Article Text

Download PDFPDF

Extended report
Acute effect of milk on serum urate concentrations: a randomised controlled crossover trial
  1. Nicola Dalbeth1,
  2. Sumwai Wong1,
  3. Greg D Gamble1,
  4. Anne Horne1,
  5. Barbara Mason1,
  6. Bregina Pool1,
  7. Lynette Fairbanks2,
  8. Fiona M McQueen3,
  9. Jillian Cornish1,
  10. Ian R Reid1,
  11. Kate Palmano4
  1. 1Bone and Joint Research Group, Department of Medicine, University of Auckland, Auckland, New Zealand
  2. 2Purine Research Unit, Guy's Hospital, London, UK
  3. 3Department of Molecular Medicine, University of Auckland, Auckland, New Zealand
  4. 4Fonterra Research Centre, Palmerston North, New Zealand
  1. Correspondence to Dr Nicola Dalbeth, Department of Medicine, Faculty of Medical and Health Sciences, University of Auckland, 85 Park Rd, Grafton, Auckland, New Zealand; n.dalbeth{at}


Objectives Recent observational studies have highlighted the beneficial role of dairy ingestion in gout prevention. The aims of this study were to determine the acute effects of milk ingestion on serum urate concentrations and examine the mechanisms of these effects.

Methods This was a short-term randomised controlled crossover trial of milk in 16 healthy male volunteers. The following products were tested (each 80 g protein): soy control, early season skim milk, late season skim milk (containing high concentrations of orotic acid, a naturally occurring uricosuric agent) and ultrafiltrated MPC 85 skim milk. Each participant received a single dose of each product in random order. Serum and urine were obtained immediately before and then hourly over a 3 h period after ingestion of each study product.

Results Ingestion of the soy control led to an increase in serum urate concentrations by approximately 10%. In contrast, ingestion of all milks led to a decrease in serum urate concentrations by approximately 10% (p<0.0001). All products (including soy) rapidly increased the fractional excretion of uric acid (FEUA). Late season milk led to a greater increase in FEUA than MPC 85 (p=0.02) and early season milk (p=0.052). There were no differences over time in serum oxypurines or purine-containing nucleosides. However, all products increased the fractional excretion of xanthine.

Conclusions Intact milk has an acute urate-lowering effect. These data provide further rationale for long-term intervention studies to determine whether such dietary interventions have an adjunctive role in the management of individuals with hyperuricaemia and gout.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


The beneficial role of dairy products in the prevention of gout has recently been highlighted. The Health Professionals Follow-up Study of more than 47 000 men reported a 21% reduction in the risk of gout per additional daily serving of total dairy products over a 12-year period. This effect was greatest with low fat and skim milk.1 Similar findings have been reported in the Nurses' Health Study which followed 92 224 women over 24 years.2 This protective effect may be due to the suppression of serum urate concentrations by dairy products. This conclusion was supported by an analysis of the Third National Health and Nutrition Examination Survey (NHANES III) which reported an inverse relationship between reported ingestion of dairy products and serum urate concentrations.3 We have recently shown that the protective effect of dairy products against gout is not due to a urate-lowering effect of dietary calcium.4

Several intervention studies have also indicated that dairy products have a urate-lowering effect. A short-term intervention study showed that certain isolated dairy proteins, but not soy protein, acutely reduced serum urate concentrations in healthy participants.5 This study also showed that both dairy and soy proteins increased uric acid excretion. A further study reported that a 4-week dairy-free diet was associated with increased serum urate levels in healthy women.6 To date, the urate-lowering effects of intact milk have not been assessed.

The uric acid and orotic acid concentrations of bovine milk have recently been reported.7 In certain dairy-producing countries (such as New Zealand and Australia), cows are primarily grass-fed and milking is seasonal with production geared to pastoral growth. In grass-feeding cows, major differences are present in these components based on the stage of lactation and time of production. In particular, the concentrations of the pyrimidine orotic acid are maximal in late season mature milk. Orotic acid is a well-recognised uricosuric compound which acts by competing with uric acid for transport through a renal tubular urate/anion exchanger, thus decreasing the reabsorption of uric acid.8 The observation that high orotic acid concentrations occur naturally in late season milk raises the possibility that this milk product may have particularly beneficial urate-lowering effects.

The primary aim of this study was to determine the acute effects of intact milk on serum urate concentrations. In addition, we wished to determine whether milk produced by grass-fed cows at various production times had a differential effect on serum urate or uric acid excretion, and to examine the mechanisms of the urate-lowering effects of milk.


This was a short-term randomised controlled crossover trial of milk in 16 healthy male volunteers. The study was designed to determine the effects of intact milk on serum urate concentrations over 3 h.


