Alcohol intake and subsequent mortality: findings from the NHANES I Follow-up Study.

OBJECTIVE: Because past research has focused primarily on populations of middle-aged men, the relationship between alcohol and mortality among women and the elderly has been less well substantiated METHOD: We examined the relationship between alcohol intake and mortality using data from the NHANES I Epidemiologic Follow-Up Study. Total mortality was examined for both sexes (N = 4,614 women, 3,573 men), but ischemic heart disease (IHD) mortality was examined only for men because the number of deaths was too small in women. Proportional hazards modeling was used to adjust for the baseline characteristics of age, race, education, body weight, smoking and physical activity. RESULTS: For men aged 40 to 64, the adjusted relative risks (RR) of death for drinking levels of .5 drinks/day, .5 to < 2 and > or = 2 (compared to the nondrinking reference group) were 0.8 (95% Confidence Interval: 0.6, 1.1), 0.9 (CI:0.6, 1.2) and 1.2 (CI: 0.9, 1.6); RRs of IHD mortality were 0.6 (CI: 0.4, 0.9), 0.5 (CI: 0.3, 0.9) and 0.7 (CI: 0.5, 1.2). For women aged 40 to 64, the RRs for death for the same exposure categories were 1.2 (CI: 0.9, 1.6), 0.9 (CI: 0.6, 1.4) and 1.9 (CI: 1.2, 3.0). Among both sexes 65 years and older, consumption of < 2 drinks/day was associated with about a 20% decrease in total mortality and IHD mortality (men only). However, this protective effect disappeared after exclusion of those with pre-existing disease. CONCLUSIONS: Our findings support a protective effect of moderate alcohol intake on IHD mortality in middle-aged men. Among both men and women, there was little evidence of a protective association between moderate alcohol intake and total mortality after excluding those with pre-existing disease.

Recreational physical activity and ten-year weight change in a US national cohort.

Clinical research has established that increases in physical activity cause weight loss among the obese, but less is known about the influence of physical activity on longer-term weight change in the general population. Data from the NHANES-I Epidemiologic Follow-up Study (1971-1975 to 1982-1984) were used to examine the relationship between self-reported recreational physical activity level (low, medium, high) and measured weight change after ten years among 3515 men and 5810 women aged 25-74 years. Cross-sectional analyses at both the baseline and follow-up surveys revealed that recreational physical activity was inversely related to body weight. Low recreational physical activity reported at the follow-up survey was strongly related to major weight gain (> 13 kg) that had occurred during the preceding ten years. The estimated relative risk of major weight gain for those in the low activity level at the follow-up survey compared to those in the high activity level was 3.1 (95% Cl = 1.6-6.0) in men and 3.8 (2.3-6.5) in women. In addition, the relative risk for persons whose activity level was low at both the baseline and follow-up surveys was 2.3 (0.9-5.8) in men and 7.1 (2.2-23.3) in women. However, no relationship was found between baseline physical activity level and subsequent weight gain among either men or women. The lack of a relationship may be due to mis-specification of physical activity because of changes in activity over time. These findings suggest that low physical activity may be both a cause and a consequence of weight gain.(ABSTRACT TRUNCATED AT 250 WORDS)

Birth-weight-adjusted infant mortality in evaluations of perinatal care: towards a useful summary measure.

Since birth-weight-specific infant mortality rates are widely considered to be measures of the effects of perinatal medical care, birth-weight-adjusted and birth-weight-specific infant mortality rates have often been used in comparisons across hospitals and geographic areas. Wilcox and Russell have provided a model which leads to the conclusion that birth weight adjustment is biased against populations with heavier birth weights and that birth-weight-specific infant mortality rates can yield misleading results. Nevertheless, evaluators and health planners still need a summary measure of the infant mortality rate in which some sort of birth weight adjustment is used to generate an (appropriately) weighted mean of birth-weight-specific relative risks. We used the 1983-85 national linked files of live births and infant deaths to investigate two new methods that extend the Wilcox-Russell approach, 'mean adjustment' and 'Z-adjustment', to compare birth-weight-specific infant mortality rates among the white populations of the states. Colorado was used as the reference population in logistic regression analyses. Statistical interactions between state and birth weight on the conventional kilogram scale, the mean-adjusted scale, and the Z scale were examined. There were substantial interactions with birth weight on all three scales. In addition, when Colorado was used as the reference population, mean adjustment shifted the odds ratios downward to implausibly low values, especially at lower weights. The Z-adjustment method incorporating the Wilcox-Russell approach appears to be a useful alternative to birth weight adjustment. However, because birth-weight-specific mortality rates do not differ uniformly across all birth weight groups, multiple summary measures are needed.

