Walden University. College of Health Sciences. This is to certify that the doctoral dissertation by. Erik Lamont Cook has been found to be complete and ...
Walden University College of Health Sciences
This is to certify that the doctoral dissertation by Erik Lamont Cook has been found to be complete and satisfactory in all respects, and that any and all revisions required by the review committee have been made. Review Committee Dr. JaMuir Robinson, Committee Chairperson, Public Health Faculty Dr. Xianbin Li, Committee Member, Public Health Faculty Dr. Scott McDoniel, University Reviewer, Public Health Faculty
Chief Academic Officer Eric Riedel, Ph.D.
Walden University 2014
Abstract Sleep Duration as an Independent Predictor of Type 2 Diabetes Mellitus among African American Adults by Erik Lamont Cook
MS, California National University, 2010 BS, California National University, 2006
Dissertation Submitted in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy Public Health
Walden University January 2014
Abstract Type 2 diabetes mellitus (T2DM) is one of the most destructive and debilitating chronic diseases of modern time. African American adults are among the individuals most negatively affected by T2DM. Previous research suggests that sleep duration is a significant predictor of T2DM and could provide a possible explanation for the higher burden of T2DM seen among African American adults. Despite this demonstrated relationship among various population subgroups, no study to date has demonstrated sleep duration role as an independent predictor of T2DM among African American adults. The social ecological model served as the theoretical framework for examining how this specific individual-level factor, sleep duration, could combine with other factors to influence T2DM. The purpose of this cross-sectional study was to examine the relationship between sleep duration and T2DM, adjusting for established T2DM risk factors (body mass index, physical activity, smoking, and socioeconomic status), among 4,529 African American adults from the 2011 National Health Interview Survey. Chisquare and binary logistic regression analyses were used. Sleep duration was not a significant predictor of diabetes prevalence after adjusting for T2DM risk factors. However, these risk factors remained significantly associated (p < .05) with diabetes prevalence in the study population. This study contributes to positive social change by providing an understanding of the ways in which sleep duration and T2DM are jointly influenced by established T2DM risk factors. The study findings may also be used for empowering African American communities to focus on diabetes prevention programs that are centered toward maintaining a normal body mass index, smoking cessation, and increased physical activity.
Sleep Duration as an Independent Predictor of Type 2 Diabetes Mellitus among African American Adults by Erik Lamont Cook
MS, California National University, 2010 BS, California National University, 2006
Dissertation Submitted in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy Public Health
Walden University January 2014
UMI Number: 3610873
All rights reserved INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted. In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion.
UMI 3610873 Published by ProQuest LLC (2014). Copyright in the Dissertation held by the Author. Microform Edition © ProQuest LLC. All rights reserved. This work is protected against unauthorized copying under Title 17, United States Code
ProQuest LLC. 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, MI 48106 - 1346
Dedication I dedicate this study to my family and all of those who suffer from, and are at risk of, type 2 diabetes mellitus.
Acknowledgments First, I would like to thank Dr. JaMuir Robinson, my Dissertation Committee Chair, Dr. Xianbin Li, Committee Member, and Dr. Scott McDoniel, University Research Reviewer, for their outstanding dedication, guidance, and support throughout this entire dissertation process. I truly would have not completed this dissertation without your unwavering support. I would like to thank the entire Walden University faculty for their superior professional guidance and support not only for this dissertation project but during my entire student tenure. I would like to thank my entire family for their support along the way. And most of all, I would like to thank God for providing me the strength, guidance, and resources to attain this academic objective.
