Disability outcomes and dose escalation with etanercept, adalimumab, and infliximab in rheumatoid arthritis patients: a US-based retrospective comparative effectiveness study.

INTRODUCTION: Rheumatoid arthritis (RA) is a chronic disease that if left untreated may substantially impair physical functioning. Etanercept, infliximab, and adalimumab are tumor necrosis factor (TNF) blockers whose FDA-approved indications in the US include moderate to severe RA. TNF-blocker dose escalation has been well documented in the literature; however, the comparative effectiveness of these agents remains uncertain. OBJECTIVE: To compare the effectiveness and dose escalation rates of etanercept, adalimumab, and infliximab in US community settings. We hypothesized that etanercept would be equivalent to infliximab and adalimumab in patient-reported disability 9-15 months after therapy initiation, and that fewer etanercept patients would experience dose escalation. METHODS: This is a retrospective analysis of the Arthritis, Rheumatism, and Aging Medical Information System (ARAMIS). Adult patients with no biologic use 6 months before TNF-blocker initiation (index) and with Health Assessment Questionnaire Disability Index (HAQ-DI) scores at index and 9-15 months after index were analyzed (218 etanercept, 93 infliximab, and 40 adalimumab). RESULTS: HAQ-DI change scores at 9-15 months did not differ by treatment (-0.12, -0.10, and -0.08 points for etanercept, infliximab, and adalimumab, respectively; p = 0.52). Dose increases were observed in 1.4% of etanercept, 10.8% of infliximab (p < 0.001), and 12.5% of adalimumab patients (p = 0.004). HAQ-DI change was associated with pre-index HAQ-DI score (p < 0.0001) and disease duration (p = 0.001). CONCLUSIONS: Fewer etanercept patients escalated dose than infliximab or adalimumab patients, but improvements in functional disability were similar. These differences may have been influenced by package labeling, mode of administration, or other factors. RA treatment with infliximab and adalimumab in community settings, characterized by dose escalation, did not yield greater disability improvements compared to etanercept, which remained at a relatively stable dose. Uncontrolled treatment selection in this observational design may have influenced outcomes, and prior methotrexate treatment may partly explain disability improvements smaller than typically observed in clinical trials.

The Arthritis, Rheumatism and Aging Medical Information System (ARAMIS): still young at 30 years.

Chronic diseases such as atherosclerosis, arthritis, diabetes, and cancer are among major public health concerns. To understand their cumulative risk factors and antecedents, a chronic disease databank consisting of time-oriented, multidisciplinary longitudinal data, prospectively collected on consecutive patients and describing their clinical courses, provides a systematic anthology of patient reported outcome (PRO) data. ARAMIS, which began in the mid-1970s, was the first large-scale chronic disease data bank system. Outcomes data are collected using the Health Assessment Questionnaire (HAQ), a well established PRO instrument that collects patient-centered data in the areas of disability, pain and other symptoms, adverse effects of treatment, economic impact, and mortality. More than 900 peer-reviewed studies have emanated from ARAMIS since its inception. In the earlier days, and even today, ARAMIS had to invent its own tools for the study of these new sciences. ARAMIS has made dominant contributions to the understanding of PROs and to helping improve treatment and health outcomes in rheumatoid arthritis (RA), osteoarthritis (OA), scleroderma, lupus, aging, and drug side effects. It continues to traverse terrain with participation in the NIH "Roadmap" project, the Patient Reported Outcome Measurement Information System (PROMIS). PROMIS is designed to provide improved assessment of health status across all chronic illnesses as part of an improved infrastructure for clinical research. As initiator of the rich history of chronic disease data banks with "rolling" consecutive open patient cohorts, ARAMIS has enabled the study of real-world PROs in rheumatology, with a wealth of resultant improved approaches to treatment, outcome, cost effectiveness, and quality of life.

Osteoarthritis: new insights. Part 2: treatment approaches.

