Association of betaine with blood pressure in dialysis patients

Abstract

Mechanisms underlying elevated blood pressure in dialysis patients are complex as a variety of non‐traditional factors are involved. We sought to explore the association of circulating betaine, a compound widely distributed in food, with blood pressure in dialysis patients. We used baseline data of an ongoing cohort study involving patients on hemodialysis. Plasma betaine was measured by high performance liquid chromatography in 327 subjects. Blood pressure level was determined by intradialytic ambulatory blood pressure monitoring. The mean age of the patients was 52.6 ± 11.9 years, and 58.4% were male. Average interdialytic ambulatory systolic and diastolic blood pressure were 138.4 ± 22.7 mm Hg and 84.4 ± 12.5 mm Hg, respectively. Mean plasma betaine level was 37.6 μmol/L. Multiple linear regression analysis revealed significant associations of betaine with both systolic blood pressure (β = −3.66, P = .003) and diastolic blood pressure (β = −2.00, P = .004). The associations persisted even after extensive adjustment for cardiovascular covariates. Subgroup analysis revealed that the association between betaine and blood pressure was mainly limited to female patients. Our data suggest that alteration of circulating betaine possibly contributes to blood pressure regulation in these patients.

1. INTRODUCTION

Patients with chronic kidney disease have increased risk of cardiovascular morbidity and mortality.1 Annual death rate in end‐stage renal disease patients on maintenance hemodialysis is up to 15%‐20%, with cardiovascular events accounting for the largest proportion.2 As a classical traditional cardiovascular risk factor, hypertension is highly prevalent in dialysis patients and is associated with adverse outcomes.3 Despite the fact that multiple interventions targeting blood pressure reduction are available today, hypertension in dialysis patients remains poorly controlled.4 This is partly due to the fact that the underlying mechanisms contributing to the elevated blood pressure in dialysis patients are complex and not fully understood. Exploring novel factors associated with blood pressure regulation in these patients may therefore provide further treatment implications.

Betaine, a zwitterionic quaternary ammonium compound with a molecular weight of 117, is abundant in many foods, including wheat, spinach, and beets.5 It can also be endogenously synthesized in the kidney and liver from choline by the mitochondrial enzyme, choline dehydrogenase.6 In humans, it functions mainly as a methyl donor and is thus involved in homocysteine metabolism.7 A large body of evidence has shown that betaine is an important nutrient for maintaining human health.

Prior data has implicated a possible role of betaine in blood pressure regulation. As early as 1951, subjects with hypertension experienced transient reduced blood pressure in the combination of betaine and guanidinoacetate.8 In a study by Schwab and colleagues,9 supplementation of betaine for 12 weeks decreased diastolic blood pressure (DBP) in obese patients, although the difference between supplementation and the control group did not reach statistical significance. In a large population‐based study, plasma betaine was inversely associated with blood pressure level in over 7000 mid‐aged and elderly subjects.10 These early findings lead us to ask the question of whether circulation betaine is associated with blood pressure in patients with renal dysfunction, given that the patients suffer from a great burden of hypertension. However, little data is available regarding this issue. Therefore, in the current study, we aimed to explore the association between circulating betaine concentration and blood pressure level in patients on maintenance hemodialysis.

2. METHODS

2.1. Study population

This study is a cross‐sectional analysis using baseline data from an ongoing cohort study. The inclusion and exclusion criteria have already been reported elsewhere.11 Briefly, patients on maintenance hemodialysis (4 hours, 3 times a week) aged 18‐80 years were recruited. Those with malignant hypertension (systolic blood pressure ≥180 mm Hg or DBP) ≥110 mm Hg) or any other situation making them unsuitable for study evaluation were excluded. The cohort study recruited a total of 368 participants from 6 tertiary hospitals in East China from July 2015 to July 2016. All subjects provided written informed consent. The Institutional Ethical Committee of the Second Affiliated Hospital of Nanjing Medical University approved the study.

