Ann Intern Med 150: 604–612, 2009
https://doi.org/10.7326/0003-4819-150-9-200905050-00006
16. Frassetto LA, Todd KM, Morris Jr RC, Sebastian A: Estimation of net endogenous noncarbonic acid production in humans from diet potassium and protein contents. Am J Clin Nutr 68: 576–583, 1998
https://doi.org/10.1093/ajcn/68.3.576
17. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC: Homeostasis model assessment: Insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 28: 412–419, 1985
https://doi.org/10.1007/BF00280883
18. Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults: Executive Summary of The Third Report of The National Cholesterol Education Program (NCEP) Expert Panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III). JAMA 285: 2486–2497, 2001
https://doi.org/10.1001/jama.285.19.2486
19. Figge J, Jabor A, Kazda A, Fencl V: Anion gap and hypoalbuminemia. Crit Care Med 26: 1807–1810, 1998
https://doi.org/10.1097/00003246-199811000-00019
20. Banerjee T, Crews DC, Wesson DE, McCulloch CE, Johansen KL, Saydah S, Rios Burrows N, Saran R, Gillespie B, Bragg-Gresham J, Powe NR: Elevated serum anion gap in adults with moderate chronic kidney disease increases risk for progression to end-stage renal disease. Am J Physiol Renal Physiol 316: F1244–F1253, 2019
https://doi.org/10.1152/ajprenal.00496.2018
21. Centers for Disease Control and Prevention (CDC). National Center for Health Statistics (NCHS). National Health and Nutrition Examination Survey Data. Hyattsville, MD: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. Available at:
https://www.cdc.gov/nchs/nhanes/index.htm. Accessed August 1, 2020.
22. Gardner J, Tuttle K, Raphael KL: Influence of medications containing acid salts on serum bicarbonate in CKD. Kidney360 1: 330–336, 2020
https://doi.org/10.34067/KID.0000532019
23. Mokhlesi B: Obesity hypoventilation syndrome: A state-of-the-art review. Respir Care 55: 1347–1362, discussion 1363–1365, 2010
24. Moranne O, Froissart M, Rossert J, Gauci C, Boffa JJ, Haymann JP, M’rad MB, Jacquot C, Houillier P, Stengel B, Fouqueray B; NephroTest Study Group: Timing of onset of CKD-related metabolic complications. J Am Soc Nephrol 20: 164–171, 2009
https://doi.org/10.1681/ASN.2008020159
25. Chen YD, Varasteh BB, Reaven GM: Plasma lactate concentration in obesity and type 2 diabetes. Diabete Metab 19: 348–354, 1993
26. Lian H, Zhuo SQ, Tian XT, Liu FC: Increased plasma lactate level is associated with subclinical cardiovascular damage in patient with non-dipping hypertension. Clin Exp Hypertens 38: 541–544, 2016
https://doi.org/10.3109/10641963.2016.1174247
27. Yin X, Subramanian S, Willinger CM, Chen G, Juhasz P, Courchesne P, Chen BH, Li X, Hwang SJ, Fox CS, O’Donnell CJ, Muntendam P, Fuster V, Bobeldijk-Pastorova I, Sookoian SC, Pirola CJ, Gordon N, Adourian A, Larson MG, Levy D: Metabolite signatures of metabolic risk factors and their longitudinal changes. J Clin Endocrinol Metab 101: 1779–1789, 2016
https://doi.org/10.1210/jc.2015-2555
28. Robergs RA, Ghiasvand F, Parker D: Biochemistry of exercise-induced metabolic acidosis. Am J Physiol Regul Integr Comp Physiol 287: R502–R516, 2004
https://doi.org/10.1152/ajpregu.00114.2004
29. Suetrong B, Walley KR: Lactic acidosis in sepsis: It’s not all anaerobic: Implications for diagnosis and management. Chest 149: 252–261, 2016
https://doi.org/10.1378/chest.15-1703
30. Kalhan SC, Guo L, Edmison J, Dasarathy S, McCullough AJ, Hanson RW, Milburn M: Plasma metabolomic profile in nonalcoholic fatty liver disease. Metabolism 60: 404–413, 2011
