Routine testing of folate levels in geriatric assessment for dementia.

To the Editor:-Folate deficiency is well established as a cause of megaloblastic anemia, but its relationship to dementia is less certain. In 1967, Strachan and Henderson' reported on two patients with dementia who were found to have megaloblastic anemias due to folate deficiency. Treatment with folic acid cured both the anemia and the dementia in one patient and resulted in marked improvement of mental status in the other patient. Sneath et al' found a significant correlation between low red-cell folate and severity of dementia on mental assessment scores. Since the dissemination of these reports and others, folic acid deficiency is usually included in the differential diagnosis of treatable dementias. Thus in many geriatric assessment/evaluation programs serum folate or red-blood-cell folate, or both, are measured as part of the routine testing. Larson et a13 found the cost of screening blood tests per diagnosis of low folate level to be $1,001, but, if determined only in those with abnormal CBCs, the cost was reduced somewhat to $842. Thus they suggested folate testing be done on a selective basis. In 60 consecutive patients with progessive intellectual deterioration, Freemon' found folate levels to be "unrewarding." Foxet a15 evaluated 40 patients with "senile dementia" with laboratory tests including routine serum folate levels. Only two patients had low levels on initial testing, one of whom showed a normal level on retesting. The other was treated and showed no improvement in his dementia. Mulley,6 in reviewing the differential diagnosis of dementia, states "red-cell folate estimations should be limited to patients whose diet was grossly inadequate . . . and who have megaloblastic anemia." Despite the findings cited above, folate levels continue to be listed in the "standard dementia workup."' In the Geriatric Assessment Service of the San Francisco Institute on Aging at Mount Zion Hospital and Medical Center, we reviewed the results of serum and red-blood-cell folate testing for all patients undergoing outpatient geriatric assessment from January 1985 to July 1987. One hundred seventy-two patients with a mean age of 77.5 years were surveyed during that time. Nine of the 172 were found to have an anemia, but none of the nine was clearly macrocytic, and none was folate-deficient. One hundred two patients, or 59.3% of the total, had serum and red-cell folate levels obtained. Of the 102 patients tested, only one had a low serum folate, and one other had a low red-blood-cell folate. Neither of these patients was macrocytic or had an anemia, but both had cognitive deficits (as did 72% of the geriatric assessment patients) by the usual clinical and mental status testing criteria (Jacob Cognitive Capacity Screening and neuropsychological evaluation). The one patient with a low serum folate of 2.0 (normalis greater than 2.3 mg/mL) had a normal red-blood-cell folate level. The patient with the low red-blood-cell folate of 104 (normal is greater than 140 mg/mL cells) had a normal serum folate level. Thus our experience over a 14-year period in testing 102 patients suggests that routine serum of red-blood-cell folate as part of a geriatric assessment is of questionable benefit.