Population-based Sample Of African Americans Research Articles

Abstract 066: A Healthy Plant-based Diet Was Associated With Slower Cognitive Decline In African Americans: A Biracial Community-based Cohort Of Older Adults

Background: African Americans (AAs) are about twice as likely to develop dementia compared to non-Hispanic whites (NHWs). Dietary modification is an effective preventive strategy to reduce the rate of cognitive decline, a hallmark feature of dementia; however, evidence from diverse populations focusing on AAs is minimal. Objective: To evaluate the association between a plant-based diet and the rate of cognitive decline in a population-based sample of AAs and NHWs. Method: This study consisted of 4753 participants (62% African American, 63% female) in a biracial population-based study in the United States (Chicago Health and Aging Project, 1993 to 2012). Plant-based diet quality was evaluated by the overall plant-based diet index (PDI), the healthful PDI (hPDI), and the unhealthful PDI (uPDI). Global cognition was assessed using a composite score of four tests of cognition. We used mixed models to examine the association between PDI, hPDI, and uPDI with the rate of decline in global cognition, perceptual speed, and episodic memory. Models were adjusted for age, sex, apoE genotype, education, cognitive activities, smoking status, calorie intake, risk factors of cardiovascular disease, time, and an interaction term of time with each covariate. Results: We demonstrated that there were distinct differences in the association of dietary patterns with cognitive decline in AAs and NHWs. AAs had higher intakes of egg, fish or seafood, sugar-sweetened beverages, and whole grains compared to NHWs (p<0.001 for all). Higher hPDI was associated with a slower rate of decline in global cognition, perceptual speed, and episodic memory in AAs but not NHWs. AAs in the highest quintile of hPDI had significantly slower rates of decline in global cognition, β = 0.0183 (p=0.032), perceptual speed β = 0.0179 (p=0.04), and episodic memory β=0.0183 (p=0.04) compared to individuals in the lowest quintile of hPDI. The rate of global cognitive decline in the highest quintile was slowed by 28.4% when compared to individuals in the lowest quintile of hPDI in AAs. There were no associations with either PDI or uPDI and the rate of cognitive decline in either racial group. Conclusion: AAs and NHWs have distinct dietary patterns. A healthy plant-based diet was associated with a slower rate of decline in global cognition, perceptual speed, and episodic memory in AAs. These results are informative in facilitating the development of tailored dietary recommendations for the prevention of cognitive decline in diverse populations.

Differences in Cardiovascular Mortality Risk among African Americans in the Minnesota Heart Survey: 1985-2015 vs The Atherosclerosis Risk in Communities Study Cohort: 1987-2015.

Geographic differences in cardiovascular disease (CVD) mortality among African Americans (AAs) are well-established, but not well-characterized. Using the Minnesota Heart Survey (MHS) and Atherosclerosis Risk in Communities (ARIC) Study, we aimed to assess whether CVD risk factors drive geographic disparities in CVD mortality among AAs. ARIC risk factors were measured between1987-1989 from a population-based sample of AAs, aged 45 to 64 years, living in Jackson, MS and Forsyth County, NC. Similar measures were made at MHS baseline, 1985, in AAs from Minneapolis-St. Paul, MN. CVD mortality was identified using ICD codes for underlying cause of death. We compared MHS and ARIC on CVD death rates using Poisson regression, risk factor prevalences, and hazard ratios using Cox regression. After CVD risk factor adjustment, AA men in MHS had 3.4 (95% CI: 2.1, 4.7) CVD deaths per 1000 person-years vs 9.9 (95% CI: 8.7, 11.1) in ARIC. AA women in MHS had 2.7 (95% CI: 1.8, 3.6) CVD deaths per 1000 person-years vs 6.7 (95% CI: 6.0, 7.4) in ARIC. A 2-fold higher CVD mortality rate remained in ARIC vs MHS after additional adjustment for education and income. ARIC had higher total cholesterol, hypertension, diabetes, and BMI, as well as less education and income than MHS. Risk factor hazard ratios of CVD death did not differ. The CVD death rate was lower in AAs in Minnesota (MHS) than AAs in the Southeast (ARIC). While our findings support maintaining low risk for CVD prevention, differences in CVD mortality reflect unidentified geographic variation.

