Protein intake and risk of urolithiasis and kidney diseases: an umbrella review of systematic reviews for the evidence-based guideline of the German Nutrition Society

This umbrella review, including 6 SRs with MA and 3 SRs without MA, examined the implications of HPI for kidney health. Key findings are that for daily protein ingestion above dietary recommendations, no convincing evidence could be ascertained for kidney function decline relevant relationships with urinary albumin excretion, renal GFR, and kidney stone risk. Also for the further assessed renal-related outcomes, none of the gradings of the overall certainty of evidence led to an assessment as ‘possible’ or ‘probable’ for detrimental HPI influences on kidney function.

According to the criteria given in Table 1, for the risk marker of CKD albumin excretion [11], the overall certainty of evidence was graded as ‘possible’ to be not elevated through HPI in both young and elderly healthy adults. Furthermore, overall certainty of evidence was graded as ‘possible’ for GFR and ‘probable’ for urinary calcium excretion, as well as for serum urea, to be physiologically (regulatorily) increased with HPI. It is noteworthy that this grading as ‘possible’ for GFR was partly due to a downgrading effect through the only SR [15] with an overall low methodological quality (AMSTAR 2) and a substantial miscategorisation of the outcome GFR [15]. Instead of 8 out of 13 RCTs with GFR determinations (< 2/3), as reported by the authors, actually 8 out of 11 RCTs with GFR determinations (> 2/3) showed higher GFRs with HPI [15], thus rather allowing an assessment of the overall certainty of evidence as ‘almost probable’ and not just ‘possible’.

As outlined below, the elevations of most of the outcomes along with HPI have to be interpreted cautiously, i.e. mostly as physiological regulatory responses and not as pathophysiological increases. However, it should be considered that an elevated urinary calcium excretion may represent a risk factor for calcium stone formation. Nevertheless, for kidney stone disease, ‘possible’ evidence was derived for an absence of an association with higher animal protein intake [17, 20].

As HPI has been associated with metabolic changes that can exhibit a risk for kidney stone formation in healthy individuals [4, 24], this issue is addressed in the following along with further specific comments on the examined outcomes. Physiological background explanations are provided for a far-reaching inappropriateness of urinary calcium excretion, urinary pH, serum urea, serum uric acid, and even of the important kidney function parameter GFR (or serum creatinine) as unbiased renal health outcomes for examinations in (mostly) healthy populations if no specific adjustments are conducted. Accordingly, the suitability of these parameters to unbiasedly reveal pathophysiologically relevant influences of HPI on kidney health will be critically appraised.

Albumin excretion

The current NutriGrade ratings of ‘low’ for the finding that the diagnostically important kidney parameter urinary albumin excretion and dietary protein intake are unrelated, definitely prompting that this potential absence of an albuminuria-elevating effect through HPI needs to be further examined and particularly studied for observation periods longer than 2 years.

Kidney stones

The prevalence of urinary stone disease in the general population has been reported to range between 4.7% and 8.8% [25, 26]. Kidney stone formation is associated with an elevated risk of chronic and end-stage kidney disease, probably due to kidney injury from obstructive nephropathy [27, 28]. An HPI may promote the risk of stone formation by providing an acid load that could lead to several metabolic changes, including decreases in urinary pH and citrate excretion, and increases in urinary calcium and uric acid excretion [24, 29,30,31]. A higher dietary net acid load, estimated by animal protein-to-potassium ratio or net acid excretion (NAE), was associated with a higher risk of kidney stone formation in large observational studies [32]. These data suggest that the proportion of the consumed amount of alkalising fruits and vegetables compared to the total amount of ingested protein could modify the risk of HPI for kidney stone formation. It is beyond controversy that fruits and vegetables have a marked alkalising potential and can in such a way relevantly neutralise the proton load, metabolically generated from ingested protein [33, 34]. High dietary acidity, resulting in lower urine pH, is a risk factor for several kidney stone types, particularly for the most common, i.e. calcium oxalate stones. The higher the urine pH, the higher is the stone-inhibiting citrate excretion and calcium-binding capacity and the lower is the urinary calcium excretion [35].

