The effect of nutrition and severity of illness on blood vitamin concentrations in critically ill patients

Objectives

To measure blood vitamin concentrations after admission to ICU and to assess the effect of our standard nutritional regimen and the severity of organ failure on subsequent vitamin concentrations.

Method

Blood vitamin concentrations were measured in 42 ventilated patients following admission and thereafter twice weekly for the duration of their ICU stay. Parenteral nutrition with vitamin supplementation (daily Solivito N and Vitlipid N) or enteral nutrition was instituted independently by the clinicians according to their normal practice. Data were collected to allow calculation of Sepsis Score and Organ System Failure Score. Vitamins A, D, E, red cell thiamine and leucocyte vitamin C were measured using standard laboratory assays.

Results

Baseline vitamin A concentrations were below the normal range in 59% of patients. However, in those patients with renal failure (dialysis or creatinine > 300 μmol/1) baseline and subsequent concentrations were up to four times the upper limit of the normal range.

Baseline vitamin C levels were below the normal range (119–301 nmol/10⁸ wbc) in 77% of the group of non-survivors (median 63 nmol) and in 43% of the group of survivors (median 146).

This difference is not statistically significant in this pilot study given the number of patients involved. Vitamin C levels failed to reach the normal range over the study period (4-31 days) in both groups. Baseline vitamin E (alpha tocopherol) concentrations were below the normal range in 50% of patients, but reached the normal range in the majority of patients over the study period. Vitamin D2, D3 and red cell thiamine concentrations were within the normal range for nearly all patients throughout the study. Thiamine levels were supranormal in patients given parentrovite for a history of alcoholism. There was no correlation between vitamin concentrations, either on admission or during the study, and severity of illness as measured by the Sepsis Score or Organ System Failure Score.

Figure

Conclusions

The initial low concentrations of vitamin A in critical illness may be explained by concurrent low concentrations of serum retinol binding protein, a negative acute phase protein, although this was not measured in the study. However, with the onset of renal failure, vitamin A reaches high concentrations strongly suggesting supplementation should be reduced, in view of potential toxicity.

Both vitamin C and E levels were low on admission and may reflect low concentrations in the general population, reduced intake with the onset of illness, increased requirement or increased utilization. The standard regimen of supplementation was not sufficient to bring leucocyte vitamin C concentrations into the normal range for `healthy individuals'. The data suggests we should give more vitamin C, and possibly vitamin E, on admission to ICU and higher maintenance doses of vitamin C. Further studies will show if this has any effect on blood concentrations or on outcome.