Decreased risk of stroke among postmenopausal hormone users. Results from a national cohort.

OBJECTIVE: To assess the impact of postmenopausal hormone use on the risk of stroke incidence and stroke mortality. DESIGN: Longitudinal study consisting of three data collection waves. The average follow up for cohort members was 11.9 years (maximum, 16.3 years). Cox proportional hazards regression models were used to estimate the relative risk of stroke for postmenopausal hormone ever-users compared with never-users. PARTICIPANTS: A national sample of 1910 (of 2371 eligible) white postmenopausal women who were 55 to 74 years old when examined in 1971 through 1975 as part of the first National Health and Nutrition Examination Survey and who did not report a history of stroke at that time. MAIN OUTCOME MEASURE: The main outcome measure was incident stroke (fatal and nonfatal). Events were determined from discharge diagnosis information coded from hospital and nursing home records and cause of death information coded from death certificates collected during the follow-up period (1971 through 1987). RESULTS: There were 250 incident cases of stroke identified, including 64 deaths with stroke listed as the underlying cause. The age-adjusted incidence rate of stroke among postmenopausal hormone ever-users was 82 per 10,000 woman-years of follow-up compared with 124 per 10,000 among never-users. Postmenopausal hormone use remained a protective factor against stroke incidence (relative risk, 0.69; 95% confidence interval, 0.47 to 1.00) and stroke mortality (relative risk, 0.37; 95% confidence interval, 0.14 to 0.92) after adjusting for the baseline risk factors of age, systolic blood pressure, diabetes, body mass index, smoking, history of hypertension and heart attack, and socioeconomic status. CONCLUSIONS: The results suggest that postmenopausal hormone use is associated with a decrease in risk of stroke incidence and mortality in white postmenopausal women.

Weight loss and mortality in a national cohort of adults, 1971-1987.

Although obesity is a risk factor for mortality, evidence that weight loss improves survival is limited. The relation between self-reported previous maximum weight, weight loss, and subsequent mortality was examined in 2,140 men and 2,550 women aged 45-74 years who participated in the First National Health and Nutrition Examination Survey (1971-1975) and survived the next 5 years. Vital status was determined through 1987. Among men and women whose maximum body mass index (weight (kg)/height (m)2) was between 26 and 29, risk of death increased with increasing weight loss, after adjustment for age, race, smoking, parity, preexisting illnesses, and maximum body mass index. Subjects who lost 15% or more of their maximum weight had over twice the mortality risk of those who lost less than 5%. At maximum body mass indices of 29 or higher, mortality risk increased with the amount of weight lost in women, but weight loss of 5% to < 15% appeared to lessen mortality risk in men. Generalization from these results is limited by the older age range of the sample and the inability to adequately distinguish voluntary from involuntary weight loss in this study. However, these findings suggest that prevention of severe overweight may be more generally effective than weight loss in reducing obesity-related mortality in the US population.

Plan and operation of the NHANES I Epidemiologic Followup Study, 1987.

This report describes the plan and operation for the 1987 data collection wave of the Epidemiologic Followup to the first National Health and Nutrition Examination Survey (NHANES I). Tracing and data collection were conducted on 11,750 persons 25-74 years of age at NHANES I who were not known to be deceased in the 1982-84 and 1986 data collection waves of the NHANES I Epidemiologic Followup Study.

Triplets and higher-order multiple births. Time trends and infant mortality.