Table of Contents List of Tables .......................................................................................................................v Chapter 1: Introduction to the Study....................................................................................1 Background of the Study ...............................................................................................3 Problem Statement .........................................................................................................5 Purpose of the Study ......................................................................................................6 Research Questions and Hypotheses .............................................................................6 Theoretical Framework ..................................................................................................7 Operational Definitions ..................................................................................................9 Scope and Delimitations ..............................................................................................12 Limitations ...................................................................................................................12 Significance of the Study .............................................................................................13 Summary and Transition ..............................................................................................14 Chapter 2: Literature Review .............................................................................................15 Introduction ..................................................................................................................15 T2DM and its Burden in African American Adults.....................................................16 An Overview of Sleep ..................................................................................................18 Sleep Duration and Health Outcomes ..........................................................................19 Sleep Duration in the U.S. Population .........................................................................20 Sleep Duration among African American Adults ................................................. 21 Studies Examining the Sleep Duration and T2DM Relationship ................................22 The Sleep Duration-T2DM Relationship by Gender ............................................ 24 i
The Sleep Duration-T2DM Relationship by Race/Ethnicity ................................ 27 Risk Factors Associated with Sleep Duration and T2DM ...........................................33 Obesity and T2DM ............................................................................................... 34 Obesity and Sleep Duration .................................................................................. 35 Physical Activity and T2DM ................................................................................ 36 Physical Activity and Sleep Duration ................................................................... 36 Smoking and T2DM ............................................................................................. 37 Smoking and Sleep Duration ................................................................................ 38 Individual SES and T2DM.................................................................................... 39 Individual SES and Sleep Duration ...................................................................... 41 Variable Measures Used in Sleep Duration-T2DM Research .....................................42 Chapter 3: Research Method ..............................................................................................45 Introduction ..................................................................................................................45 Research Design and Rationale ...................................................................................45 Setting and Sample ......................................................................................................46 The National Health Interview Survey ................................................................. 46 Selection of Study Participants ............................................................................. 47 Power Analysis ..................................................................................................... 48 Data Collection ............................................................................................................48 Instrumentation and Materials .....................................................................................49 Dependent Variable .............................................................................................. 51 Independent Variable ............................................................................................ 51 ii
Risk Factors .......................................................................................................... 51 Data Analysis ...............................................................................................................53 Threats to Validity .......................................................................................................55 Protection of Human Participants ................................................................................57 Summary ......................................................................................................................57 Chapter 4: Results ..............................................................................................................59 Introduction ..................................................................................................................59 Data Collection ............................................................................................................60 Results ..........................................................................................................................60 Demographic Characteristics ................................................................................ 60 Sleep Duration ...................................................................................................... 64 Diabetes Prevalence .............................................................................................. 65 Research Question 1 ............................................................................................. 68 Research Question 2 ............................................................................................. 69 Post-Hoc Analysis ................................................................................................. 75 Summary and Transition ..............................................................................................77 Chapter 5: Discussion, Conclusions, and Recommendations ............................................79 Introduction ..................................................................................................................79 Summary of Findings ...................................................................................................79 Interpretation of Findings ............................................................................................80 Socioeconomic Status ........................................................................................... 80 Overweight/Obesity .............................................................................................. 81 iii