2.2. Blood pressure

Blood pressure level was determined by midweek interdialytic ambulatory blood pressure monitoring. Ambulatory blood pressure has been demonstrated to be a more potent prognostic indicator than dialysis‐unit blood pressures.3 Although 44‐hour monitoring covering the whole interdialytic period would give more detailed information, we did not adopt this strategy in the current study since very long monitoring durations usually lead to more complaints of discomfort and less compliance from the patients. The current study's monitoring program began in the morning of a midweek non‐dialysis day and was terminated before the next dialysis session using a SpaceLabs 90217 monitor. We have previously shown that 24‐hour ambulatory blood pressure monitoring starting from the non‐dialysis day agrees well with 44‐hour monitoring and therefore can be used as a good surrogate for 44‐hour monitoring.12 Patients were instructed to keep their arm still at measurement and follow their daily activity. Blood pressure was measured at 20‐minute intervals in the daytime (6:00 am‐10:00 pm) and at 30‐minute intervals in the nighttime (10:00 pm‐6:00 am). For those with <6 readings (n = 6), predialysis blood pressure averaged over 2 weeks before enrollment (6 dialysis session) was used instead.

2.3. Routine laboratory tests

All blood samples were obtained from vascular access before dialysis treatment for routine biochemical tests, including hemoglobin, albumin, total cholesterol, triglyceride, high‐density lipoprotein cholesterol (HDL‐C), low‐density lipoprotein cholesterol (LDL‐C), calcium, phosphorus, and parathyroid hormone. For patients who did not have these tests at inclusion, we recorded their most recent results (within 3 months) from their local center.

2.4. Plasma betaine measurement

Plasma samples were obtained and sent to the Laboratory for Cardiovascular disease at Peking University (Beijing). Betaine level was measured by high performance liquid chromatography‐mass spectrometry. 20 μL plasma were aliquoted to a 1.5 mL Axygen tube and mixed with 80 μL of 10 μmol/L internal standard comprised of d11‐betaine in methanol. D11‐Betaine was purchased from Cambridge Isotope Laboratories. All other reagents were obtained from Sigma‐Aldrich, unless otherwise specified. Protein in the samples was precipitated by vortexing for 1 minute, the supernatant was then recovered following centrifugation at 20 000 g at 4°C for 10 minutes. Supernatants (70 μL) were analyzed by injection onto a silica column (2.0*150 mm, Luna 5 μ Silica 100A) at a flow rate of 0.4 mL/min using a LC‐20AD Shimadazu pump system and SIL‐20AXR autosample interfaced with an API 5500Q‐TRAP mass spectrometer (AB SCIEX). A discontinuous gradient was generated to resolve betaine by mixing solvent A (0.1% propanoic acid in water) with solvent B (0.1% acetic acid in methanol) at different ratios starting from 2% B linearly to 95% B over 5 minutes, hold for 1 minute, and then back to 2% B. Betaine was monitored using electrospray ionization in positive‐ion mode with multiple reaction monitoring (MRM) of precursor and characteristic product‐ion transitions of betaine at m/z 118→57.7, d11‐betaine at m/z 129→65.9 respectively. In order to get the precise betaine concentration, standard curve was generated. 20 μL various concentration standards (0‐100 μmol/L) process the same procedures. Standard curves were acceptable when coefficient of determination (R2) reached 0.99 (Figure S1). The accuracy of the measurement at several concentration levels were presented in Table S1.

2.5. Statistical analyses

Data was expressed as mean ± standard deviation or median (interquartile range) for numerical variables and counts (%) for categorical variables. Numerical variables with a skewed distribution were logarithm transformed in further analysis if needed. Comparison of plasma betaine level between subgroups was performed using Mann‐Whitney U test. Associations between betaine and blood pressures were determined by multiple linear regression analyses, with or without adjustment for several covariates, including age, gender, body mass index, smoking status, diabetes mellitus, previous history of cardiovascular disease, use of antihypertensives, use of erythropoietin, total and high‐density lipoprotein cholesterol, and interdialytic weight gain. All analyses were performed using SPSS 19.0. A 2‐tailed P value <.05 was considered statistically significant.