OBJECTIVES: To describe changes in rates of higher-order multiple births (triplets and higher) between 1972 and 1989, to compare infant mortality rates in infants of higher-order multiple births and singletons born from 1983 through 1985, and to compare infant mortality rates among higher-order multiples born from 1983 through 1985 with rates among those born in 1960. RESEARCH DESIGN: Population-based analysis of live births (1972 through 1989) and infant deaths (1960 and 1983 through 1985) in the United States. The rate of higher-order multiple births was calculated per 100,000 live births. DATA SOURCE: Computerized national natality files for 1972 through 1989 and national linked birth/infant death data sets for 1960 and 1983 through 1985 from the National Center for Health Statistics, Centers for Disease Control. POPULATION: Live births to white and black women in the United States. INTERVENTIONS: None. MAIN RESULTS: Between 1972 through 1974 and 1985 through 1989 the rate of higher-order multiple births increased by 113% among infants of white mothers and by 22% among infants of black mothers. In whites the increase was mostly age specific and was not due to the upward shift in the maternal age distribution. The increase was particularly large in white women aged 30 through 34 years (152%) and 35 through 39 years (165%) and in more highly educated mothers. In blacks the modest increase in the rate of higher-order multiple births was mostly due to an upward shift in the maternal age distribution. From 1983 through 1985, mortality of infants of higher-order multiple births was about 15 times that of singletons. This was due almost entirely to the lower birth weight distribution of infants of higher-order multiple births. Their weight-specific mortality compared favorably with that of singletons. At 500 through 999 g, mortality was about the same. In weight categories between 1000 and 1999 g, mortality rates in higher-order multiple births were much lower: weight-specific relative risks ranged from 0.30 to 0.73. Between 1960 and 1983 through 1985 infant mortality in higher-order multiple births declined by about 50%. CONCLUSIONS: It is likely that much of the increase in the incidence of higher-order multiple births is due to the rise in the use of ovulation-inducing drugs for the treatment of infertility. This increase and the decline in mortality risk have created a much greater need for medical and social services for infants of higher-order multiple births and their families.

Mortality among infants of black as compared with white college-educated parents.

BACKGROUND: In the United States, black infants are twice as likely to die as white infants; this difference reflects both black infants' higher rates of low birth weight and the higher mortality among black infants of normal birth weight. We studied mortality in infants born to college-educated parents in order to investigate this gap while controlling for sociodemographic variables. METHODS: We used the National Linked Birth and Infant Death Files for 1983 through 1985 to calculate infant mortality rates for children born to college-educated parents. The study population consisted of 865, 128 white infants and 42,230 black infants. A separate effect of birth weight was assessed by examining mortality rates before and after the exclusion of infants weighing less than 2500 g at birth (low-birth-weight infants). RESULTS: In this population, the infant mortality rate was 10.2 per 1000 live births for black infants and 5.4 per 1000 live births for white infants; the adjusted odds ratio for death among black infants was 1.82 (95 percent confidence interval, 1.64 to 2.01). The rate of low birth weight was more than twice as high among blacks (7 percent) as among whites (3 percent), although the mortality rate in this group was not higher among blacks than among whites. Black infants were three times as likely as white infants to die of causes attributable to perinatal events, including prematurity. They were no more likely to die of the sudden infant death syndrome. After the exclusion of low-birth-weight infants, the mortality rates for black and white infants were equal. CONCLUSIONS: In contrast to black infants in the general population, black infants born to college-educated parents have higher mortality rates than similar white infants only because of their higher rates of low birth weight. Black and white infants of normal birth weight have equivalent mortality rates.

The low cholesterol-mortality association in a national cohort.

The relationship of low serum cholesterol and mortality was examined in data from the NHANES I Epidemiologic Followup Study (NHEFS) for 10,295 persons aged 35-74, 5833 women with 1281 deaths and 4462 men with 1748 deaths (mean (followup = 14.1 years). Serum cholesterol below 4.1 mmol/l was associated with increased risk of death in comparison with serum cholesterol of 4.1-5.1 mmol/l (relative risk (RR) for women = 1.7, 95% confidence interval (CI) = (1.2, 2.3); for men RR = 1.4, CI = (1.1, 1.7)). However, the low serum cholesterol-mortality relationship was modified by time, age, and among older persons, activity level. The low serum cholesterol-mortality association was strongest in the first 10 years of followup. Moreover, this relationship occurred primarily among older persons (RR for low serum cholesterol for women 35-59 = 1.0 (0.6, 1.8), for women 70-74, RR = 2.1 (1.2, 3.7); RR for low serum cholesterol for men 35-59 = 1.2 (0.8, 2.0), for men 70-74, RR = 1.9 (1.3, 2.7)). Among older persons, however, the low serum cholesterol-mortality association was confined only to those with low activity at baseline. Factors related to underlying health status, rather than a mortality-enhancing effect of low cholesterol, likely accounts for the excess risk of death among persons with low cholesterol. The observed low cholesterol-mortality association therefore should not discourage public health programs directed at lowering serum cholesterol.

