Screening prescription drugs for possible carcinogenicity: eleven to fifteen years of follow-up

Using computerized pharmacy records from 1969 to 1973 for a cohort of 143,574 members of the Kaiser Permanente Medical Care Program, we have been testing associations of 215 drugs or drug groups with subsequent incidence of cancer at 56 sites. This paper presents findings with follow-up through 1984. There were 227 statistically significant (P less than 0.05, two-tailed) associations: 170 positive, 57 negative. Some were undoubtedly chance findings; others were likely due to confounding by unmeasured covariables. However, several associations suggested hypotheses for further studies and/or the need for continued observation. Most notable among findings not previously reported were associations of several antibiotics, both oral and topical, with lung cancer. These associations could not be explained by indications for drug use or by differences in smoking habits between users and nonusers, and suggest a possible link between the occurrence of bacterial infections and risk for cancer. In general, our results continue to suggest that most medications used during that period did not affect cancer incidence substantially. However, for less frequently prescribed medications, our power to detect moderate increases in cancer risk was quite low.

Body burdens of lead in hypertensive nephropathy

Chronic lead exposure resulting in blood lead concentrations that exceed 1.93 mumol/l (40 micrograms/dl) or chelatable urinary lead excretion greater than 3.14 mumol (650 micrograms) per 72 h has been associated with renal disease. A previous study had found greater chelatable urine lead excretion in subjects with hypertension and renal failure than in controls with renal failure due to other causes, although mean blood lead concentrations averaged 0.92 mumol/l (19 micrograms/dl). To determine if chelatable urinary lead, blood lead, or the hematologic effect of lead (zinc protoporphyrin) were greater in hypertensive nephropathy (when hypertension precedes elevation of serum creatinine) than in other forms of mild renal failure, we compared 40 study subjects with hypertensive nephropathy to 24 controls having a similar degree of renal dysfunction due to causes other than hypertension. Lead burdens were similar in both the study and control groups as assessed by 72-h chelatable urinary lead excretion after intramuscular injection of calcium disodium EDTA (0.74 +/- 0.63 vs. 0.61 +/- 0.40 mumol per 72 h, respectively), and by blood lead (0.35 +/- 0.23 vs. 0.35 +/- 0.20 mumol/l). We conclude that subjects from a general population with hypertensive nephropathy do not have greater body burdens of lead than renal failure controls.

Evidence of genetic influence on central body fat in middle-aged twins

The heritability of centrally and peripherally deposited subcutaneous body fat, as measured by thickness of subscapular and triceps skinfolds respectively, was examined in 173 monozygotic and 178 dizygotic pairs of white male twins, ages 54 to 65 years, who participated in the second examination of the National Heart, Lung, and Blood Institute's Twin Study. The heritability of two indices of body fat distribution (subscapular/triceps ratio and subscapular-triceps difference) and two indices of overall obesity (body mass index and sum of skinfolds) were also assessed. Evidence for a genetic influence on central deposition of body fat was suggested in that the classical estimate of heritability for subscapular skinfold thickness was 0.77 (p less than 0.0001). After adjusting subscapular skinfold for the overall level of obesity, heritability was reduced but remained highly significant (0.40, p = 0.003). Heritability estimates for triceps skinfold thickness and for the two fat distribution indices were substantially lower and were not statistically significant after adjustment for overall obesity. High classical estimates of heritability were also observed for both measures of overall obesity: 0.70 for BMI and 0.73 for sum of skinfolds. However, these two estimates were biased upward because of lower total variances among monozygotic compared to dizygotic twins in this sample. The more conservative and unbiased among-component estimates also suggested substantial heritability for each measure (0.35, p = 0.08 and 0.53, p = 0.01, respectively). The heritability of overall obesity emphasizes the importance of adjusting measures of fat distribution for overall obesity before assessing its heritability.

Precursors of essential hypertension. The role of body fat distribution pattern

Progression from normotension between 1964 and 1972 to essential hypertension by age 55 years was documented in 1,031 adult members of the Kaiser Permanente Medical Care Program (Northern California region) from computerized multiphasic health checkup records and medical record review. Each case was matched to a persistently normotensive control on age, sex, race, number, and dates of multiphasics. In 609 pairs with baseline measurements of subscapular and triceps skinfolds, mean interval from baseline to the case's first hypertensive multiphasic was 5.7 years, and mean age at onset of hypertension was 47 years. Baseline measures of body mass index, subscapular skinfold, and triceps skinfold were each predictive of development of hypertension (odds ratios 3.85, 3.75, and 2.29 respectively, comparing highest with lowest quintiles, p less than 0.0001 for each). When both skinfolds were included in the same model, subscapular skinfold was highly predictive and triceps skinfold was no longer related to risk. When the authors controlled for overall obesity (body mass index), subscapular skinfold remained highly predictive (p less than 0.0001). In 330 pairs who also had skinfold measurements at the hypertensive multiphasic, weight gain was a strong predictor of hypertension. Increase in subscapular skinfold conferred a small increase in risk in women only. The authors conclude that centrally deposited body fat increases risk for developing essential hypertension independent of the overall level of obesity, while peripherally deposited fat does not.