Abstract P057: Optimal Modifiable Lifestyle Risk Factor Scores Are Associated with Lower Risk of Type 2 Diabetes Mellitus in African Americans - The Jackson Heart Study

Background: The associations of combined modifiable risk factors for incident diabetes (physical activity, television watching, dietary intake, sleep disordered breathing and smoking) are less well investigated in African Americans (AAs). Hypothesis: We hypothesized that an optimal modifiable lifestyle risk factor score would be inversely associated with incident diabetes among AAs. Design and Methods: Data on modifiable risk factors was collected by questionnaire at baseline (2000-2004) in a population-based sample of AAs in the Jackson Heart Study. Incident diabetes (fasting glucose ≥126 mg/dl, physician diagnosis, use of diabetes drugs, or HbA1c ≥ 6.5%) was assessed over 12 years, among adults without prevalent diabetes at baseline. Participants were excluded for missing data on baseline covariates or diabetes follow-up. Incidence rate ratios (IRR) were estimated using Poisson regression modeling adjusting for age, sex, education, current occupation status, systolic blood pressure and body-mass index. Modifiable lifestyle factors (regular exercise, healthy diet, smoking avoidance, lower amounts of television watching and low sleep disordered breathing burden) were combined in risk score categories of poor (0-3 points), average (4-7 points), optimal (8-11 points). Results: Among 3,252 adults (mean age 53.3 years, 64% female) there were 560 incident diabetes cases (median follow-up 7.6 years). An average or optimal compared to a poor modifiable lifestyle risk score was associated with a 21% (IRR 0.79, 95% CI: 0.62, 0.99) and 31% (IRR 0.69, 95% CI: 0.48, 1.01) lower risk of diabetes, respectively, in a monotonic fashion (p=0.03). Body-mass index (BMI) and glycemic status at baseline modified the association of lifestyle risk score with diabetes - among participants with BMI < 30 kg/m 2 , IRRs for average or optimal compared to poor categories were 0.60 (95% CI: 0.40, 0.91) and 0.53 (95% CI: 0.29, 0.97), respectively, compared to 0.90 (95% CI 0.67, 1.21) and 0.83 (95% CI: 0.51, 1.34) among participants with BMI ≥ 30 kg/m 2 . For participants with normoglycemia (normal fasting glucose and HbA1c) at baseline, the IRRs for average or optimal compared to poor categories were 0.64 (95% CI: 0.43, 0.96) and 0.57 (95% CI: 0.31, 1.04), respectively, compared to 0.90 (95% CI 0.69, 1.19) and 0.80 (95% CI: 0.52, 1.23) among participants with prediabetes at baseline. Conclusions: Modifiable lifestyle factors are associated with a lower risk of diabetes among AAs, with greater effects among those with lower adiposity and normoglycemia. Lifestyle interventions to reduce obesity have focused on individuals with high BMI and/or prediabetes (high risk approach). Our study suggests that AAs at the lower end of the diabetes risk spectrum may derive significant long-term benefit from diabetes prevention strategies focused on the outlined modifiable lifestyle risk factors.

Clinic and ambulatory blood pressure in a population-based sample of African Americans: the Jackson Heart Study

Blood pressure (BP) can differ substantially when measured in the clinic versus outside of the clinic setting. Few population-based studies with ambulatory blood pressure monitoring (ABPM) include African Americans. We calculated the prevalence of clinic hypertension and ABPM phenotypes among 1016 participants in the population-based Jackson Heart Study, an exclusively African-American cohort. Mean daytime systolic BP was higher than mean clinic systolic BP among participants not taking antihypertensive medication (127.1[standard deviation 12.8] vs. 124.5[15.7] mm Hg, respectively) and taking antihypertensive medication (131.2[13.6] vs. 130.0[15.6] mm Hg, respectively). Mean daytime diastolic BP was higher than clinic diastolic BP among participants not taking antihypertensive medication (78.2[standard deviation 8.9] vs. 74.6[8.4] mm Hg, respectively) and taking antihypertensive medication (77.6[9.4] vs. 74.3[8.5] mm Hg, respectively). The prevalence of daytime hypertension was higher than clinic hypertension for participants not taking antihypertensive medication (31.8% vs. 14.3%) and taking antihypertensive medication (43.0% vs. 23.1%). A high percentage of participants not taking and taking antihypertensive medication had nocturnal hypertension (49.4% and 61.7%, respectively), white-coat hypertension (30.2% and 29.3%, respectively), masked hypertension (25.4% and 34.6%, respectively), and a nondipping BP pattern (62.4% and 69.6%, respectively). In conclusion, these data suggest hypertension may be misdiagnosed among African Americans without using ABPM.