Healthy male volunteers were recruited by public advertisement. They were not included in the study if any of the following criteria were present: (1) abnormal renal function (serum creatinine >120 μmol/l), (2) anaemia, (3) previous gout, (4) diabetes mellitus, (5) diuretic use and (6) lactose intolerance. Potential participants attended a screening visit where a general health questionnaire was completed and baseline measurements (including weight, height and waist circumference) and physical examination were performed. Screening blood tests were also obtained. Three prospective participants were excluded following the screening visit (one patient with gout and two declined to enter the study). The first participant was randomised in June 2007 and the last participant completed the study in May 2008. All visits took place at a clinical research facility in a tertiary medical centre. The study was registered by the Australian Clinical Trials Registry (ACTRN12607000187448).


The trial protocol was modified from a previous study of dairy protein ingestion.5 The study specifically recruited men in order to identify those with higher serum urate concentrations at baseline. The study followed a crossover design with four treatments and four time periods (complete block design) where each of the 16 participants received all four products in random order. Participants were allocated a study number according to the sequence of their enrolment. Sixteen out of 24 possible dose orders were chosen so that each treatment was administered four times in each time period and there was an approximate balance with respect to treatments immediately following each other. Randomisation was performed prior to the commencement of the study by the study statistician. Study products were preweighed into 64 containers labelled with individual 3-digit codes. The study product order for each participant was chosen at random and was provided in sealed envelopes. The staff member who dispensed the study product into the numbered containers had no direct contact with other study staff or with the trial participants. Participants and investigators were blinded to the identity of the study products, which were identically packaged. Unblinding occurred at the end of the study. Adherence to the study products was complete, as the study nurse directly observed ingestion of each product.

At each study visit a venous catheter was inserted for blood collection. Following an overnight fast, participants consumed the study product between 08:00 and 09:00 and blood was obtained for urate, creatinine, urea, albumin and serum storage for purine metabolites prior to ingestion and then 60 min, 120 min and 180 min after ingestion. Urine volume was measured and urine was obtained at these time points for testing of uric acid, creatinine and storage for purine metabolites. After voiding urine, participants received a volume of water equivalent to the recorded urine volume. Adverse events were recorded at each visit. Each study visit was separated by at least a week, and all visits were completed within 6 weeks. No other dietary modifications were made throughout the study period.

The following products were tested: soy control (PRO-FAM 873; ADM, Decatur, Illinois, USA), early season skim milk, late season skim milk and MPC 85 skim milk (an ultrafiltrated skim milk product). All milk products were produced by Fonterra Co-operative Group, Palmerston North, New Zealand. Each product dose was equivalent to 80 g protein and was administered in an 800 ml suspension. Before commencement of the study, the purine and orotic acid content of all products was measured using high-performance liquid chromatography (HPLC). Lactose was measured by the ferricyanide reduction method using a Technicon AutoAnalyser II9 and xanthine oxidase was measured using the Amplex Red xanthine oxidase kit (Invitrogen, Carlsbad, California, USA).

Serum and urine chemistry were tested using the Roche Modular P (Hitachi) analyser (Basel, Switzerland). Purine testing (other than urate) of clinical samples was performed at the Purine Research Unit, Guy's Hospital, London, UK. Serum was tested for inosine, guanosine, hypoxanthine and xanthine by HPLC, and urine was tested for hypoxanthine and xanthine by ultra-performance liquid chromatography. The fractional excretion of uric acid (FEUA) was calculated; this is the ratio between the renal clearance of uric acid to the renal clearance of creatinine, expressed as a percentage.

Statistical analysis

The primary end point of the study was change in serum urate concentration, and the secondary and exploratory end points were change in serum urea concentration, change in FEUA, change in serum concentrations of other purines and change in the fractional excretion of oxypurines.

This study was powered (80%) at the 5% significance level to detect an effect size difference between any of the treatment arms of at least 0.73 should the sample size remain ≥15. This corresponded with a large effect size10 and was comparable to the effect size observed in the study by Garrel et al.5 Because of the prolonged follow-up in the present study, the overall sample size was increased by one to provide protection for loss to follow-up.

Statistical analysis was performed by a statistician independent of the product manufacturers. Data are presented as mean (SD) for descriptive purposes; however, measures of effect are presented with the appropriate 95% CI. Data were analysed using a mixed models approach to repeated measures. Significant group effects were explored using the method of Tukey. All analyses were performed using SAS Version 9.2 (SAS Institute Inc, Cary, North Carolina, USA) on an intention-to-treat basis. p<0.05 was considered significant and all tests were two-tailed.


Baseline patient characteristics

All participants completed the study. At screening the median age was 34 years (range 20–56), body mass index 24.6 kg/m2 (range 20.0–30.9), waist circumference 87 cm (range 71–101) and serum creatinine 89 μmol/l (range 73–129). Baseline laboratory values for serum urate concentrations and other measures for each treatment arm are shown in table 1.