Osteoarthritis is the most common form of arthritis, affecting millions of people in the United States. It is a complex disease whose etiology bridges biomechanics and biochemistry. Evidence is growing for the role of systemic factors, such as genetics, diet, estrogen use, and bone density, and local biomechanical factors, such as muscle weakness, obesity, and joint laxity. These risk factors are particularly important in the weight-bearing joints, and modifying them may help prevent osteoarthritis-related pain and disability. Major advances in management to reduce pain and disability are yielding a panoply of available treatments ranging from nutriceuticals to chondrocyte transplantation, new oral anti-inflammatory medications, and health education. This article is part 2 of a two-part summary of a National Institutes of Health conference that brought together experts in osteoarthritis from diverse backgrounds and provided a multidisciplinary and comprehensive summary of recent advances in the prevention of osteoarthritis onset, progression, and disability. Part 2 focuses on treatment approaches; evidence for the efficacy of commonly used oral therapies is reviewed and information on alternative therapies, including nutriceuticals and acupuncture, is presented. Biomechanical interventions, such as exercise and bracing, and behavioral interventions directed toward enhancing self-management are reviewed. Current surgical approaches are described and probable future biotechnology-oriented approaches to treatment are suggested.

Reducing need and demand for medical services in high-risk persons. A health education approach.

We undertook this study to identify persons with high medical use to target them for health promotion and self-management interventions specific to their problems. We compared the reductions in cost and health risk of a health education program aimed at high-risk persons with a similar program addressed to all risk levels. We compared health risk and use in 2,586 high-risk persons with those of employee (N = 50,576) and senior (N = 39,076) groups and contrasted results in specific high-risk disease or behavior categories (modules)--arthritis, back pain, high blood pressure, diabetes mellitus, heart disease, smoking, and obesity--against each other, using validated self-report measures, over a 6-month period. Interventions were a standard generic health education program and a similar program directed at high risk individuals (Healthtrac). Health risk scores improved by 11% in the overall high-risk group compared with 9% in the employee group and 6% in the senior group. Physician use decreased by 0.8 visits per 6 months in the high-risk group compared with 0.05 and 0.15 visits, respectively, per 6 months in the employee and senior groups. Hospital stays decreased by 0.2 days per 6 months in the high-risk group compared with 0.05 days in the comparison groups. The duration of illness or confinement to home decreased by 0.9 days per 6 months in the high-risk group and 0.15 and 0.25, respectively, in the employee and senior groups. Using imputed costs of $130 per physician visit, $1,000 per hospital day, and $200 per sick day, previous year costs were $1,138 in direct costs for the high-risk groups compared with $352 and $995 in the employee and senior groups, respectively. At 6 months, direct costs were reduced by $304 in the high-risk group compared with $57 and $70 in the comparison groups. Total costs were reduced $484 in the high-risk groups compared with $87 in the employee group and $120 in the senior group. The return on investment was about 6:1 in the high-risk group compared with 4:1 in the comparison groups. Effective health education programs can result in larger changes in use and costs in high-risk persons than in unscreened persons, justifying more intensive educational interventions in high-risk groups.

Aging, health risks, and cumulative disability.

BACKGROUND: Persons with lower health risks tend to live longer than those with higher health risks, but there has been concern that greater longevity may bring with it greater disability. We performed a longitudinal study to determine whether persons with lower potentially modifiable health risks have more or less cumulative disability. METHODS: We studied 1741 university alumni who were surveyed first in 1962 (average age, 43 years) and then annually starting in 1986. Strata of high, moderate, and low risk were defined on the basis of smoking, body-mass index, and exercise patterns. Cumulative disability was determined with a health-assessment questionnaire and scored on a scale of 0 to 3. Cumulative disability from 1986 to 1994 (average age in 1994, 75 years) or death was the measure of lifetime disability. RESULTS: Persons with high health risks in 1962 or 1986 had twice the cumulative disability of those with low health risks (disability index, 1.02 vs. 0.49; P<0.001). The results were consistent among survivors, subjects who died, men, and women and for both the last year and the last two years of observation. The onset of disability was postponed by more than five years in the low-risk group as compared with the high-risk group. The disability index for the low-risk subjects who died was half that for the high-risk subjects in the last one or two years of observation. CONCLUSIONS: Smoking, body-mass index, and exercise patterns in midlife and late adulthood are predictors of subsequent disability. Not only do persons with better health habits survive longer, but in such persons, disability is postponed and compressed into fewer years at the end of life.