Fruit and Vegetable Intake and Physical Activity as Predictors of Disability Risk Factors in African-American Middle-Aged Individuals.

To investigate fruit and vegetable intake (FVI) and different dimensions of physical activity (PA) as predictors of change in disabilities and other known precursors of progressive disability in a population-based sample of African Americans. Longitudinal investigation of the independent associations of reported FVI and PA with six-year changes in disabilities and other known precursors of progressive disability. Longitudinal study of a population-representative cohort of late middle-aged African Americans. 432 cohort participants with complete information on all measures. Measurements and Analytic Approach: During wave 8 (2008), FVI was measured using 2005 Behavioral Risk Factor Surveillance System questions and PA dimensions using the Yale Physical Activity Survey (YPAS). Disability measures included basic activities of daily living (ADLs) and instrumental ADLs (IADLs); other precursors included measured gait speed, grip strength, and short physical performance battery (SPPB) and reported lower body functional limitations (LBFLs) and FRAIL scale; these were measured at wave 4 (2004) and wave 10 (2010). Residual-change score linear regression was used to identify FVI and PA factors that were independently associated with six-year changes in disability and other precursors. The study cohort was less active than the YPAS-development group. Longitudinally, leisurely walking was independently associated with better ADL, IADL, grip strength, SPPB, LBFL, and frailty outcomes; standing with better IADL and SPPB; intake of vegetables other than carrots, salads, or potatoes with better grip strength and frailty; and fruit juice intake with worse grip strength and frailty. In this relatively inactive cohort, leisurely walking was associated with multiple beneficial outcomes. Benefits were also seen with vegetables other than potato intake, and fruit juice intake was associated with detrimental effects. This study highlights the importance of finding strategies to help this population increase PA (especially leisurely walking) and intake of whole fruits and vegetables.

Hypertension and alterations in left ventricular structure and geometry in African Americans: the Jackson Heart Study

African Americans (AAs) have an increased risk for hypertension-related cardiovascular outcomes compared with whites, which may be related to abnormal left ventricular (LV) structure. We examined the association of prevalent hypertension with concentric remodeling (CR; normal LV mass index [LVMI] and increased relative wall thickness [RWT]), eccentric hypertrophy (increased LVMI and normal RWT), and concentric hypertrophy (CH; increased LVMI and increased RWT) within the Jackson Heart Study. Among 4721 participants (mean ± SD, age 55.7 ± 12.7 years), 2841 (60.2%) had prevalent hypertension, defined as mean clinic blood pressure ≥140/90 mm Hg or antihypertensive medication use. Prevalent hypertension was associated with a statistically significantly increased odds for having CR (odds ratio [OR] = 1.78, 95% confidence interval [CI] = 1.42–2.24), eccentric hypertrophy (OR = 1.68; 95% CI = 1.15–2.44), and CH (OR = 3.86, 95% CI = 2.28–6.54) after multivariable adjustment. In conclusion, in a population-based sample of AAs, hypertension was associated with increased odds for having abnormal LV structure, particularly CH.

Prevalence, Determinants, and Clinical Significance of Masked Hypertension in a Population-Based Sample of African Americans: The Jackson Heart Study.

The disproportionate rates of cardiovascular disease in African Americans may, in part, be due to suboptimal assessment of blood pressure (BP) with clinic BP measurements alone. To date, however, the prevalence of masked hypertension in African Americans has not been fully delineated. The purpose of this study was to evaluate masked hypertension prevalence in a large population-based sample of African Americans and examine its determinants and association with indices of target organ damage (TOD). Clinic and 24-hour ambulatory BP monitoring were conducted in 972 African Americans enrolled in the Jackson Heart Study. Common carotid artery intima-media thickness, left ventricular mass index, and the urinary albumin:creatinine excretion ratio were evaluated as indices of TOD. Masked hypertension prevalence was 25.9% in the overall sample and 34.4% in participants with normal clinic BP. All indices of TOD were significantly higher in masked hypertensives compared to sustained normotensives and were similar between masked hypertensives and sustained hypertensives. Male gender, smoking, diabetes, and antihypertensive medication use were independent determinants of masked hypertension in multivariate analyses. In this population-based cohort of African Americans, approximately one-third of participants with presumably normal clinic BP had masked hypertension when BP was assessed in their daily environment. Masked hypertension was accompanied by a greater degree of TOD in this cohort.

