Appendix I. Laboratory Values
Population norms can be established for any physiologic parameter and its laboratory measurement. The normal value for a laboratory measurement is usually defined as its mean value ± 2 SD (which includes 95% of results) in a population of healthy persons. Thus, 5% of results obtained from healthy persons are classified as "abnormal," even though they merely represent variability within a normal population. If each test in a battery of tests is independent, the probability of a healthy person having completely normal results is relatively low (eg, 54% in a battery of 12 tests).
The determination of normal laboratory values in the elderly is complicated by the high prevalence of disease (latent or overt and multisystem) and by age-related physiologic and anatomic changes. In addition, as persons age, they become less like each other. Thus, extrapolating from population norms to individuals is less useful for elderly than for younger persons. With age, organ function declines, but the rate of decline varies from person to person and from system to system within one person. For example, between ages 30 and 70 years, nerve conduction velocity decreases by only about 10%, but renal function decreases on average by nearly 40%. Nonetheless, age-related declines in cardiac, pulmonary, renal, and metabolic function can be correlated with changes in normal laboratory values (see Table below). Large or sudden changes in function frequently signify a disorder.
Determining the likelihood that an abnormal laboratory test result represents a particular disorder may be difficult in the elderly. Likelihood depends largely on the pretest probability of the disorder (based on clinical findings and prevalence of the disorder) and the operating characteristics of the test (sensitivity, specificity, likelihood ratios). Pretest probability and a test's operating characteristics can be difficult to determine because precise information about prevalence, sensitivity, specificity, and likelihood ratios for the elderly is often unavailable. Inaccurate interpretations tend to lead to excess false-positive results.
Drugs commonly used by the elderly may spuriously alter the results of laboratory tests. For example, isoniazid, levodopa, morphine, vitamin C, nalidixic acid, and penicillin G may lead to false-positive urine glucose results. Levodopa may produce a spurious increase in serum bilirubin and uric acid, also leading to false-positive results.
A low serum albumin level in healthy elderly persons is usually due to diet. However, low levels may also indicate serious disease, especially when accompanied by undernutrition.
Fasting blood glucose levels increase with age, although values remain within the nondiabetic range. Glucose tolerance decreases gradually with age. However, lack of exercise, obesity, and the use of some drugs may affect glucose tolerance more than aging per se.
The renal threshold for glucose decreases with age. Consequently, glucose may be detected in the urine of elderly patients, even when blood glucose is considerably less than 200 mg/dL (11.1 mmol/L).
The erythrocyte sedimentation rate (ESR) is not affected by aging per se, but the ESR tends to be higher in the elderly, probably because prevalence of disease is higher. Elevations in ESR usually indicate elevations in serum proteins due to chronic inflammatory activity (including infections), monoclonal gammopathy, or cancer. If elderly patients have an unexplained elevation in the ESR, a monoclonal gammopathy should be sought. This disorder occurs in about 10% of persons aged 62 to 95 who are apparently healthy but in about 35% of those who have an elevated ESR.
The prostate-specific antigen (PSA) level increases with age, and it can be increased by benign conditions. Consequently, its value in screening for prostate cancer is debatable. For men with an abnormal prostate detected by rectal examination, PSA may be measured, but this measurement may be less valuable than the ratio of PSA to gland size, and abnormal results are likely to lead to more expensive, invasive tests. An increase in PSA of > 5 to 8% in a year may be highly significant and usually requires rectal ultrasonography and biopsy. Before PSA is measured, patients should be warned about the possibilities of ambiguous and false-positive results.
CBCs and measurement of thyroid-stimulating hormone may be useful as screening tests in the elderly because of the high incidence of occult anemia and hypothyroidism.
Protein:creatinine ratios in single-voided urine samples can be used to estimate the magnitude of proteinuria (the incidence of trace proteinuria decreases during a person's 20s, then increases during old age). These spot measurements correlate with determinations of 24-hour urinary protein excretion and are particularly useful in elderly patients when 24-hour specimens are difficult to obtain. Protein:creatinine ratios > 3.0 indicate massive proteinuria (> 3.5 g/24 hour); ratios < 0.2 usually indicate insignificant proteinuria.
Creatinine clearance decreases with age, but serum creatinine remains relatively stable because most elderly persons lose muscle mass as they age. Therefore, when serum creatinine is measured, renal function in the elderly tends to be overestimated, and creatinine clearance should be used to assess renal function, according to the Cockcroft-Gault formula: (see also page 59)
For women, the calculated value is multiplied by 0.85. Creatinine clearance may need confirmation (eg, with timed collections), particularly in patients who are malnourished, who have lost muscle mass, who are severely ill, or who are very elderly.