Risk factors for adult Still's disease.

OBJECTIVE: To assess risk factors for adult Still's disease (ASD). METHODS: A matched case-control study of 60 patients with ASD and 60 same sex siblings closest in age was conducted. Subjects were recruited from cohorts in Eastern Canada, Pittsburgh, and the Arthritis, Rheumatism, and Aging, Medical Information Systems (ARAMIS). A questionnaire was used to obtain data on demographic characteristics, education, income, occupation, exposure to toxic substances, stress, and medical history. RESULTS: 116 patients with ASD were identified, of which 104 participated. 86 identified same sex siblings, of which 60 replied. When compared to same sex siblings, ASD patients were similar with respect to education and occupation but had a trend to higher median income. There were no significant associations of ASD with smoking, alcohol consumption, individual toxic substances, vaccination, blood transfusion, minor or major surgery, pregnancy, or diet in the year preceding disease onset. There were no significant associations with tonsillectomy or adenoidectomy, appendectomy, asthma, hay fever, allergy shots, or pregnancy at any time preceding the onset of disease. There was a statistically nonsignificant increase in a history of exposure to coal dust [odds ratio (OR) 3.0; 95% confidence interval (CI) 0.30 to 28.84], in allergy preceding the onset of disease (OR 2.67; 95% CI 0.71 to 10.05), and in oral contraceptive use in the year preceding onset (OR 2.00; 95% CI 0.18 to 22.06). Stressful life events (OR 2.56; 95% CI 1.18 to 5.52) in the year preceding onset was significantly associated with increased risk for ASD. This positive association should be treated with caution unless confirmed by a separate study. CONCLUSION: This exploratory study of risk factors for ASD draws attention to stress as a potentially important risk factor, while likely excluding a considerable number of others.

Prevention of osteoporotic fractures: possibilities, the role of exercise, and limitations.

The Compression of Morbidity paradigm seeks to reduce lifetime illness and morbidity by compressing the dominant morbidity, that of the senior years, between an increasing age of onset of morbidity and a more slowly increasing average age at death. Fractures, often associated with osteoporosis, cause a substantial part of this morbidity. For morbidity resulting from fractures to be reduced, the age-specific incidence of fractures needs to decline, since treatment of fractures after they occur is not likely to prove a major benefit. Thus, the risk factors need to be identified and appropriate preventive interventions undertaken. The medical model seeks to diagnose, then to treat those with disease. In considering prevention, many apply the medical model. The disease is "osteoporosis", we must identify people with this disease and then treat them. The public health model, in contrast, seeks to prevent "disease" in all susceptibles. The disease is "morbidity resulting from fractures". The fatal flaws in the medical screening approach will be discussed, together with a lament that this conference was not entitled: "Recent Progress in the Prevention of Morbidity Associated with Fractures". Osteoporosis is only one of many factors associated with increased morbidity resulting from fractures. A fracture management model for reduction in this morbidity will be presented. Osteoporosis finds its genesis in many well-identified risk factors, including age, sex, estrogen levels, and exercise levels, together with positive (e.g. calcium, estrogen) and negative (corticosteroids) effects of medications. Falls, the other main branch of the model, find their genesis is such risk factor as slippery floors, medication side effects, and co-morbid conditions, often with their own antecedent risk factors. Together, over twenty preventable risk factors contribute to the major morbidity associated with fractures.

Relationship of running to musculoskeletal pain with age. A six-year longitudinal study.

