The renal haemodynamics response to a meat meal is reduced and delayed in severe obesity

Introduction

It is accepted that GFR and RPF are reduced with obesity. However, data on severe obesity are scarce and no study has been conducted on the kidney hemodynamic response to a protein meal in this population; also tubular function is unknown.

Furtherore, it is largely known that kidney function varies after several interventions (drugs, water assumption). Among these, one ‘physiologically valid’ stimulus is the intake of a protein meal, which is known to induce several adjustments in kidney function, possibly through the action of glucagon, insulin, prostaglandins, nitric oxide, kallikrein.

Since no studies of renal tubular function in obese subjects has been carried out, the present study was devised with the triple goal to study (i) RPF and GFR in severe obese patients, (ii) their time course changes following a protein meal and (iii) the tubular function with Lithium clearance.

To this aim we have used obese subjects with intact kidney function at the time of the observation (non-albuminuric, preserved eGFR) (De Santo 1992 [1]).

Materials and methods

Subjects

We studied 28 obese subjects. Inclusion criteria were BMI>40 kg/m2, blood creatinine levels < 1mg/dl, blood fasting glucose levels <90 mg/dl, and albuminuria <30mg/24h and WHR indicating central-type obesity. The control group consisted of 20 subjects matched for age, sex, height, with normal arterial blood pressure and without microalbuminuria. Anthropometric measurements and major blood chemistry data are reported in Table 1. Urinary sodium excretion averaged 100±21 mM/day in control subjects and 116±14.8 mM/day in obese subjects.

Table 1: characterization of control and obese group

Controls (mean ±SD)

Obese (mean ±SD)

p (t-test for non-paired data)

n

20

28

Female/male

9/11

14/14

Age (yrs)

42±7

44±9

0.46

Weight (kg)

69.3±8.6

118.8±21.7

<<0.001

Height (cm)

164.8±9.4

161±10.4

0.2

BMI (Kg/m2)

25.7±4

45.9±8.3

<<0.001

Waist Hip ratio

0.89±0.03

1.20±0.03

<<0.001

Lean weight (Kg)

52.8±6.7

56.7±9.2

0.09

Fatty weight (Kg)

16.5±6.3

61.2±23

<<0.001

Total body water

57.5±4.2

41.4±9.3

<<0.001

Systolic blood pressure (mmHg)

127±9.9

134±14.8

0.06

Diastolic blood pressure (mmHg)

80.2±6.4

86±10.6

0.02

Serum Creatinine (mg/dl)

0.95±0.1

0.97±0.07

0.46

Blood urea nitrogen (g/L)

0.34±0.09

0.35±0.06

0.68

Glucose (mg/dl)

81.2±4.3

82.5±3.3

0.24

Insulin (µg/ml)

11±2

11.4±4

0.68

Glucagon (pg/ml)

100±23

98.6±20

0.82

Creatinine Clearance/height (ml/min·m)

76.8±10.8

71.7±17

0.21

eGFR (MDRD) ml/min·m2

83.3 ±16.9

79.3±12.4

0.4

eGFR (CKD-EPI) ml/min·m2

87.7±14.6

83.6±14.4

0.34

GFR (Cl inulin)/height (ml/min·m)

70.8±8

52.6±8.7

<<0.001

RPF/height (L/min·m)

2.9±0.5

2.4±0.7

0.005

FF

0.24±0.05

0.23±0.05

0.6

Procedure

All subjects have been maintained with a daily diet made of natural foods, providing 100mM Na+ and 1.15±0.22g/kg BW of proteins for 7 days; Na+ intake was controlled measuring the urinary excretion of Na+.

At 10PM of the day before the experiments, three hour after the evening meal, the subjects received 8.1mmol of elemental Lithium under the form of Lithium carbonate as detailed elsewhere.

Clearance studies were performed after overnight fasting. RPF was measured as the clearance of paraaminohippuric acid (PAI) and GFR as inulin clearance as described elsewhere (Anastasio 2000 [2]). In detail, the studies were started at 8:30AM by administering an oral water load of 20mg/kg body weight. Subsequently additional water was continuously administered to replace urine volume.

The urine was collected by means of bladder catheterization. Small Teflon (DuPont, Wilmington, DE) cannula was placed in the antecubital vein of each arm for inulin and blood sampling.

A bolous of inulin (40mg/kg BW), and p-aminohippurate (PAH, 12mg/kg) was administered followed by continuous inulin infusion, through constant speed infusion pump (Braun, Melsungen FRG) to maintain plasma concentrations of inulin and PAH at 20mg/dl at 2mg/dl respectively.

After an equilibration period of 1.5 hrs, 3 blood sampling to calculate inulin, PAH and lithium concentrations were done at regular intervals of 30 minutes.