3. RESULTS

After the exclusion of 41 patients without sufficient blood samples, a total of 327 subjects were included in the present analysis. Demographic and clinical characteristics of the study patients are depicted in Table 1. The mean age of the patients was 52.6 ± 11.9 years, 58.4% were male, 21.7% were current smokers, and 16.8% had diabetes mellitus. Only 10% subjects had a previous history of cardiovascular disease. More than two‐thirds of the patients took antihypertensive drugs. Routine laboratory tests revealed characteristics of a typical dialysis population.

Table 1.

Patients' characteristics

	Statistical descriptions
Age, y	52.6 ± 11.9
Male	178 (54.4%)
BMI, kg/m2	22.0 ± 3.4
Current smokers	71 (21.7%)
Dialysis duration, mo	58.0 (31.0‐91.0)
Interdialytic weight gain, kg	2.4 (1.9‐3.0)
Diabetes mellitus	55 (16.8%)
Previous history of CVD	33 (10.1%)
Use of antihypertensives	227 (69.4%)
Use of statins	14 (4.3%)
Use of EPO	291 (89.0%)
Hemoglobin, g/L	108.6 ± 16.4
Albumin, g/L	39.4 ± 4.4
Total cholesterol, mmol/L	3.96 ± 0.94
Triglycerides, mmol/L	1.64 (1.16‐2.30)
HDL‐C, mmol/L	1.00 ± 0.29
LDL‐C, mmol/L	2.08 ± 0.60
Phosphorus, mmol/L	1.82 ± 0.51
Calcium, mmol/L	2.29 ± 0.26
Parathyroid hormone, pg/mL	289.6 (112.6‐569.4)
Systolic blood pressure, mm Hg	138.4 ± 22.7
Diastolic blood pressure, mm Hg	84.4 ± 12.5
Heart rate, bpm	77.0 ± 9.6

BMI, body mass index; bpm, beats per minute; CVD, cardiovascular disease; EPO, erythropoietin; HDL‐C, high‐density lipoprotein cholesterol; LDL‐C, low‐density lipoprotein cholesterol.

The median (interquartile range) plasma betaine concentration in the patients was 37.6 (30.1‐47.6) μmol/L. The distribution of betaine level was shown in Figure (A). We first evaluate betaine concentration within several subgroups, including age (<60 vs ≥60), gender (female vs male), body mass index (<25 vs ≥25), diabetes mellitus (no vs yes), and previous history of cardiovascular disease (no vs yes). As shown in Table 2, there was no significant difference of plasma betaine level across these subgroups (all P > .05).

Figure 1.

Distribution of circulating betaine level (A) and its crude correlation to systolic blood pressure (B)

Table 2.

Betaine level in all patients and in subgroups

	Median (interquartile range), μmol/L	P
All	37.6 (30.1‐47.6)
Age, y
<60 (n = 229)	36.8 (29.3‐46.6)	.07
≥60 (n = 98)	39.0 (32.4‐49.1)
Gender
Female (n = 149)	37.5 (30.5‐46.7)	.86
Male (n = 178)	37.9 (29.5‐47.6)
BMI, kg/m2
<25 (n = 270)	37.7 (30.2‐46.4)	.58
≥25 (n = 57)	37.6 (29.7‐53.8)
Diabetes mellitus
No (n = 272)	37.9 (30.0‐47.6)	.65
Yes (n = 55)	36.0 (30.2‐44.5)
Previous history of CVD
No (n = 294)	37.9 (30.4‐47.6)	.13
Yes (n = 33)	34.4 (27.5‐42.4)

BMI, body mass index; CVD, cardiovascular disease.