OBJECTIVE: To determine, by longitudinal study, whether long-distance running, maintained for many years, is associated with increased musculoskeletal pain with age. METHODS: A 6-year prospective longitudinal study of 410 runners' club members and 289 community controls, age 53-75 years at study initiation, was conducted. Subjects were also categorized as ever-runners (n = 488) and never-runners (n = 211). The primary dependent variable was pain score as indicated on a horizontal double-anchored analog scale; data for this variable were available beginning in 1987. Statistical adjustment for age, education level, smoking, alcohol consumption, history of arthritis, and presence of other major medical conditions was done by analysis of covariance. Further analyses of previously reported associations of regular vigorous physical activity with decreased disability and mortality after 9 years were performed. RESULTS: The degree of musculoskeletal pain was slightly lower in the exercise group compared with controls, and the difference was statistically significant for women but not for men. Average adjusted pain scores for men were 18.3 (SEM 0.8) in runners' club members, 20.2 (1.2) in controls, 18.6 (0.8) in ever-runners, and 20.3 (1.6) in never-runners. For women, these scores were 17.5 (1.8) in runners' club members versus 22.8 (1.4) in controls (P < 0.05), and 17.2 in ever runners versus 23.7 (1.5) in never-runners (P < 0.002). Disability had continued to develop in runners' club members at a rate only one-third that in the controls after 9 years of observation. Mortality over 9 years consisted of 51 deaths, of which 41 were in the control group and only 10 were among runners' club members. CONCLUSION: Vigorous running activity over many years is not associated with an increase in musculoskeletal pain with age, and there may be a moderate decrease in pain, particularly in women. Vigorous physical activity is associated with greatly decreased levels of disability and with decreased mortality rates.

Lymphoma and luekemia in rheumatoid arthritis: are they associated with azathioprine, cyclophosphamide, or methotrexate?

Incident cases of lymphoma and leukemia in a cohort of 3824 rheumatoid arthritis (RA) patients from the Arthritis, Rheumatism and Aging Medical Information System (ARAMIS) database were identified, and the use of azathioprine, cyclophosphamide, and methotrexate was compared in a matched case-control study. Controls were matched on age, sex, year of study entry, disease duration, center, and years of follow-up. Twenty-four cases of lymphoma and 10 cases of leukemia were identified: 21% of patients with cancer versus 9% of controls had taken azathioprine [McNemar statistic 1.50 (p = 0.22), odds ratio 5.0 (95% confidence interval 0.6,236.5)]. Equal numbers of cases and controls (6% each) had taken cyclophosphamide and 18% of cases and 12% of controls had taken methotrexate [McNemar statistic 0.13 (p = 0.72), odds ratio 1.7 (0.3, 10.7)]. Results suggest but do not prove that RA patients taking azathioprine and methotrexate may have an increased risk of developing lymphoma. However, even if this increased risk can be confirmed, it accounts for only a small proportion of the greatly increased incidence of these malignancies in RA.

Running and the development of disability with age.

