To the Editor: Independent of the use of antipsychotic medication, people with severe and persistent mental illness often have metabolic disarray, which increases cardiovascular risk (1). With use of antipsychotic medications, which are associated with weight gain and diabetes, the risk of metabolic abnormalities is even greater. For this reason the New York State Office of Mental Health (OMH) began an initiative in 2008 to summarize metabolic data from electronic patient records on a quarterly basis and provide the data to medical directors of OMH-operated hospitals as part of a quality improvement effort (2). Surprisingly, as detailed below, we found that cholesterol values among inpatients were significantly lower than the national norm.

In 2008 OMH began aggregating data from its electronic record system to report to OMH-operated hospitals on patient weight, fasting cholesterol, and fasting blood glucose (weight is monitored every three months, and fasting cholesterol and fasting blood glucose are monitored on admission—and annually for long-stay patients). Laboratory values for fasting cholesterol reported here were from licensed laboratories that follow national standards for processing and chain of possession. As of April 1, 2010, the 25th, 50th, and 75th percentiles for length of stay for adults (excluding individuals with a forensic status) were 3.4, 12.4, and 57.3 months, respectively.

Of the 3,792 adult (age 18 and older) inpatients hospitalized between May 2010 and August 2010 for whom we had data, 17% (N=648) had a total cholesterol level (measured on admission or within the past year, whichever was more recent) of ≥200, the cutoff for elevated cholesterol used by the National Cholesterol Education Program. The percentage of patients with elevated cholesterol ranged from 11% to 36% across the 17 OMH-operated hospitals. In contrast, U.S. national data indicate that among persons age 20 and older, from 40% to 51% have a cholesterol level of ≥200, depending on ethnic background (3).

We had expected that cholesterol levels would be higher than the national average in this population of hospitalized psychiatric patients, given the large proportion taking antipsychotic medications. We were surprised to find that cholesterol values for inpatients were significantly lower than the national norm, and in most hospitals, rates were less than half that seen in the adult general population. We therefore commenced investigations to verify data accuracy, including verifying the chain of possession of blood samples and confirming the accuracy of laboratory processes. In addition, we examined how patients' cholesterol levels varied by age, because an inpatient population younger than the national average might account for the difference. However, we found no significant difference in cholesterol levels across inpatient age groups. Another possible reason for lower-than-expected cholesterol levels is exposure of these inpatients to interventions that may be more likely to occur in an inpatient setting than in the general population, including healthy diets, prescription of statins, and adherence to prescribed statins. Given the number of the hospitals involved and the geographic diversity of the patients, the results appear unlikely to stem from local variations in diet or prescribing practices. Rather, it appears that inpatient stays, many of which last several months or more, offer some protection from the elevated cholesterol so prevalent in the United States.

The authors thank Sheila Donahue, M.A., Carol Lanzara, M.S., J.D., Hailing Li, M.S., Christina Mangurian, M.D., Gregory Miller, M.D., and Lloyd Sederer, M.D., for their efforts in support of this project.

The authors report no competing interests.