We next explored the crude correlation between betaine and blood pressure. As shown in Figure (B), betaine level was correlated with systolic blood pressure in the patients (r = −.16, P = .003). The precise association of plasma betaine with blood pressure was determined by multiple linear regression analysis (Table 3). Since normal distribution check (the Kolmogorov‐Smirnov test) revealed that the distribution of betaine is skewed (P = .009), it was logarithm transformed in the following analyses. In an unadjusted model (Model‐1), log‐transformed betaine was significantly and negatively associated with both systolic (β = −3.66, 95% confidence interval: −6.10 to −1.22, P = .003) and DBP (β = −2.00, 95% confidence interval: −3.35 to −1.22, P = .003; β are calculated per SD increase in log‐transformed betaine level). In the other 2 models with adjustment for covariates (Model‐2: adjust for age and gender; Model‐3: adjust for age, gender, body mass index, smoking status, diabetes mellitus, previous history of cardiovascular disease, use of antihypertensives, use of erythropoietin, total and high‐density lipoprotein cholesterol, and interdialytic weight gain), the significant negative associations between log‐transformed betaine and blood pressures remained essentially unchanged (all P ≤ .015).

Table 3.

Associations of betaine level with blood pressures

	Systolic blood pressure			Diastolic blood pressure
	β	95% CI	P	β	95% CI	P
Model‐1	−3.66	−6.10 to −1.22	.003	−2.00	−3.35 to −0.66	.004
Model‐2	−3.79	−6.21 to −1.37	.002	−1.56	−2.82 to −0.30	.015
Model‐3	−3.35	−5.56 to −1.13	.003	−1.61	−2.83 to −0.39	.010

CI, confidence interval.

Model‐1: unadjusted model. Model‐2: adjust for age and gender. Model‐3: adjust for age, gender, body mass index, smoking status, diabetes mellitus, previous history of cardiovascular disease, use of antihypertensives, use of erythropoietin, total, and high‐density lipoprotein cholesterol and interdialytic weight gain. β are calculated per SD increase in log‐transformed betaine level.

A previous study has suggested that in vivo metabolism of betaine differs between females and males,13 therefore, we performed a gender‐stratified subgroup analysis. As shown in Table 4, the significant association between betaine and blood pressure was mainly limited to female patients (systolic blood pressure: β = −4.86, 95% confidence interval: −7.72 to −1.99, P = .001; DBP: β = −1.50, 95% confidence interval: −3.01‐0.02, P = .053), whereas no significant associations were observed in male patients (P = .223 for systolic blood pressure and .097 for DBP).

Table 4.

Gender‐stratified subgroup analysis of the association between betaine and blood pressure

	Systolic blood pressure			Diastolic blood pressure
	β	95% CI	P	β	95% CI	P
Female	−4.86	−7.72 to −1.99	.001	−1.50	−3.01‐0.02	.053
Male	−2.15	−5.62‐1.32	.223	−1.66	−3.63‐0.30	.097

β are calculated per SD increase in log‐transformed betaine level and were adjusted for age, body mass index, smoking status, diabetes mellitus, previous history of cardiovascular disease, use of antihypertensives, use of erythropoietin, total and high‐density lipoprotein cholesterol and interdialytic weight gain.

To explore whether the ambulatory blood pressure recording number affects the association between betaine and blood pressure found in our study, we performed a sensitivity analysis including only those with ≥48 recording (the median recording number of all the patients). After adjustment for covariates (age, gender, body mass index, smoking status, diabetes mellitus, previous history of cardiovascular disease, use of antihypertensives, use of erythropoietin, total and high‐density lipoprotein cholesterol, and interdialytic weight gain), the association between betaine and blood pressure remained essentially unchanged (P = .002 for systolic blood pressure and 0.010 for DBP).

4. DISCUSSION

In the current study, we explored the association of plasma betaine and blood pressure in patients on maintenance hemodialysis. Our results revealed a significant negative association between betaine and blood pressure level in these patients, and the association was significant mainly in females rather than in males. Our data suggest a potential role for betaine in blood pressure regulation in dialysis patients, which needs to be confirmed by future longitudinal studies or randomized clinical trials.