OBJECTIVE: To determine, by longitudinal study, whether regular vigorous running activity is associated with accelerated, unchanged, or postponed development of disability with increasing age. STUDY DESIGN: 8-year prospective, longitudinal study with yearly assessments. PARTICIPANTS: 451 members of a runners' club and 330 community controls who were initially 50 to 72 years old (also characterized as "ever-runners" [n = 534] and "never-runners" [n = 247], respectively). MEASUREMENTS: The dependent variable was disability as assessed by the Health Assessment Questionnaire and separately validated in these participant cohorts. Covariates included age, sex, body mass index, comorbid conditions, education level, smoking history, alcohol intake, mean blood pressure, initial disability level, family history of arthritis, and radiologic evidence of osteoarthritis of the knee in a subsample. RESULTS: Initially, the runners were leaner, reported joint symptoms less frequently, took fewer medications, had fewer medical problems, and had fewer instances of and less severe disability, suggesting either that the average previous 12 years of running had improved health or that self-selection bias was present. After 8 years of longitudinal study, the differences in initial disability levels (0.026 compared with 0.079; P < 0.001) had steadily increased to 0.071 for runners compared with 0.242 for controls (P < 0.001). The difference was consistent for men and women. The rate of development of disability was several times lower in the runners' club members than in community controls; this difference persisted after adjusting for age, sex, body mass, baseline disability, smoking history, history of arthritis, or other comorbid conditions (slopes of progression of disability for the years 1984 to 1992, after adjusting for covariates: men in the runners' club, 0.004 [SE, 0.002]; community controls, 0.012 [SE, 0.002]; women in the runners' club, 0.009 [SE, 0.005]; community controls, 0.027 [SE, 0.004]; P < 0.002 for both sets of comparisons). In addition to differences in disability, there were significant differences in mortality between the runners' club members (1.49%) and community controls (7.09%) (P < 0.001). These differences remained significant after adjusting for age, sex, body mass, comorbid conditions, education level, smoking history, alcohol intake, and mean blood pressure (P < 0.002, conditional risk ratio for community controls compared with the runners, 4.27; 95% CI, 1.78 to 10.26). CONCLUSIONS: Older persons who engage in vigorous running and other aerobic activities have lower mortality and slower development of disability than do members of the general population. This association is probably related to increased aerobic activity, strength, fitness, and increased organ reserve rather than to an effect of postponed osteoarthritis development.

Predictors of physical disability after age 50. Six-year longitudinal study in a runners club and a university population.

Predictors of disability were studied over 6 years among 50- to 80-year-old members of a runners club (N = 407) and a university population (N = 299). Data have been collected annually since 1984 on sociodemographic characteristics, health habits, medical history, medication use, family history, psychological parameters, and physical disability as measured by the Health Assessment Questionnaire. Members of the runners club, compared to university participants, had better overall health and less disability at baseline (0.03 versus 0.08) and at 6-year follow-up (0.04 versus 0.24). Predictors of greater subsequent disability among university participants were greater baseline disability, greater medication use, greater number of pack-years of cigarette smoking, older age, being unmarried, higher blood pressure, history of arthritis, and less physical activity compared to one's peers. In addition, changes in characteristics during follow-up that were independently associated with greater disability were development of joint pain, arthritis, or bone fracture and increased body mass index. Predictors of greater disability in the runners group included greater baseline disability, being a nonrunner at baseline, greater dietary salt intake, more years of running at baseline, and greater frequency of physician visits for running injuries. Greater disability in this group also was associated with increases in medication use, declining alcohol consumption, and development of joint pain over 6 years. Results of this study suggest that physical disability is linked to a constellation of characteristics, health habits, medical history, comorbidities, and marital status. While self-selection bias cannot be ruled out entirely, these data are consistent with the hypothesis that those who engage in high levels of physical activity beyond middle age will continue to maintain better functional abilities.

The mortality of rheumatoid arthritis.

OBJECTIVE: To determine the risk and causes of death and to quantify mortality predictors in patients with rheumatoid arthritis (RA). METHODS: RA patients (n = 3,501) from 4 centers (Saskatoon n = 905, Wichita n = 1,405, Stanford n = 886, and Santa Clara n = 305) were followed for up to 35 years; 922 patients died. RESULTS: The overall standardized mortality ratio (SMR) was 2.26 (Saskatoon 2.24, Wichita 1.98, Stanford 3.08, Santa Clara 2.18) and increased with time. Mortality was strikingly increased for specific causes: infection, lymphoproliferative malignancy, gastroenterologic, and RA. In addition, as an effect of the SMR of 2.26, the expected number of deaths was increased nonspecifically across all causes (except cancer), with a large excess of deaths attributable to cardiovascular and cerebrovascular diseases. Independent predictors of mortality included age, education, male sex, function, rheumatoid factor, nodules, erythrocyte sedimentation rate, joint count, and prednisone use. CONCLUSION: Mortality rates are increased at least 2-fold in RA, and are linked to clinical severity.