The Roles of IL-6, IL-10, and IL-1RA in Obesity and Insulin Resistance in African-Americans

The aim of the study was to investigate the associations between IL-1 receptor antagonist (IL-1RA), IL-6, IL-10, measures of obesity, and insulin resistance in African-Americans. Nondiabetic participants (n = 1025) of the Howard University Family Study were investigated for associations between serum IL (IL-1RA, IL-6, IL-10), measures of obesity, and insulin resistance, with adjustment for age and sex. Measures of obesity included body mass index, waist circumference, hip circumference, waist-to-hip ratio, and percent fat mass. Insulin resistance was assessed using the homeostasis model assessment of insulin resistance (HOMA-IR). Data were analyzed with R statistical software using linear regression and likelihood ratio tests. IL-1RA and IL-6 were associated with measures of obesity and insulin resistance, explaining 4-12.7% of the variance observed (P values < 0.001). IL-1RA was bimodally distributed and therefore was analyzed based on grouping those with low vs. high IL-1RA levels. High IL-1RA explained up to 20 and 12% of the variance in measures of obesity and HOMA-IR, respectively. Among the IL, only high IL-1RA improved the fit of models regressing HOMA-IR on measures of obesity. In contrast, all measures of obesity improved the fit of models regressing HOMA-IR on IL. IL-10 was not associated with obesity measures or HOMA-IR. High IL-1RA levels and obesity measures are associated with HOMA-IR in this population-based sample of African-Americans. The results suggest that obesity and increased levels of IL-1RA both contribute to the development of insulin resistance.

Urinary and Fecal Incontinence and Quality of Life in African Americans

To investigate associations between quality of life (QoL) and incontinence in a population-based African-American sample. Cross-sectional survey. Metropolitan St. Louis, Missouri. Eight hundred fifty-three non-institutionalized African Americans aged 52 to 68 in the African American Health study. Respondents who reported having involuntarily lost urine over the previous month were classified as having urinary incontinence (UI), and respondents who reported having lost control of their bowels or stool over the past year were classified as having fecal incontinence (FI). QoL was measured using the Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36) and the 11-item Center for Epidemiologic Studies Depression Scale (CES-D). Prevalences of UI and FI were 12.1% (weighted n=102/841) and 5.0% (weighted n=42/841). Participants with UI and those with FI had worse SF-36 scores than their referent groups (physical function -15.5 and -38.1 points, respectively; role physical -13.2 and -26.5 points; bodily pain -15.7 and -24.5 points; general health perceptions -15.5 and -27.6 points; vitality -15.0 and -16.5 points; social functioning -18.4 and -25.6 points; role emotional -13.2 and -22.1 points; mental health -12.2 and -17.5 points; all Ps<.001), adjusting for age, sex, body mass index, and chronic conditions. Proportions with clinically relevant levels of depressive symptoms were also higher in both groups (UI+17.9%; P<.001) and FI (+37.2%; P<.001) than in their referent groups. UI and FI were strongly associated with worse health-related QoL as well as symptoms of depression in this population-based sample of African Americans.

Informant-based Dementia Screening in a Population-based Sample of African Americans

An informant-based screening tool for dementia may be useful in population-based studies of minority populations. Investigate the feasibility of screening for very mild dementia in a community sample of African Americans using an informant-based screening tool (AD8). Cohort study. One hundred forty-seven persons from the African American Health (AAH) project were screened for dementia; 61 of 93 who were invited had follow-up clinical assessments for dementia diagnosis. The AD8, Mini-Mental State Examination, Short Blessed Test, Brief Instrument for Dementia Detection, and a neuropsychologic battery were administered at visit 1. The Clinical Dementia Rating (CDR) was administered at visit 2 by clinicians blinded to visit 1 results; the presence of dementia was determined by a CDR greater than 0. Four hundred sixty-five individuals from the AAH cohort were sent a letter describing the study and, among this group, 252 individuals were contacted by phone to request participation in this study. Six percent (14/252) of the participants contacted by phone were unable to identify an informant (required for the AD8). One hundred fifty individuals agreed by phone to participate of which 2% (n=3) did not have an informant available at the time of participation. The AD8 alone was effective at discriminating between CDR 0 and CDR 0.5 (area under the curve=0.847; P<0.001; 95% confidence interval, 0.73-0.96). A brief informant-based instrument, the AD8, has high sensitivity and specificity for distinguishing CDR 0 from CDR 0.5 in the community. Informant availability may not be a barrier for using the AD8 in an African American community sample; however, further study in larger samples with a higher response rate, different community settings (eg, community clinics), and among older age groups (eg, age 75+) is warranted to confirm this.