Table 1

Baseline laboratory values for each treatment arm

Compositional analysis of tested products

The purine, orotic acid, xanthine oxidase and lactose contents of each study product were tested before commencement of the study (table 2). The soy control contained high concentrations of the purine-containing nucleotides adenosine monophosphate (AMP) and guanosine monophosphate (GMP), and also the purine nucleosides adenosine and guanosine. No uric acid, orotic acid or lactose was detected in the soy product. In contrast, purines, orotic acid and lactose were absent or virtually absent from the ultrafiltrated MPC 85 milk. Early season milk had the highest concentration of uric acid and moderate orotic acid concentration. Late season milk had the highest concentration of orotic acid, with a lower concentration of uric acid. Both early and late season milks had detectable AMP and high lactose concentrations. Xanthine oxidase levels were low in soy and were higher in all milks.

Table 2

Analysis of products tested

Serum urate concentrations

Following ingestion of the soy control, the serum urate concentration increased by approximately 10% within 2 h and remained raised throughout the study period (figure 1). In contrast, following ingestion of the milks the serum urate concentration gradually reduced over the 3 h follow-up period with an overall reduction in serum urate by approximately 10%. All milks were associated with a significant reduction in the serum urate concentration over time and in comparison with the soy control, and there was no difference between the three milks with respect to the intensity of the urate-lowering effect. There was no relationship between the age of the participants and changes in serum urate concentration (p=0.29–0.90).

Figure 1

Effects of soy and milks on (A) percentage change in serum urate concentration from baseline values and (B) serum urate concentrations over time. Data are presented as mean (95% CI).

Serum urea concentrations

All products led to a significant increase in the serum urea concentration (figure 2A). The soy control led to significantly greater increases in serum urea concentrations than early season milk (differences in least square means post hoc analysis, adjusted p=0.04), late season milk (p<0.0001) and MPC 85 milk (p=0.01). The increase in serum urea was not associated with an increase in serum creatinine for any product (data not shown), suggesting that the increases in serum urea concentrations were related to catabolism of a protein load rather than changes in renal function. There was no change in serum albumin over time or between groups (data not shown).

Figure 2

Effects of soy and milks on (A) percentage change in serum urea concentration from baseline values and (B) percentage change in fractional excretion of uric acid (FEUA) from baseline values. Data are presented as mean (95% CI).

FEUA of uric acid

All products (including soy) increased the FEUA (figure 2B). The increased FEUA was maximal 2 h after ingestion of the products. The soy control led to a significant increase in FEUA compared with early season milk and MPC 85 milk (both p<0.0001), but did not differ from late season milk (p=0.13). Late season milk led to a significantly greater increase in FEUA compared with MPC 85 (p=0.02) and a trend towards increased FEUA compared with early season milk (p=0.052). There was no relationship between the age of the participants and changes in FEUA (p=0.66–0.88).

Serum concentrations of purine-containing nucleosides and oxypurines

Overall, there were no significant differences over time or between groups in changes in the serum concentrations of the oxypurines (xanthine and hypoxanthine) or purine-containing nucleosides (guanosine and inosine) (figure 3). Large variations in these values were observed between individuals and over time.

Figure 3

Effects of soy and milks on percentage changes from baseline in serum concentrations of purine-containing nucleosides and oxypurines: (A) guanosine; (B) inosine; (C) hypoxanthine; (D) xanthine. Data are presented as mean (95% CI).

Fractional excretion of urate precursors

There was no significant difference over time or between groups with respect to changes in the fractional excretion of hypoxanthine (figure 4A). However, all products led to an increase in the fractional excretion of xanthine over time (figure 4B). The significant increases in the fractional excretion of xanthine occurred 1–2 h after ingestion of the products.

Figure 4

Effects of soy and milks on percentage change from baseline values of (A) fractional excretion of hypoxanthine (FEHX) and (B) fractional excretion of xanthine (FEX). Data are presented as mean (95% CI).

Adverse events

No participant withdrew from the study due to adverse events. Headache,3 diarrhoea5 and nausea4 were the only reported adverse events. There was no difference between groups in the frequency of reported adverse events.


This study has shown that intact milk has an acute urate-lowering effect. These observations are of relevance in view of previous observational studies which have reported a protective effect of milk on the development of gout, and also an inverse relationship between serum urate concentrations and milk ingestion. Although another intervention study has demonstrated an acute urate-lowering effect with certain isolated dairy proteins, the urate-lowering effects of intact milk have not previously been analysed. Our data indicate that the observed urate-lowering effect of milk occurs due to its low purine content in combination with increased excretion of both uric acid in response to a protein load.