Correlations between education and arthritis in the 1971-1975 NHANES I.

Data from the National Health and Nutrition Examination Survey I, 1971-1975 (NHANES I) were used to analyze associations among highest education level and arthritis. The dependent variables indicated whether the respondent had ever been diagnosed with any form of arthritis by a physician (10,678 women and 7243 men) or whether physician X-ray readings suggested arthritis of the knee (3491 women and 3119 men). These variables did not distinguish between osteo- and rheumatoid arthritis. It is likely that the great majority of the sample reporting or diagnosed with arthritis had osteoarthritis. There were strong univariate correlations between answers to the general arthritis question and the knee question on the one hand and gender, age, body mass, schooling, income and employment on the other. Respondents' education level was found to be strongly and negatively associated with self-reported arthritis in the larger samples both before and after controls were entered for employment, income and potential biological risk factors. The association between self-reported arthritis or arthritis of the knees and education was weaker for men, but not for women after employment and income were accounted for. When body mass was accounted for, the association between self-reported arthritis or arthritis of the knees and education was weaker among women but not men. Long-run preventive strategies to combat osteoarthritis ought to consider investments in education.

Education, gender, and the compression of morbidity.

According to the Compression of Morbidity (CM) hypothesis, people who exercise, eat nutritiously, do not smoke, and maintain good weight, i.e., people who practice healthy habits, will be more likely to live free of disabling diseases and injuries up until the last few months or years of life. The Increasing Misery (IM) hypothesis, on the other hand, holds that preventive health measures will extend life expectancy but will also increase the number of infirm years. The CM theory implies that curves of morbidity or disability with age should become increasingly "rectangular" for groups who practice healthy habits in the broadest sense. The IM theory does not. This Rectangularization hypothesis is examined with cross-sectional data measuring disability from the Epidemiological Follow-up to the National Health and Nutrition Examination Survey, I (NHEFS), using years of schooling as the independent variable proxy representing favored health status, and examining interactions with age. A modified version of the Disability Index (DI) from the Stanford Health Assessment Questionnaire (HAQ) is used to measure disability. In some analyses, deceased subjects were assigned the worst disability score. Four subsamples of women and men, fifty years old and over, alive and deceased in 1982-84, were analyzed. Female, and especially male, subsamples which included the deceased provided evidence for the CM hypothesis. Results for the subsamples of those remaining alive in 1982-84 were ambiguous. However, lifetime (over age 50) cumulative disability was 21 to 60 percent less for the more educated than the less educated, depending upon whether deceased were included or excluded. If higher education level is an appropriate surrogate for the effect of good health practices, then extending such practices will result in less, rather than more, lifetime disability.

Unilateral ptosis as an initial manifestation of D-penicillamine induced myasthenia gravis.

We describe 2 patients presenting with isolated unilateral ptosis without other signs of cranial or peripheral nerve involvement or sympathetic denervation. Both patients (one case of progressive systemic sclerosis and one of rheumatoid arthritis) were currently taking D-penicillamine. In these cases, the ptosis was reversed a few minutes after a Tensilon test, hallmark of myasthenia gravis. Antibodies to acetylcholine receptors were present. Myasthenia gravis should be suspected with ptosis without other cranial nerve involvement or miosis, even if the ptosis is unilateral. Thus, unilateral ptosis can be the first manifestation of a toxic side reaction to D-penicillamine.

Seasonal variation in the onset of Wegener's granulomatosis, polyarteritis nodosa and giant cell arteritis.