The androgen receptor CAG and GGN repeat polymorphisms and prostate cancer susceptibility in African-American men: results from the Flint Men's Health Study.

Repeat lengths of the CAG and GGN microsatellites in exon 1 of the androgen receptor (AR) gene have been hypothesized to be associated with prostate cancer risk. In vitro studies have showed an inverse association between AR CAG and GGN repeat length and activity levels of the AR product. It is known that men of African descent have a higher incidence of and greater mortality from prostate cancer than men of Caucasian or Asian descent and, on average, a smaller number of repeats at AR CAG and GGN. Consistent with these findings, studies have also found increased AR protein expression levels in benign prostatic hyperplasia and prostatic diseased tissues from men of African descent. Despite these findings, limited studies have been conducted to evaluate the association between repeat lengths at AR CAG and prostate cancer risk in African Americans. Our study is the first such study to examine whether repeat length of the AR GGN repeat is associated with prostate cancer risk in African Americans. We found no evidence for an association between AR CAG or GGN repeat lengths and prostate cancer risk in a population-based sample of African Americans.

Demographic, Behavioral, Psychosocial, and Dietary Correlates of Cancer Screening in African Americans

We examined demographic, behavioral, psychosocial, and dietary correlates of prostate, breast, and colorectal cancer screening in a population-based sample of African Americans in North Carolina (n=405). Cross-sectional data of African Americans in North Carolina were analyzed using linear regression models. Seventy-eight percent of men had obtained prostate screening (PSA test) and 81% of women had received breast cancer screening (mammogram) during the previous two years. Screening rates for CRC were lower (48% women, 31% men in the previous ten years). Older age, college education, and being married were associated with all screening modalities, as was belief in a diet-cancer relationship with PSA testing and mammography. There were correlations of former smoking, obesity, and healthy eating self-efficacy with PSA testing, and family/personal cancer history with CRC screening. Screening for all cancers was associated with lower fat consumption, and PSA testing with high fruit/vegetable intake (p<.05). Cancer screening was prevalent, but rates did not approach national targets. Knowledge of these correlates can be used to design effective cancer screening interventions for African Americans.

Genetic testing for colon cancer among African-Americans in North Carolina

Objective. To describe attitudes and correlates of intention to take a genetic test for colon cancer in a population-based sample of African-Americans. Methods. African-Americans ( n = 658), age 18–70, in North Carolina completed an 11-page questionnaire between June–October 2003 that assessed attitudes (familiarity, perceived benefits and risks, anxiety, and confidentiality) and intention to take a genetic test for colon cancer and various participant characteristics. Results. Respondents expressed favorable attitudes and high intention regarding genetic testing for colon cancer: 87% would definitely/probably take a genetic test, although only 42% had read/heard a lot or some about genetic testing. Most agreed that genetic test results should be available to healthcare providers (79%) but not to health insurers (62%) or employers (82%). About a third were concerned that genetic testing could lead to discrimination. Correlates of intention differed by sex. Perceived benefits were significantly positively associated with intention among all respondents. However, being married (OR = 2.1, 95% CI: 1.2, 3.7), doctor as the main source of health information (OR = 2.4, 95% CI: 1.4, 3.9), and colon cancer family history (OR = 4.3, 95% CI: 1.6, 11.6) were significant only for women; some college education (OR = 4.1, 95% CI: 1.7, 9.7), importance of sharing test results with relatives (OR = 5.5, 95% CI: 1.6, 18.7), and colon cancer screening history (OR = 3.4, 95% CI: 1.6, 7.5) were only significant for men. Conclusions. Respondents expressed high interest in genetic testing for colon cancer risk, although confidentiality of test results is a concern. Guidelines and policies for genetic testing specific to African-Americans should be established and future research should examine the prevalence of genetic testing.