The plasma concentration of inulin, PAH and lithium were determined as average of the initial and final concentration for each clearance period. The basal clearance of inulin and lithium were therefore measured as the average of three clearance 30 min apart.

Immediately after the last blood sampling, subjects have been asked to eat within 25 min period an oral protein load of 2g/kg body weight in the form of cooked red meat without salt and fat and without cereals, bread and vegetables as decribed elsewhere [13].

Four additional blood sampling after the protein meal were collected every 30 minutes to measure the clearance of lithium, PAH and inulin. An additional blood sample was obtained 60 min after the last blood sampling, to measure a fifth clearance period (as detailed in Figure 1).

Data were then analyzed to retrieve: (i) the time of peak GFR after the proteic meal, (ii) the absolute amount of GFR at the peak time, (iii) the relative amount of GFR at peak time, expressed as percent of the baseline levels and (iv) the cumulative postmeal GFR expressed as ml/180min.

GFR and RPF were normalized as a function of the body height.

Tubular functions

Tubular functions were measured by lithium clearance (CLi) and by its derived formulae as follows:

CLi as C = U x V/P , where U is the concentration of Lithium in urine, P is the concentration of Lithium in plasma and V is the urinary flow rate.

Lithium fractional excretion (FELi) was calculated as CLi/GFR x 100, which corresponds to the fractional excretion of sodium (FPRNa) .

Statistical analysis

Results are expressed as mean±SEM. Comparisons between normal and obese subjects at the basal

levels have been done using multivariate statistics. Time-dependent variables (GFR, urinary flux (V), RPF, CLi, FPRNa, FF) have been analyzed separately using repeated measures ANOVA with group (control-obese) as a two-level factor. Analysis of the relationship between BMI and change in RPF has been done using Pearson’s correlation index and linear regression. Variables have been tested for homogeneity of variance between groups. Rejection level has been set to p<0.05.

Results

As shown in Table 1 the obese subjects did not differ significantly from controls for systolic blood pressure, serum creatinine levels, blood urea and glucose. Similarly, in obese subjects plasma levels of insulin and glucagon basal were not different. On the contrary, body mass index and weight were significantly higher in the experimental group. As indicated in Table 1, creatinine learance and eGFR by MDRD and CKD-EPI in obese persons were not statistically different from those in controls. GFR (by inulin) was significantly lower in obese persons (p<0.01).

In obese persons the RPF (by PAH) was significantly lower (p<0.005) than in controls. GFR (ml/minx m) following a meat meal increased significantly over baseline (Fig 2A) but in lower extent in obese subjects, as shown by the cumulative GFR or area under the curve (Table 3).

Table 2: Effects of proteic meal (repeated measures ANOVA)

Variable

Effect of time

Effect of group

Interaction effect (time x group)

FF

F=1.36

p= 0.24

F=1.773

p=0.195

F=1.31

p=0.264

FPRNa

F=0.461

p=0.804

F=0.464

p=0.504

F=0.55

p=0.735

CLi

F=0.699

p=0.625

F=0.324

p=0.576

F=0.93

p=0.462

V

F=2.2

p=0.087

F=2.6

p=0.123

F=1.1

p=0.35

RPF/height

F=2.3

p=0.04

F=25.7

p=<<0.001

F=5.1

p=<<0.001

GFR/height

F=10.5

p<<0.001

F=33.1

p<<0.001

F=3.0

p=0.03

Table 3: Effects of proteic meal on variables describing the response curve (data represent mean ±SD)

Controls (mean ±SD)

Obese (mean ±SD)

Student’s t (p)

Cumulative postmeal GFR variation (ml/180min)

729±80

551±91

t=6 ( p<<0.01)

Time to peak (min)

60±24.4

80±30.8

t=2.1 (p=0.04)

Maximum GFR (ml/min·m)

90.7±14.8

69.7±10.7

t=4.3 (p<<0.001)

Maximum increase of GFR above the basal GFR (renal reserve) (ml/min·m)

26.3±14.9

26.3±13.5

t=0.02 (0.98)

The cumulative postmeal GFR variation (ml/180min) was significantly lower in obese persons compared to controls (Table 3). The time to peak GFR response was prolonged in obese subjects (80min vs 60 min in control subjects, p=0.04). The maximum GFR response averaged 90.7±14 in controls and 69.7±10.7 in obese subjects (p<<0.01). However, in obesity the peak response was significantly delayed (Table 3).

Specifically, GFR was increased at 30 min and peaked at 60 min after the proteic meal in controls. In obese subjects the GFR increased at 30 and 60 min, but peaked later at 90 min (Fig 3B).

RPF in obese subjects decreased at 30 and 60 min compared to the pre-meal values, whereas in normal subjects the RPF increased.