OBJECTIVE: The hypothesis of a seasonal pattern in the onset of symptoms for some vasculitides has been raised in previous small studies. METHODS: Using the data collected by the American College of Rheumatology (ACR) Subcommittee on Classification of Vasculitis, we specifically tested for a higher proportion of onset of symptoms in winter, lower in summer, and intermediate for the other seasons for polyarteritis nodosa (PAN) and Wegener's granulomatosis. We also tested for a higher proportion of onset of symptoms in the spring-summer months for giant cell arteritis (GCA). RESULTS AND CONCLUSIONS: The results of our study support the hypothesis of a seasonal trend for the onset of symptoms of Wegener's granulomatosis (p = 0.04) as described previously. No seasonal pattern was found for the other vasculitides studied (PAN and GCA).

Risk factors for physical disability in an aging cohort: the NHANES I Epidemiologic Followup Study.

Successful improvement in health in our increasingly aged population will depend in substantial part on reduction of age specific disability levels. In turn, the epidemiologic model suggests that this requires identification of risk factors, development of intervention models, and testing of these models. We attempted to identify risk factors for physical disability among 4,428 50-77-year-olds using baseline data collected in the first National Health and Nutrition Examination Survey (NHANES I) (1971-1975) linked to disability data collected 10 years later in the NHANES I Epidemiologic Followup Study. Results of forward stepwise linear regression analysis showed that the major characteristics contributing to greater disability (explaining at least 1% of the variability in scores) were older age at baseline, less nonrecreational activity, arthritis history, less education, female sex, and greater body mass index at age 40. Other factors associated with greater disability included a history of asthma, cardiovascular disease, abnormal urine test, less recreational activity, higher sedimentation rate, rheumatic fever history, lower caloric intake, positive musculoskeletal findings, histories of polio and allergies, lower family income, elevated blood pressure, lower serum albumin, history of tuberculosis, glucose in the urine, and histories of hip or spine fracture, chronic pulmonary disease, and kidney disease.

Strategies for reduction of morbidity.

The future health of our increasingly senior populations depend upon the interrelationship between two critical points: the onset time of the first major disease, infirmity, or disability and the time of death. Reduction of morbidity requires compressing the average period between these points and reducing the average level of morbidity during this period. The goal of compression of morbidity currently is being achieved in some areas. Life expectancy increases in the United States above age 65 y have plateaued, with further increases becoming ever more difficult. Some major chronic diseases, such as atherosclerosis and lung cancer, now occur later in life. Work disability prevalence has begun to decline. Intergenerational comparisons demonstrate improved health at specific ages. Randomized-controlled trials of primary prevention have failed to decrease total mortality in risk subjects while markedly decreasing the morbidity experienced by the same subjects. Compression has been documented for higher socioeconomic class subpopulations. These observations have major implications for health policy and mandate initiatives directed at prevention of disability and infirmity.

Gender and race differences in the correlation between body mass and education in the 1971-1975 NHANES I.

STUDY OBJECTIVE: Differences in the correlation between body mass index and education across four gender and race groups were investigated while simultaneously accounting for occupation, income, marital status, and age. DESIGN: The study used analysis of covariance techniques to calculate average body mass and confidence intervals within education categories while simultaneously adjusting for the covariates: age, square of age, family income, marital status, and occupation. SETTING: Data were drawn from the US National Health and Nutrition Examination Survey (NHANES I), 1971-1975. NHANES I is a national probability sample designed to gather information on the non-institutionalised US civilians, ages 1-74 years. SUBJECTS: Samples of 8211 white women, 1673 black women, 6188 white men, and 1023 black men were drawn from the NHANES I, 1971-1975. MAIN RESULTS: Data in the female samples indicate a strictly inverse relation between body mass and years of schooling among white women and an inverted "U" association among black women, achieving a maximum around 8 to 11 years of schooling. In the male samples data indicate inverted "U" relations among both black and white men, reaching maxima between 12 and 15 years of schooling. The sides of the "U" curve are much steeper for black than for white men. CONCLUSIONS: The four gender/race categories display four different body mass index and education associations. These four associations are only slightly altered by simultaneously adjusting for two additional measures of socioeconomic status: occupation and income.