Cognitive Ability and Physical Performance in Middle‐Aged African Americans

To investigate the association between cognitive ability and physical performance in a population-based sample of middle-aged African Americans. Cross-sectional study, 2000/2001. St. Louis, Missouri. Nine hundred ninety-eight African Americans born between 1936 and 1950. Cognitive function was measured using the Mini-Mental State Examination (MMSE) and the Animal Naming Test of verbal fluency. Physical performance was measured using eight tests: chair stand, semitandem stand, tandem stand eyes open, tandem stand eyes closed, one-leg stand, usual gait speed, grip strength, and peak expiratory flow. There was a statistically significant and monotonic (progressively worsening) trend of the eight physical performance measures across cognitive tertiles in all eight MMSE analyses and five of eight Animal Naming analyses, controlling for age, sex, education, geographic area, depressive symptoms, and comorbid conditions. The association between physical performance and cognitive function appears robust. The results extend previous reports for adults aged 65 and older to a measure of verbal fluency and to a population-based sample of African Americans aged 49 to 65. Further research is needed to disentangle the temporal sequence and identify potential interventions to prevent declines in function.

Left ventricular structure and systolic function in African Americans: the Atherosclerosis Risk in Communities (ARIC) study.

To estimate prevalence of left ventricular (LV) hypertrophy and its relation to systolic function in a population-based sample of African Americans. A baseline 2D guided M-mode echocardiogram was conducted as part of a longitudinal cohort study to assess prevalence and cross-sectional relationships between echocardiographic and clinical parameters. Data were collected as part of the Atherosclerosis Risk in Communities study. Analysis is limited to 1543 African Americans, aged 51-70 years, without clinically apparent cardiovascular or echocardiographically determined valvular disease. LV hypertrophy prevalence was defined as LV mass/ height2.7 > or = 51 g/m2.7. LV systolic chamber function was assessed at the midwall using the ratio of observed midwall fractional shortening (MWS%) to the value predicted from circumferential end-systolic stress. The prevalence of LV hypertrophy was 33% in men, 38% in women. The prevalence of concentric hypertrophy (LV hypertrophy with relative wall thickness > or = 0.45) was greater than that of eccentric hypertrophy (men: 24% vs 9%; women: 27% vs 11% women). Observed/predicted (O/P) MWS% was strongly and inversely related to LV mass/ height2.7 (P<.001) and LV hypertrophy (P<.001). The O/P MWS% was inversely related to LV mass/height2.7 quartile: O/P MWS% was 106% and 99% in the first and 97% and 89% in the fourth quartile of LV mass/height2.7 for men and women, respectively. Adjusting for age, adiposity, diabetes, blood pressure, antihypertensive medication use, and smoking did not remove association between O/P MWS% and LV mass/height2.7. LV hypertrophy was highly prevalent in this population-based middle-aged sample of African Americans and was associated with poorer LV systolic chamber function.

The Stages of Change of Rural African-American Smokers

Cigarette smoking remains the most important preventable cause of death in the United States, and the burden is especially great for African Americans. A promising approach to better understanding smoking behavior involves the application of a stages-of-change model. The purpose of this study was to analyze the smoking behavior of a population-based sample of African Americans in two rural southern counties using this model. Data were collected using a door-to-door household survey. The study sites were two rural counties in central Virginia. A total of 556 African-American smokers and recent quitters completed a questionnaire concerning their smoking behavior and beliefs about quitting. Overall, 51% of the respondents were in the pre-contemplation stage, 28% in the contemplation stage, 17% in the preparation stage, and 4% in the action stage. In a multivariate model, eight factors were significantly associated with the stage of change among current smokers: age, number of prior cessation attempts; confidence in becoming a nonsmoker; scales measuring both the pros and cons of smoking; perceived desires of others; believing quitting would improve health; and number of doctor visits in the past year. This study provides support for applying a stages-of-change model to African Americans who smoke. Many of the predictors of the stage of change are the same as those found in other populations. In addition, finding an association between stages and both physician visits and believing stopping smoking can improve health has important implications for providing smoking cessation services in health care settings.