Regional variation in ischemic heart disease incidence.

This study examines the relationship between cardiovascular risk factors and regional variation in IHD incidence among white males 55-74 years of age from the NHANES I Epidemiologic Followup Study. The age-adjusted IHD incidence rate was lowest in the west (31.3 per 1000 persons years of followup). The rates in the northeast, midwest, and south were similar and so they were combined into one region, the non-west, with a rate of 42.4. Differences in risk factors (smoking, educational level, hypertension, serum cholesterol, diabetes mellitus, and body mass index) did not explain the regional differences in IHD incidence. After adjusting for baseline risk factors using proportional hazards model, the risk of IHD incidence was still 38% higher in the non-west compared to the west. However, the effect of hypertension, diabetes, and body mass index on IHD incidence varied by region.

Is weight loss a modifier of the cholesterol-heart disease relationship in older persons? Data from the NHANES I Epidemiologic Follow-up Study.

The relationship between cholesterol and 14-year incidence of coronary heart disease was compared for men and women of two age groups, 25 to 64 years and 65 to 74 years. While cholesterol levels of 6.2 mmol/L or higher were associated with a risk of coronary heart disease in the younger group, this was not true for either men or women aged 65 to 74. Further analyses for older persons showed that weight loss modified the cholesterol-heart disease relationship. Those with stable weight showed a positive relationship between cholesterol and coronary heart disease, similar to the younger age group (relative risk [RR] = 1.8 [95% confidence interval: 1.1, 2.9] for men; RR = 1.6 [.7, 3.4] for women). Among those with a weight loss of 10% or more, the relationship of cholesterol to heart disease was inverse (RR = .8 [.5, 1.2] for men; RR = .6 [.3, 1.0] for women). These data suggest that the relationship of cholesterol to coronary disease in healthier older persons may be similar to that in younger persons, and that health status should be considered in analyses of cholesterol risk in old age.

Health service areas for the United States.

The objectives of this report are to document methods used to identify health service areas for the United States and to describe and evaluate these areas. A health service area is defined as one or more counties that are relatively self-contained with respect to the provision of routine hospital care. Service areas that include more than one county are characterized by travel between the counties for routine hospital care.

Postural change in blood pressure associated with age and systolic blood pressure. The National Health and Nutrition Examination Survey II.

The prevalence of postural change in blood pressure and its association with age and systolic blood pressure were examined in data from 8,574 White nondiabetic persons aged 25-74 who participated in the second National Health and Nutrition Examination Survey (1976-1980). Postural change in blood pressure was defined as a drop of 20 mm Hg or more on change from supine to seated position. In subjects on no antihypertensive medications (n = 7,316), the prevalence of postural change in blood pressure increased with older age and with higher blood pressure levels, regardless of age. However, systolic blood pressure levels also increased with age. In logistic regression models, level of supine systolic blood pressure was strongly related to postural change in blood pressure (Relative odds (RO) = 1.59, 95% confidence interval (CI) = 1.49, 1.70 for a 10 mm Hg increase in systolic blood pressure) whereas age was not related to postural change in blood pressure (RO for age = 1.07, Cl = .89, 1.19 for a 10-year increase in age). Results were similar for those medicated for hypertension. All results were unchanged by addition of health status indicators, including reports of hospitalization and number of medical conditions, to the model. These data suggest that the age-related increase in the prevalence of postural hypotension previously reported may be partially explained by age-associated increases in systolic blood pressure.

Is the serum cholesterol-coronary heart disease relationship modified by activity level in older persons?