Lower-extremity amputation in people with diabetes. Epidemiology and prevention

The age-adjusted rate of lower-extremity amputation (LEA) in the diabetic population is approximately 15 times that of the nondiabetic population. Over 50,000 LEAs were performed on individuals with diabetes in the United States in 1985. Among individuals with diabetes, peripheral neuropathy and peripheral vascular disease (PVD) are major predisposing factors for LEA. Lack of adequate foot care and infection are additional risk factors. Several large clinical centers have experienced a 44-85% reduction in the rate of amputations among individuals with diabetes after the implementation of improved foot-care programs. Programs to reduce amputations among people with diabetes in primary-care settings should identify those at high risk; clinically evaluate individuals to determine specific risk status; ensure appropriate preventive therapy, treatment for foot problems, and follow-up; provide patient education; and, when necessary, refer patients to specialists, including health-care professionals for diagnostic and therapeutic interventions and shoe fitters for proper footwear. Programs should monitor and evaluate their activities and outcomes. Many issues related to the etiology and prevention of LEAs require further research.

Precursors of essential hypertension: body weight, alcohol and salt use, and parental history of hypertension

Body mass index, alcohol and salt consumption, and parental history of hypertension were examined as possible predictors of the development of essential hypertension in 1,031 persons, ages 30-49 years at entry, with documented normotension followed by documented hypertension after a mean interval of 6 years. In a comparison with 1,031 matched persistently normotensive persons initial body mass index and percentage increase in body mass index were each predictive of hypertension. Consumption of three or more alcoholic drinks a day at baseline was also predictive, more so if this level of intake persisted than if it diminished. Heavy salt intake as crudely estimated at baseline by one question was also associated with the development of hypertension. Parental history of hypertension was also predictive, more so for hypertension in the mother than for hypertension in the father, and the association was apparent only in female subjects. These characteristics at baseline showed independent associations with subsequent hypertension in multivariate analysis. When follow-up data were included in the multivariate analysis, alcohol consumption at the hypertensive examination was much more strongly related than at the baseline examination, suggesting a short-term effect, and heavy salt consumption was no longer predictive, possibly because of a marked loss of subjects due to missing follow-up data. This large study confirms longitudinally the importance of obesity, weight gain during adulthood, alcohol, family history, and, to some extent, salt as predictive and possibly causal factors for essential hypertension.

Epidemiologic evidence of an association between body iron stores and risk of cancer

Biologic evidence suggests that high body iron stores could promote development of cancer. Because a previous study had shown an association between prescribed iron medication and lung cancer risk in men, we examined recent iron use as well as 2 additional indirect measures of body iron stores, anemia and the total iron-binding capacity (TIBC) of plasma, in relation to subsequent risk of cancer in a larger cohort of 174,507 persons. Women, but not men, who reported recent iron use had a lower risk of lung cancer than those who did not [RR = 0.60, 95% confidence limits (CL) 0.37, 0.97] after adjustment for age and cigarette smoking. Women who had used iron appeared to remain relatively iron-depleted. Risk for other cancers was slightly, but not significantly, lower in women who used iron. Anemia (hemoglobin less than 12 g) was also associated with lower risk of lung cancer in women (RR = 0.61, 95% CL 0.61, 0.98), but not in men. TIBC, which is inversely related to body iron stores, was inversely related to risk of lung cancer in women in a graded fashion (RR = 0.41, 95% CL 0.23, 0.73 comparing highest with lowest quartile). In men, a protective effect of higher TIBC against lung cancer was suggested, but did not reach statistical significance. These indirect measures of body iron stores appeared to reflect iron stores better in women than in men, probably because variability in iron stores is greater in women and iron deficiency more prevalent. A possible alternative explanation for our findings is incomplete adjustment for the confounding effects of cigarette smoking. This could apply to iron use and hemoglobin level which were related to smoking, but not to TIBC, which was not. These data, which indicate lower risk of cancer in iron-depleted women, lend epidemiologic support to the hypothesis that high iron stores may increase cancer risk, at least for lung cancer.