https://doi.org/10.1016/j.metabol.2010.03.006
31. DiGirolamo M, Newby FD, Lovejoy J: Lactate production in adipose tissue: A regulated function with extra-adipose implications. FASEB J 6: 2405–2412, 1992
https://doi.org/10.1096/fasebj.6.7.1563593
32. Heinonen S, Jokinen R, Rissanen A, Pietiläinen KH: White adipose tissue mitochondrial metabolism in health and in obesity. Obes Rev 21: e12958, 2020
https://doi.org/10.1111/obr.12958
33. Kelley DE, He J, Menshikova EV, Ritov VB: Dysfunction of mitochondria in human skeletal muscle in type 2 diabetes. Diabetes 51: 2944–2950, 2002
https://doi.org/10.2337/diabetes.51.10.2944
34. Spahr L, Negro F, Leandro G, Marinescu O, Goodman KJ, Rubbia-Brandt L, Jordan M, Hadengue A: Impaired hepatic mitochondrial oxidation using the
13C-methionine breath test in patients with macrovesicular steatosis and patients with cirrhosis. Med Sci Monit 9: CR6–CR11, 2003
35. Madias NE: Lactic acidosis. Kidney Int 29: 752–774, 1986
https://doi.org/10.1038/ki.1986.62
36. Adeva-Andany M, López-Ojén M, Funcasta-Calderón R, Ameneiros-Rodríguez E, Donapetry-García C, Vila-Altesor M, Rodríguez-Seijas J: Comprehensive review on lactate metabolism in human health. Mitochondrion 17: 76–100, 2014
https://doi.org/10.1016/j.mito.2014.05.007
37. Asrih M, Jornayvaz FR: Metabolic syndrome and nonalcoholic fatty liver disease: Is insulin resistance the link? Mol Cell Endocrinol 418: 55–65, 2015
https://doi.org/10.1016/j.mce.2015.02.018
38. Wang T, Chen K, Yao W, Zheng R, He Q, Xia J, Li J, Shao Y, Zhang L, Huang L, Qin L, Xu M, Zhang Z, Pan D, Li Z, Huang F: Acetylation of lactate dehydrogenase B drives NAFLD progression by impairing lactate clearance. J Hepatol 74: 1038–1052, 2021
https://doi.org/10.1016/j.jhep.2020.11.028
39. Newby FD, Sykes MN, DiGirolamo M: Regional differences in adipocyte lactate production from glucose. Am J Physiol 255: E716–E722, 1988
https://doi.org/10.1152/ajpendo.1988.255.5.E716
40. Wesson DE, Pruszynski J, Cai W, Simoni J: Acid retention with reduced glomerular filtration rate increases urine biomarkers of kidney and bone injury. Kidney Int 91: 914–927, 2017
https://doi.org/10.1016/j.kint.2016.10.023
41. Bobulescu IA, Park SK, Xu LHR, Blanco F, Poindexter J, Adams-Huet B, Davidson TL, Sakhaee K, Maalouf NM, Moe OW: Net acid excretion and urinary organic anions in idiopathic uric acid nephrolithiasis. Clin J Am Soc Nephrol 14: 411–420, 2019
https://doi.org/10.2215/CJN.10420818
42. Maalouf NM, Cameron MA, Moe OW, Sakhaee K: Metabolic basis for low urine pH in type 2 diabetes. Clin J Am Soc Nephrol 5: 1277–1281, 2010
https://doi.org/10.2215/CJN.08331109
43. Gabrielsen AM, Lund MB, Kongerud J, Viken KE, Røislien J, Hjelmesaeth J: The relationship between anthropometric measures, blood gases, and lung function in morbidly obese White subjects. Obes Surg 21: 485–491, 2011
https://doi.org/10.1007/s11695-010-0306-9
44. Cogan MG: Chronic hypercapnia stimulates proximal bicarbonate reabsorption in the rat. J Clin Invest 74: 1942–1947, 1984
https://doi.org/10.1172/JCI111614
45. Juraschek SP, Bower JK, Selvin E, Subash Shantha GP, Hoogeveen RC, Ballantyne CM, Young JH: Plasma lactate and incident hypertension in the atherosclerosis risk in communities study. Am J Hypertens 28: 216–224, 2015
https://doi.org/10.1093/ajh/hpu117
46. Aryal D, Roy T, Chamcheu JC, Jackson KE: Chronic metabolic acidosis elicits hypertension via upregulation of intrarenal angiotensin II and induction of oxidative stress. Antioxidants 10: 2, 2020
https://doi.org/10.3390/antiox10010002
