October 2011 E-Journal Club


We had a wonderful group of trainees here in October who hailed from Scranton, PA; New Bern, NC; Albuquerque, NM; and Charlotte, NC.  Our Journal club article this month deals with a still controversial (but hopefully soon to be passé) topic – the validity of using prealbumin, or prealbumin and C-reactive protein as an indicator of the adequacy of nutrition support.

October Citation:

Davis CJ, Sowa D, Keim KS, et al.  The Use of Prealbumin and C-Reactive Protein for Monitoring Nutrition Support in Adult Patients Receiving Enteral Nutrition in an Urban Medical Center.  JPEN J Parenter Enteral Nutr. 2011 Jul 28. [Epub ahead of print]


This was a retrospective review of 154 adult patients that were receiving enteral nutrition (EN) to determine if prealbumin (PAB), or PAB adjusted for C-reative protein correlated with the adequacy of nutrition provided.  The population studied was primarily critically ill patients, with 85.7% having spent some time in a medical, surgical, or neuro ICU during their hospital stay and a mean APACHE II score of 17.4 ± 5.8 for ICU patients. The sample was split into 2 groups based on those who received above and below an average of 60% of prescribed calorie and protein needs.

Patients were also divided into tertiles of percent calories and protein delivered.  The sequential changes in PAB and CRP were trended over time and compared to the tertiles of calorie and protein provision.  PAB levels were obtained within 24 hours of starting EN and every Tuesday and Friday thereafter. CRP values were obtained if PAB levels were <18 mg/dL to assess the presence and severity of inflammation.

Inclusion and Exclusion Criteria were:

Inclusion criteria:

Adult patients (≥18 years) requiring EN.

Exclusion criteria:

Patients that received EN for <3 days, had <2 measures of PAB, received parenteral nutrition (PN) or an oral diet while receiving EN or patients admitted to the psychiatric or maternity units.

Major Results reported by authors:

Change in PAB did not correlate with percent calories delivered or percent protein delivered.  There was no significant difference in change from first to last PAB for those receiving ≥60% of calorie needs (2.74 ± 9.50 mg/dL) compared to <60% of calorie needs (2.48 ± 9.36 mg/dL); P = 0.86. For those who achieved adequate protein intake, there was a nonsignificant trend toward a greater increase in PAB (4.12 ± 8.05 mg/dL) compared to those who did not achieve adequate protein intake (1.48 ± 10.16 mg/dL); P = .09.

Only the change in CRP was able to significantly predict change in PAB levels. The change in CRP explained 29.6% of the variance (R2 = 0.296) in the change of PAB after adjusting for either percent calories delivered (B = -0.051, P < .001) or percent protein delivered (B = -0.051, P < .001).

A subgroup analysis of 62 patients that had repeated measures of both CRP and PAB allowed multivariate analysis of the tertiles of calories and protein provided while controlling for CRP.  There was a significant difference in the pattern of PAB change based on the tertiles, with the lowest tertile of both percent calories and percent protein delivered having the largest increase in PAB over time.

Author’s Conclusions:

¨      PAB may not be an accurate marker for the adequacy of nutrition support in critically ill patients in the presence of inflammation.

¨     Changes in PAB do not reflect the delivery of adequate calories and protein in the presence of an inflammatory condition and changes in PAB correlated only with changes in inflammatory status.

¨      Until a more reliable marker for monitoring the adequacy of nutrition support can be identified, clinicians should use their skills in subjective global assessment and clinical judgment to determine whether a nutrition prescription should be adjusted to meet the needs of patients on nutrition support.


This is a valuable and much needed study that adds to our knowledge regarding the utility of checking PAB in the hospitalized patients.  Although patients were not randomized to a calorie/protein intake, the retrospective design allows investigation into the correlation between nutrition provision, degree of inflammation (CRP) and serum PAB.  One of the strengths of this study is that the actual volumes of EN formula that were provided to the patients were monitored on a daily basis.

One of the limitations of this study is that most patients did not meet full calorie or protein goals – on average patients met 57% of calorie needs and only 45.5% of patients achieved > 60% of goal calories.  However, the relatively generous calorie goal of 30-35 kcals/kg for many patients somewhat mitigates this limitation.  Additionally, the change in PAB for the tertile of 48 patients that received 69-114% of estimated needs was not significantly different that the tertile fed 4-50% of calorie needs.  The other potential limitation of this study is that only 39 patients (25.3%) received an indirect calorimetry study.

Our group did find it interesting that in the subgroup of 62 patients with repeated PAB and CRP measures that allowed multivariate analysis actually demonstrated that PAB increased the most in the tertile that received the least calories and protein.  While at first this may seem counterintuitive, we need to remember that serum proteins tend to be within normal limits in marasmic malnutrition because decreased calories and protein cause muscle breakdown and continue to supply the liver with substrate for serum protein synthesis.  Considering that those patients receiving the least nutrition would likely have the greater muscle catabolism, perhaps PAB synthesis was supported by muscle breakdown.  Regardless of the mechanism for the greater change in PAB levels in those fed the least, it certainly illustrates the limitation of using PAB as an indicator of nutrition adequacy.

Our Take Home Message

¨      Prealbumin is not a valid indicator of adequate nutrition delivery in hospitalized patients receiving EN.

¨      It is hard to justify increasing blood draws or spending any money in an already compromised healthcare system for a test that is not valid; UVAHS costs on  10/3/11:

o   Prealbumin – $124.00

o   C-reactive protein (high sensitivity CRP and “regular” CRP) – both $95.00.


Other News on the UVAHS GI Nutrition Website (

Upcoming Webinars for Fall 2011:

·         November 15: Vitamin and Mineral Issues in Gastric Bypass–Kelly O’Donnell, MS, RD, CNSD

·         December 6: The Basics of GI Anatomy & Physiology, Part II: The Malabsorption Workup–Carol Parrish, MS, RD

Check out our new:

·         “Nutrition Support Blog”

·         “Resources for the Nutrition Support Clinician”


Joe Krenitsky MS, RD

Carol Rees Parrish MS, RD


PS – Please feel free to forward on to friends and colleagues.