Sigmoidoscopy and mortality from colorectal cancer: the Kaiser Permanente Multiphasic Evaluation Study

The Kaiser Permanente Multiphasic Evaluation Study is often cited as evidence from a randomized trial that screening sigmoidoscopy reduces mortality from colorectal cancer. To examine the role of sigmoidoscopy in this reduction, we reviewed the 110 incident cases of colorectal cancer occurring among the 10,713 subjects from randomization in 1964 through 1982. Tumor stage at diagnosis, location, mode of discovery, and current mortality status were determined for each. We also reanalyzed chart review data for the years 1965 through 1974 to assess the difference in exposure to sigmoidoscopy between groups. Study group subjects, who were urged to have annual multiphasic health checkups (MHC), had both a lower cumulative incidence (4.3 vs 6.7 cases per 1000 persons) and a better stage distribution (86 vs 54% Stage B or better) than nonurged control subjects for colorectal cancers arising within reach of the sigmoidoscope. The lowered incidence accounted for two-thirds of the total difference in mortality. No appreciable difference in removal of colorectal polyps was seen between groups. Only a slight excess in exposure to sigmoidoscopy was seen in the study group (30 vs 25% of subjects examined at least once between 1965 and 1974), which was unlikely to account for more than a small fraction of the study group's decrease in mortality. Although the Multiphasic Evaluation Study did find a significantly lower mortality from colorectal cancer in the study group, it did not achieve a substantial difference in exposure to sigmoidoscopy. Its results are therefore inconclusive with respect to sigmoidoscopy and should not be used as evidence either for or against sigmoidoscopic screening.

Concordance for type 2 (non-insulin-dependent) diabetes mellitus in male twins

Concordance for Type 2 (non-insulin-dependent) diabetes was determined in 250 monozygotic and 264 dizygotic white male twin pairs who participated in the National Heart, Lung, and Blood Institute Twin Study. These twins were born between 1917 and 1927 and were identified from military records without regard to disease status. We examined surviving members of the cohort twice--at mean ages of 47 and 57 years--and obtained 1-h post-load glucose tests and medication histories. Diagnostic criteria for Type 2 diabetes included a glucose value greater than or equal to 13.9 mmol/l or current use of antidiabetic medication; possible Type 1 (insulin-dependent) diabetic twins were excluded. A strong genetic predisposition to Type 2 diabetes was suggested by 3 lines of evidence from the second examination: (1) 58% of monozygotic co-twins of diabetic twins were themselves diabetic compared with an expected prevalence of 10%; (2) only 1 of 15 originally disease-discordant, monozygotic twin pairs remained discordant for diabetes; and (3) 65% of non-diabetic monozygotic co-twins of diabetic twins had elevated glucose values. Because concordance for diabetes was less than 100% for twins aged 52-65 years and because twins varied in age at onset of disease, non-genetic factors may also influence diabetes development. Among the 19 monozygotic twins pairs discordant for diabetes, diabetic twins did not differ from their non-diabetic co-twins in obesity, diet, alcohol consumption, or education. However, compared with unrelated non-diabetic twins of the same ages, non-diabetic co-twins of diabetic twins gained more weight as adults (p less than 0.02) and had higher glucose levels (p less than 0.03).

Environmental and behavioral determinants of fasting plasma glucose in women. A matched co-twin analysis

Comparing results of an unmatched analysis in a sample of unrelated persons created by randomly selecting one twin from each pair with results of a matched monozygotic co-twin analysis provides a means of distinguishing environmental influences from genetic influences on a dependent variable of interest. Determinants of the fasting plasma glucose level were investigated in a large sample of adult female nondiabetic twins from the Kaiser Permanente Twin Registry. In the subsample of unrelated persons, fasting plasma glucose was significantly and positively associated with age, body mass index, uric acid level, and alcohol intake and was significantly inversely related to leisure time physical activity. When monozygotic twins were treated as matched pairs, only the effects of body mass index and alcohol consumption remained significant, whereas the association with leisure time physical activity was greatly reduced and no longer statistically significant. These results suggest that behavioral and therefore alterable aspects of obesity and alcohol consumption influence the fasting plasma glucose level; alcohol consumption may be a risk factor for non-insulin-dependent diabetes mellitus; and common underlying genetic factor(s) may be responsible for the negative association between physical activity and fasting plasma glucose.