In humans, circulation betaine is mainly eliminated by metabolism rather than renal excretion.14 Previous evidence showed that renal excretion of betaine increased in subjects with kidney injury.15, 16 There is very few available information regarding betaine level in patients with end‐stage renal disease. In a study by Missailidis and colleagues,17 plasma betaine level decreased with declined renal function, with the lowest levels observed in CKD stage 5 patients. The median (interquartile range) betaine concentration in their study was 21.5 (10.5‐52.1) μmol/L in CKD stage 5 patients, being 4‐fold lower when compared to controls. In our study, the median predialysis plasma betaine concentration in patients was 37.6 (30.1‐47.6) μmol/L. Given that betaine is a small molecule with a molecular weight of 117.2, hemodialysis should easily remove it. Thus, the actual interdialytic circulating betaine load would be even lower than the predialysis level. It is therefore possible that the association between betaine and blood pressure in our study implies a link between betaine insufficiency and elevated blood pressure. It should be noted that, residual renal function would affect both the level of circulating betaine and blood pressure. However, given that the patients were all on hemodialysis three times a week, with median dialysis duration of nearly 5 years, residual renal function and its confounding effect on the association between betaine and blood pressure should be neglectable.

Biological function of betaine in vivo could shed light on the hypothetical mechanism by which betaine contributes to blood pressure regulation. In humans, betaine serves primarily as a methyl donor and is involved in homocysteine metabolism. Hyperhomocysteinemia is a well‐known cardiovascular risk factor and has been associated with the development of hypertension.18, 19 In vivo, betaine can transfer the methyl group to homocysteine to form methionine via the betaine homocysteine methyl transferase. Betaine supplementation alone is sufficient to reduce the plasma homocysteine level in healthy men and women with mildly elevated homocysteine.20 However, few published studies have evaluated the effects of betaine supplementation on blood pressure. In the study by Schwab and colleagues,9 42 obese subjects were randomly assigned to betaine supplementation group or placebo for 12 weeks. The authors noted a non‐significant decline in DBP (from 85.4 ± 7.9 mm Hg to 80.5 ± 7.1 mm Hg) in the betaine supplementation group compared to the controls. Limited sample size and a near‐to‐normal blood pressure level (122.5/85.4 mm Hg in the betaine supplementation group) at baseline may contribute to the absence of a significant decline in blood pressure. Another explanation is that the possible blood pressure lowering effect may only be limited to those with betaine deficiency. Our unpublished data reveal that betaine supplementation also failed to lower blood pressure in spontaneously hypertensive rats.

Several previous studies have examined the effects of betaine supplementation on homocysteine levels in chronic renal failure patients.21, 22, 23 It appears that betaine supplementation has no effect in reducing fasting plasma homocysteine concentration, which may be due to the inhibition of betaine homocysteine methyl transferase activity by dimethylglycine in these patients.24 However, data regarding the effect of betaine on blood pressure were absent in these studies.

The gender‐specified subgroup analysis in our study showed that the association between plasma betaine and blood pressure was mainly potent in female patients. The reason for this gender‐based difference is unknown. The study by Lever and colleagues13 indicates a gender difference in betaine metabolism. However, we did not observe a difference of plasma betaine levels between female and male patients in our study. Also, caution should be taken since the subgroup analyses were limited by the small sample size in each group.

There are several limitations in our study. First, the cross‐sectional analysis precludes any causality interference. Future longitudinal studies and randomized clinical trials are warranted to explore the exact relationship between betaine and blood pressure in dialysis patients. Second, we only measured plasma betaine concentration in a time spot instead of taking serial measurements. This may be less precise in estimating the actual betaine level in the patients. Third, the sample size is limited, especially for subgroup analysis.

In conclusion, reduced plasma betaine levels were associated with elevated blood pressure in a group of hemodialysis patients, and the association was mainly strong in females. Our results suggest a possible role for betaine deficiency in the development of hypertension in dialysis patients and warrant further confirmation in longitudinal studies or clinical trials.

CONFLICT OF INTEREST

None to declare.

Supporting information

 

Wang L, Zhao M, Liu W, et al. Association of betaine with blood pressure in dialysis patients. J Clin Hypertens. 2018;20:388–393. 10.1111/jch.13190