In conclusion, a number of protein intake-related, metabolic, and idiopathic risk factors and confounders, such as low or high urine pH, hypercalciuria, hypocitraturia, hyperuricosuria, hyperoxaluria and further dietary/environmental risk factors, such as high sodium chloride intake and low urine volume [36, 37], all complicate a straight examination of ‘the inherent impact of protein’ on stone formation. However, most of these risk factors can at least partly be avoided or reduced by the respective changes in dietary habits, e.g. by increasing the habitual intake of metabolically alkalising fruits and vegetables [32].

GFR

Increases in GFR frequently occur during the first years after onset of diabetes mellitus type 1 or 2 [38]. This phenomenon is termed glomerular hyperfiltration. With advancing duration of the disorder, hyperfiltration regresses again and frequently turns into a pathophysiological decline of GFR. Increased body fatness and obesity also lead to elevations in GFR, independent of hyperglycaemia and other metabolic and hormonal signals also present in diabetes.

Another major stimulus of GFR is protein intake. GFR increases, lasting for several hours, occur after protein-rich meals [39], implying that if HPI and blood sampling are temporally relatively far apart (overnight fasting or even longer), GFR increases can, but may not necessarily, be any longer detectable with the use of mere serum measurement-based estimates (eGFR), although they would be observable by 24-h urine-based GFR measures. This is one of a number of explanations why in all MAs and SRs that included the outcome GFR, at least some primary studies were present which did not find GFR increases following increased protein ingestion by healthy subjects. In principle, elevations in GFR are basic, physiologically adaptive mechanisms induced by HPI in case of normal kidney function state [39,40,41,42,43].

In line herewith, none of the SRs of this umbrella review found clear indications for a GFR reduction due to HPI. Accordingly, one could classify GFR increases or at least GFR stability as a very probable consequence of raises in protein intake above dietary recommendations in the healthy state, despite the fact that the formal use of the modified grading system [8] only resulted in a grading of ‘possible’ for the overall certainty of evidence for the absence of GFR reductions.

Among others, not only younger age (until around 35 years) [44] and HPI [39,40,41,42,43], but also increases in BW [41], BMI or fat mass [45, 46], insulin resistance [47, 48], insulin secretion [49], and sodium chloride intake [45, 50, 51] all have a GFR-elevating potential. Accordingly, the examination of a potential kidney function decline by using GFR reduction as a marker or an outcome in initially metabolically healthy subjects appears – at first glance – not ideal for an exposure that by itself biologically raises GFR. GFR changes, however, should be studied in the future (as far as possible bias free) as a major outcome for the assessment of gradual kidney function decline over periods of more than 5–10 years by more appropriately controlling relevant confounders.

Urinary calcium excretion

Various dietary factors affect urinary calcium excretion, particularly the intakes of calcium, protein, and sodium chloride. In healthy subjects, intestinal calcium absorption is approximately 25% [52]. However, intestinal hyperabsorption of calcium is frequently diagnosed in stone formers [53]. Higher dietary protein intakes are consistently reported to increase urinary calcium excretion [18, 54, 55], in part due to the increased GFR [56] (see above). Apart from calcium and protein intake, urinary calcium excretion is also related to urinary NAE and acidotic stimuli [57]. While the administration of 1.5 g/d L-methionine did not significantly raise urinary calcium excretion in healthy subjects [58], the supplementation of 3 g/d L-methionine resulted in a significant increase in urinary calcium excretion by about 1 mmol/d (40 mg/d) in parallel with a rise in urinary NAE of 40 mEq/d [59]. Accordingly, without specific adjustments for the aforementioned confounding influences, the utilisation of urinary calcium excretion as an important urolithiasis-related renal health outcome appears to be less useful.