Although coronary heart disease remains a leading cause of death and disability in old age, the relationship of serum cholesterol level to risk of coronary heart disease in old age is controversial. Data for 2,388 white persons aged 65-74 who participated in the National Health and Nutrition Examination Survey (NHANES) I Epidemiologic Follow-up Study (NHEFS) were examined to determine the relationship of serum cholesterol level to coronary heart disease incidence and whether activity level would modify this relationship. While there was no overall relationship between serum cholesterol level and coronary heart disease risk in either men or women, the relationship between serum cholesterol level and coronary heart disease differed within activity groups. For persons who were more active, serum cholesterol level was associated with a graded increase in risk of coronary heart disease, from 1.3 (95% CI 0.7, 2.3) in those with serum cholesterol level of 4.7-5.1 to 1.7 in those with serum cholesterol level of 6.2 mmol/L or more (95% CI 1.0, 2.7), when compared with those with serum cholesterol level below 4.7. For the least active persons, all levels of cholesterol were associated with a significant inverse relative risk, including cholesterol of 6.2 mmol/L or more (Relative risk = 0.4 (95% CI 0.2, 0.7]. These data suggest that factors such as activity level may modify the serum cholesterol-coronary heart disease association in old age. The serum cholesterol-coronary heart disease association in more active older persons resembles that seen in younger populations, whereas the association in less active persons is that of serum cholesterol level and risk of cancer or death. The modification of the serum cholesterol-coronary heart disease association by activity level may have implications for appropriate clinical management as well as appropriate design of research studies of this association.

Double jeopardy: twin infant mortality in the United States, 1983 and 1984.

The United States Linked Birth/Infant Death Data Sets: 1983 and 1984 Birth Cohorts from the National Center for Health Statistics were used to identify maternal and infant characteristics related to twin infant mortality; 41,554 white and 10,062 black live-born matched twin pairs were evaluated. Twin birth weight distribution was skewed with 48% of white and 63% of black twins born weighing less than 2500 gm. Overall infant mortality rates were 47.1 and 79.3 deaths per 1000 live births for white and black twins, respectively (five times the rates for singletons). Three fourths of deaths were among twins weighing less than 1500 gm. White like-gender twins had about twice the risk of both twins dying compared with unlike-gender twins. Likewise, white twin pairs with greater than 25% birth weight disparity had a 40% to 80% increased risk of both twins dying, compared with twins whose weights were within 10% of each other. Twins born to high-risk women (on the basis of demographic factors) were twice as likely to die as twins born to low-risk women. Thus strategies to decrease twin infant mortality must address both maternal and infant risk factors.

Smoking cessation and severity of weight gain in a national cohort.

BACKGROUND: Many believe that the prospect of weight gain discourages smokers from quitting. Accurate estimates of the weight gain related to the cessation of smoking in the general population are not available, however. METHODS: We related changes in body weight to changes in smoking status in adults 25 to 74 years of age who were weighed in the First National Health and Nutrition Examination Survey (NHANES I, 1971 to 1975) and then weighed a second time in the NHANES I Epidemiologic Follow-up Study (1982 to 1984). The cohort included continuing smokers (748 men and 1137 women) and those who had quit smoking for a year or more (409 men and 359 women). RESULTS: The mean weight gain attributable to the cessation of smoking, as adjusted for age, race, level of education, alcohol use, illnesses related to change in weight, base-line weight, and physical activity, was 2.8 kg in men and 3.8 kg in women. Major weight gain (greater than 13 kg) occurred in 9.8 percent of the men and 13.4 percent of the women who quit smoking. The relative risk of major weight gain in those who quit smoking (as compared with those who continued to smoke) was 8.1 (95 percent confidence interval, 4.4 to 14.9) in men and 5.8 (95 percent confidence interval, 3.7 to 9.1) in women, and it remained high regardless of the duration of cessation. For both sexes, blacks, people under the age of 55, and people who smoked 15 cigarettes or more per day were at higher risk of major weight gain after quitting smoking. Although at base line the smokers weighed less than those who had never smoked, they weighed nearly the same at follow-up. CONCLUSIONS: Major weight gain is strongly related to smoking cessation, but it occurs in only a minority of those who stop smoking. Weight gain is not likely to negate the health benefits of smoking cessation, but its cosmetic effects may interfere with attempts to quit. Effective methods of weight control are therefore needed for smokers trying to quit.