We also analysed the differential effect of certain milk products obtained from grass-fed cows at different production times. This study has shown that late season milk, which contains high concentrations of orotic acid, has a preferential effect on uric acid excretion compared with the other milks tested. The clinical relevance of this observation is uncertain, noting that this milk did not preferentially reduce serum urate concentrations compared with the other milks in the short term. It is conceivable that this preferential effect on uric acid excretion may have longer term benefits with respect to the urate lowering effect. Studies of longer duration are required to address this possibility.

In addition to hyperuricaemia, serum hypoxanthine and xanthine concentrations are increased in patients with gout.11 Furthermore, patients with primary gout have impaired ability to renally excrete the oxypurines.11 Previous studies have shown that other dietary factors associated with hyperuric-aemia such as beer also lead to increased plasma hypoxanthine and xanthine concentrations.12 For this reason, the effects of soy and milk products on serum oxypurine concentrations and renal clearance of these urate precursors were studied. It is interesting that, despite the striking changes in serum urate concentration, differential changes in serum oxypurine and purine-containing nucleoside concentrations were not identified following ingestion of the soy and milk products. These findings do not exclude alterations in hepatic purine concentrations following protein ingestion and during purine nucleotide synthesis and degradation to urate. Furthermore, additional effects on hepatic xanthine oxidase following ingestion of milk products cannot be excluded by the current analysis. However, it is reasonable to conclude that the xanthine oxidase content of the ingested products did not influence serum urate concentrations, noting that the high xanthine oxidase concentrations in milk compared with soy were not associated with increased serum urate concentrations.

The role of carbohydrate ingestion in the regulation of serum urate concentrations and the risk of gout has recently been emphasised.13 In particular, oral intake of fructose has been shown to increase serum urate concentrations.14,,16 Lactose is the principal carbohydrate in milk and it is conceivable that the protective effects of milk against gout are related to a differential effect of lactose on renal handling of uric acid. Our analysis indicates that the urate-lowering effect of milk is not due to a lactose effect, as no difference in serum urate concentrations or FEUA was observed between the ultrafiltrated MPC 85 milk (containing negligible lactose) and the high lactose-containing early season milk.

The uricosuric effect of a protein load is well recognised and may explain the observed uricosuric effects of all the products tested.5 17 18 It is also possible that the large fluid load related to soy and milk ingestion had a uricosuric effect. The observation that soy, with its high purine content, led to a significant increase in serum urate concentration despite promoting uric acid excretion suggests that hepatic urate production has a dominant effect on serum urate concentrations, at least in the short term following ingestion. However, the results suggest that the uricosuric effect of milk ingestion does become relevant in the context of the low purine content of milk. Our study has also shown that both soy and milk products promote renal excretion of the urate precursor xanthine. We observed large increases in xanthine excretion of up to 200% following ingestion of the products, occurring 1–2 h after ingestion. It is conceivable that increased xanthine excretion may contribute to the urate-lowering effect of milk by reducing the availability of substrates for urate production. It has previously been reported that xanthine excretion is strongly correlated with uric acid excretion.11 Whether the observed increase in xanthine excretion occurs via the same renal mechanisms as uric acid excretion is unknown and warrants further analysis.

We acknowledge that this study has some limitations. In particular, due to intensive study protocol, the effects of milk were only examined over a short time period and it is uncertain whether the urate-lowering effects observed would persist over longer time points. The volume of milk was high and may not be easily incorporated into a standard diet. The daily intake of 800 ml milk is similar to that reported as the highest quintile of dairy ingestion (≥3 servings/day) in the analysis of NHANES III, which was associated with significantly reduced serum urate concentrations.3 Furthermore, it is conceivable that patients with existing hyperuricaemia and gout may respond in a different way to healthy volunteers with normal renal function and purine metabolism. Despite these limitations, we have identified a consistent and significant effect of milk on serum urate, suggesting that dietary modification through increased dairy intake may have meaningful effects on serum urate concentrations. On the basis of these results, it will now be interesting to determine whether the acute effects are sustained over time and to study the long-term clinical effects of milk ingestion in patients with gout and hyperuricaemia. Further work should also address the optimal dose of milk that can be tolerated and has a clinical benefit in this patient group.

In summary, this study has shown that intact milk has significant acute urate-lowering effects. These data, together with observational data supporting the benefits of dairy ingestion for prevention of gout, provide a rationale for long-term intervention studies to determine whether such dietary interventions can play an adjunctive role in the management of hyperuricaemia and gout.



  • Funding This study was funded by LactoPharma (a joint venture between Fonterra Ltd, Fonterra R&D Ltd and Auckland UniServices Ltd) and the New Zealand Government Foundation for Research Science and Technology.

  • Competing interests KP is an employee of Fonterra Co-operative Group Ltd.

  • Ethics approval This study was conducted with the approval of the Northern X regional ethics committee (NTX/07/02/005) and each participant gave written informed consent.

  • Provenance and peer review Not commissioned; externally peer reviewed.