Urinary pH

Urinary pH marks the small amount of free hydrogen ions (H+) not buffered by ammonia and titratable acid (i.e. mostly phosphate) and reflects, to some degree, the overall excess of H+ that is renally secreted. The overall, i.e. the buffered amount of H+ daily eliminated by the kidney is quantified as NAE [60,61,62]. Although 24-h urine pH and NAE/d usually show good correlations [34, 63], a variation in renal buffer supply can markedly change the usual pH–NAE relationship. One major confounder in this regard is protein intake itself. The higher the protein intake, the higher is the kidney´s capacity to excrete surplus H+ [64, 65]. Accordingly, if protein intake increases and NAE is constant (through higher alkali intake), the ammonia buffer is much more easily renally provided. This means that a lower free proton stress (a lower H+ signalling) is required to increase buffer provision, i.e. ammoniagenesis. Correspondingly, in subjects without kidney disease, urine pH will be higher with HPI for every given acid load, i.e. for a constant potential renal acid load (PRAL) or a constant NAE. Thus, 24-h urine pH can only be used as a marker for kidney function change if measurements of renal 24-h NAE or PRAL are concurrently performed and appropriately adjusted for [66]. Even with HPI of around 80 to 100 g/d, mean 24-h urine pH can be kept at ≥ 6 through moderate alkali equivalent ingestion [33, 64]. Besides this, HPI with a higher NAE, higher age (> 50 years) [67], higher BMI or body fat [66, 68, 69], and other features of the metabolic syndrome including insulin resistance [68, 70] each contribute to urine pH reductions. Thus, the NutriGrade rating of ‘low’ obtained for a potential constancy of the urine pH along with rises in protein intake [18] suggests that HPI does not necessarily increase renal “free proton stress”. However, the examination of urinary free protons at least in combination with (reliable markers of) protein intake and the related net acid load can be a valuable tool to assess renal acid excretion function as well as stone formation risk [66, 71].

Serum urea and serum uric acid

Increases in urea, mostly within the normal physiological range, have been reported in almost all primary studies after protein intake was raised. Since elevation of serum urea above the upper limit of the normal range primarily depends on functional GFR reduction [72] and further confounders like hydration status, circulating urea rather represents an insensitive indicator of kidney function [72]. Next, serum uric acid shows varying interdependences with protein ingestion [73], purine intake, hydration status [74], and GFR [75], as well as with metabolic syndrome [76] and diabetes mellitus type 2 [77]. Thus, irrespective of the rating of the overall certainty of evidence as ‘probable’ or ‘insufficient’ for effects due to increases in protein intake, the utilisation of serum urea and uric acid, respectively, for the assessment of potential influences of HPI on kidney health appears to be a less specific approach.

Strengths and limitations

A strength of this umbrella review is that six of the included hitherto published SRs on the relevance of HPI for kidney health comprise, either exclusively or primarily, RCTs. A further strength is that we critically examined in more detail the suitability, as well as relevant physiological confounders, of those kidney parameters commonly used to investigate kidney health. However, several limitations have to be taken into account when assessing the findings of the SRs included in this umbrella review. We applied NutriGrade instead of the GRADE approach (Grading of Recommendations, Assessment, Development and Evaluation) because an important novelty of NutriGrade (published in 2016) was the modified classification for MA of RCTs and cohort studies compared with the traditional GRADE approach (initially classifying RCTs with an initial high score and cohort studies with a low score) [78]. We are aware that in the meantime, the GRADE approach was amended (adjustments published in 2019, but after the guideline methodology was established in 2017) in a way that cohort studies can now also be assigned an initially high score, when risk of bias tools such as ROBINS-I are used [79]. The intervention duration of the primary RCTs and also the protein intake levels varied considerably with ranges from 1 week to 2 years, and intakes from 12.5 to 40 En% solely in the high-protein groups, respectively. Although dietary protein sources have been provided in most primary studies, more specific statements regarding the relevance of animal vs. plant vs. dairy protein could not be drawn, particularly due to an insufficient number of corresponding specific data analyses. Furthermore, the substantial degree of heterogeneity, present for the different outcomes, could not be further assessed. Further important limitations of the current umbrella review are that (i) major primary studies, not included in the SRs or MAs, remained unconsidered and (ii) that, of the nine SRs that could be included, only 3 examined the most diagnostically conclusive outcome variable, i.e. albuminuria as well as kidney stones. The various other kidney function-related outcomes that were examined, however, showed clear weaknesses regarding a specific, i.e. an unconfounded assessment of possible kidney function impairments. Their increases (or potential urinary pH reductions) along with HPI are biologically plausible, but without direct specific pathophysiological relevance.

留言 (0)

沒有登入
gif