Differences in infant mortality by race, nativity status, and other maternal characteristics.

The objective of this study was to examine the effects of nativity status (native vs foreign born) and other maternal characteristics (age, parity, education, and marital status) on infant, neonatal, and postneonatal mortality among white and black mothers. The design of this nonrandomized cohort study was based on birth and death certificates. The setting involved live births among US residents (excluding California, Texas, and Washington) in 1983 and 1984. The participants included white mothers with 4.4 million births and black mothers with 926,000 births in single deliveries. There were no interventions. With regard to measurements (the main results), after adjusting for other risk factors, neonatal mortality risk was 22% lower among the black foreign-born mothers than among the black native-born mothers, while among white infants, there was no risk difference by nativity. Relative risks were more similar for postneonatal mortality, ie, 24% lower among black foreign-born mothers and 20% lower among white foreign-born mothers. Combining the several categories of risk factors into three broad maternal risk groups, there was a near-doubling of black and near-tripling of white infant mortality rates between the low and high levels of maternal risk. We concluded that if the infant mortality rate in the low-risk groups could be achieved by the moderate- and high-risk groups, there would be a 30% reduction in infant deaths within each race. Since the black infant mortality rate is twice the white infant mortality rate and black foreign-born mothers have much lower rates than black native-born mothers, it is likely that further improvement is possible among black infants.

Reduction of cardiovascular disease-related mortality among postmenopausal women who use hormones: evidence from a national cohort.

A national sample of 1944 white menopausal women greater than or equal to 55 years old from the epidemiologic follow-up of participants in the National Health and Nutrition Examination Survey was reviewed to investigate the role of hormone therapy in altering the risk of death from cardiovascular disease. Women in the study were observed for up to 16 years after the baseline survey in 1971 to 1975. By 1987 631 women had died; 347 of these deaths were due to cardiovascular disease. History of diabetes (relative risk, 2.38; 95% confidence interval 1.73 to 3.26), previous myocardial infarction (relative risk, 2.12; 95% confidence interval 1.56 to 2.86), smoking (relative risk, 2.18; 95% confidence interval, 1.69 to 2.81), and elevated blood pressure (relative risk, 1.49; 95% confidence interval, 1.14 to 1.94) were strong predictors of cardiovascular disease-related death in this cohort. After adjusting for known cardiovascular disease risk factors (smoking, cholesterol, body mass index, blood pressure, previous myocardial infarction, history of diabetes, age) and education, the use of postmenopausal hormones was associated with a reduced risk of death from cardiovascular disease (relative risk, 0.66; 95% confidence interval, 0.48 to 0.90). The same protective effect provided by postmenopausal hormone therapy was seen in women who experienced natural menopause (relative risk, 0.69; 95% confidence interval, 0.45 to 1.06).

Comparison of infant mortality among twins and singletons: United States 1960 and 1983.

Infant mortality among US black and white twins and singletons was compared for 1960 and 1983 using the Linked Birth/Infant Death Data Sets from the National Center for Health Statistics. Both twin and singleton infant mortality rates showed impressive declines since 1960 but almost all of the improvement in survival for both twins and singletons was related to increased birth weight-specific survival rather than improved birth weight distribution. One-half of white twins and two-thirds of black twins weighed less than 2,500 g at birth, and 9% of white twin births and 16% of black twin births were in the very low (less than 1,500g) birth weight category. In 1983, twin infant mortality rates were still four to five times that of singletons. However, twins had a survival advantage in the 1,250-3,000 g range, which persisted after adjustment for gestational age. Cause-specific mortality among twins was considerably higher for every major cause of death: twin mortality risks due to newborn respiratory disease, maternal causes, neonatal hemorrhage, and short gestation/low birth weight were six to 15 times that of singletons. The lowest twin-to-singleton mortality ratios observed were for congenital anomalies and sudden infant death syndrome with relative risks twice that of singletons. The data underscore the need to develop effective strategies to decrease infant mortality among twins.