We directly tested a correlation between body weight and the modification of RPF at 30 min in Fig 4: it is evident a strong linear correlation between body mass index and RPF adjustment at 30 min after the protein meal (Pearson correlation -0.616, p<<0.01).

Proximal tubular function of obese patients and controls

CLi has been measured as a marker of Na+ reabsorption along the proximal tubule. Figure 2E shows that control subjects and obese patients had no different CLi (31.3±4.28 ml/min and 26.6±4.9 ml/min respectively, p>0.05, NS). Consequently, also the FPRNa (Fig 2F) between the two groups was similar (27.4±3.4 ml/min and 32.8±6.2 ml/min, in controls and obese patients, p>0.05, NS).

Discussion

This study indicates that in non-albuminuric obese patients GFR (by inulin method) is reduced in comparison to healthy controls, whereas eGFR and creatinine clearance were normal. Obese patients also show reduced and delayed GFR response to a meat meal. Lithium clearance as well as the fractional readsorption of Na in the proximal tubule is normal.

Several epidemiological data suggest that obesity is an independent risk factor of CKD; however, the exact mechanism linking the excess of adipose tissue and renal dysfunction is only partially understood. Since obesity is commonly accompanied by other morbidities (hypertension and diabetes) that may affect renal hemodynamic, it is difficult to dissect apart the respective roles of fat excess and its co morbidities on renal (dys)function.

The present study has been conducted in a population of obese subjects in the absence of significant comorbidities (non proteinuric obese subjects). Besides, GFR (inulin clearance) and the RPF (PAH clearance), the study has focused also tubular function.

Interestingly, in this setting renal hemodynamic did not show any significant difference compared with controls. Proximal tubule Na+ handling has been measured as CLi, , because Lithium is reabsorbed in parallel with sodium and water along the proximal tubule. In our clinical settings, all subjects were submitted to the same dietary sodium regimen; therefore the lack of significant differences in tubular functions is unlikely due to a bias in the CLi. Therefore, the lack of basal modifications of CLi and FPRNa in obese subjects suggests that the tubulo-glomerular feedback is still intact in our patients.

To further study the physiology of the kidney in obese subjects, we stimulated the kidney function using a proteic meal. It is widely known that dietary protein intake can modulate renal hemodynamic. The effect of dietary protein on rates of urea excretion has been observed since 1923 and on GFR later, in 1934. The protein-induced hyperfiltration is thought to originate from a combination of hormonal interactions and tubule-glomerular feedback. Among hormonal factors, increased glucagon secretion in response to protein administration has been repeatedly postulated as an important hyperfiltrating mediator subsequent to a cascade of events referred to as “pancreato-hepatorenal cascade”.

This physiological response of the kidney hemodynamic has been exploited as a mean to measure the ‘renal reserve’, indexed by the maximum GFR compared to the baseline level.

Previous observations have analyzed the renal reserve (the kidney hemodynamic adjustment to a protein meal) in obese hypertensive subjects. Hypertensive obese subjects have been found to bear a lower renal reserve when compared with that of lean hypertensives, which was accompanied by the inability to elevate the nitric oxide serum levels and lower increase in urinary kallikrein. Moreover, in obese subjects on a hyperproteic diet both the kidney size and the GFR were significantly increased from that measured at baseline, without changes in albumin excretion [23].

Our data show that in obese subjects actually the renal reserve, or the maximum value of the GFR, is not impaired compared to controls. However, more subtle modifications of the response of the kidney to the meal are evident: in normal subjects a protein meal induced acutely glomerular hyperfiltration, at 30 and 60 minutes, than the GFR went back to the baseline. Interestingly, as in controls, in obese subjects GFR increased from the baseline and it continued to increase up to 90 minutes after the meal. More dramatic was the difference in RPF between obese subjects and control. The protein meal induced a significant reduction of the RPF after 30 and 60 minutes, with a subsequent increase at 90 minutes in obese subjects, while the controls showed a rapid increased RPF from the baseline starting at 30 minutes.

It is difficult to explain the origin of this different response to the protein meal. Speculatively, the observed modifications in RPF adjustments might derive from (i) a different absorption of the aminoacids at gut level (ii) different hormonal response (e.g. glucagon) to the aminoacid load (iii) different baseline level of other regulatory hormones (leptin, inflammation-related prostaglandins). Clearly, disentangling these pathways may be important for the prevention of kidney impairment in obese subjects and is currently matter of ongoing research.

Our data pertain to a selected population of non-albuminuric obese subjects. Overall, these data indicate that, at basal, severely obese subjects (in the absence of the common comorbidities) show normal glomerular function, renal perfusion and proximal tubule Na+ handling; in contrast, the renal hemodynamic response to a protein meal differs from non obese subjects. Further studies are needed unravel the clinical valence of these findings, as they may represent the earliest hallmark of future kidney dysfunction.