Physical Activity ................................................................................................... 82 Smoking Status ..................................................................................................... 82 Diabetes Prevalence .............................................................................................. 83 Sleep Duration ...................................................................................................... 84 Research Question 1 ............................................................................................. 84 Research Question 2 ............................................................................................. 86 Post-Hoc Analysis ................................................................................................. 87 Limitations ...................................................................................................................90 Recommendations for Future Studies ..........................................................................91 Implications for Social Change ....................................................................................92 Conclusion ...................................................................................................................93 References ..........................................................................................................................96 Curriculum Vitae .............................................................................................................121
iv
List of Tables Table 1. Summary of 2011 NHIS Survey Questions for Variables of Interest................. 50 Table 2. Statistical Test Summary .................................................................................... 55 Table 3. Summary of Study Participant Demographic Characteristics ............................ 61 Table 4. Summary of Study Participant Average Age and Body Mass Index .................. 62 Table 5. Chi-Square Analysis of Sleep Duration and Demographic Characteristics ....... 65 Table 6. Demographic Characteristics and Diabetes Prevalence...................................... 66 Table 7. Chi-Square Analysis for Established T2DM Risk Factors and Diabetes ........... 67 Table 8. Unadjusted Odds Ratio for Sleep Duration and Diabetes ................................... 68 Table 9. Sleep Duration-Diabetes Association Controlling for each Risk Factor ............ 70 Table 10. Post-Hoc Logistic Regression Analysis: Predictors of Diabetes Prevalence ... 76
v
1 Chapter 1: Introduction to the Study Chronic disease presents the most significant challenge to public health today. Chronic diseases are those illnesses distinguished by prolonged illness, disability or impairment, inability to cure, and multiple risk factors (Remington, Brownson, & Wegner, 2010). In the last century, diseases such as heart disease, cancer, and diabetes have replaced infectious disease as the leading cause of morbidity and mortality in the United Sates (Remington et al., 2010). An estimated 1 in 2 Americans have at least one chronic disease (Centers for Disease Control and Prevention [CDC], 2009). Additionally, an estimated 75% of health care costs in the United States are associated with chronic diseases (CDC, 2009). Finally, chronic diseases account for approximately seven in every 10 deaths in the United States (Kung et al., 2008). Of all racial/ethnic groups in the United States, African American adults bear the most significant burden of chronic disease. According to the Office of Minority Health (OMH; 2012), many chronic diseases such as heart disease, hypertension, cancer, and diabetes disproportionately affect African American adults. Of these, type 2 diabetes mellitus (T2DM) presents one of the greatest burdens among African American adults. T2DM is one of the most destructive and debilitating chronic diseases of modern time. The medical costs, incidence, and mortality rates of T2DM have grown exponentially over the last half century, and continue to grow at an alarming rate. Cases of diagnosed diabetes among adults in the United States have increased from 1.6 million diagnosed cases in 1958 to 21 million by 2010 (CDC, 2011b). It is projected by the year 2050 that 39 million Americans will have diabetes (Honeycutt et al., 2003). This disease
2 put a significant burden on the U.S. health care system as direct medical costs reached $116 billion in 2007 (CDC, 2011c). Diabetes has significantly contributed to the increased number of health complications that include heart disease, hypertension, blindness, kidney disease, amputations, nervous system diseases, and pregnancy complications (CDC, 2011c). Furthermore, diabetes ranks seventh in the top 15 causes of death in the United States as it was responsible for over 68,000 deaths in the year 2010 alone (National Vital Statistic Reports [NVSS], 2012). African American adults are among those most negatively affected by T2DM. These negative effects are evident in T2DM prevalence, incidence, morbidity, and related mortality among African American adults. Over the last 3 decades, the prevalence of diagnosed T2DM among African American adults increased from 4.5% in 1980 to nearly 9.5% by the end of 2010, compared to the T2DM prevalence of 2.6% in 1980 to 6.0% in 2010 among their European American counterparts (CDC, 2011b). Diabetes incidence among African American adults in 2010 was 13.0 per 1,000, compared to 7.7 per 1,000 among European Americans (CDC, 2011b). With regard to morbidity, African American adults harbor disproportionate rates of a host of diabetes related complications that include End-Stage Renal Disease (Collins et al., 2010), Hyperglycemic Crisis (Smiley, Chandra, & Umpierrez, 2011), amputations, and vision impairment (Lanting et al., 2005). Finally, African Americans had an age-adjusted T2DM mortality rate of 38.7 deaths per 100,000, compared to 19.0 deaths per 100,000 among European Americans in 2010 (NVSS, 2012). This significant T2DM burden among African American adults suggests a better understanding of its etiology is needed.
3 This chapter begins by outlining the background of the study, problem statement, purpose of the study, and nature of the study. This chapter will then focus on a detailed description of research questions and hypotheses, theoretical framework, operational definitions, assumptions, limitation, and delimitations. Finally, this chapter concludes with a discussion of the significance of this study. Background of the Study In recent years, a significant interest has been taken in sleep duration and its effects on human health (Alvarez & Ayas, 2004; Hobson, 2005; Mignot, 2008; Stickgold, 2005; Yoo et al., 2007; Zepelin, 2005). Although the exact purpose of sleep is not well understood, it has been demonstrated to be essential to the normal operation of human metabolic (Redwine et al., 2000; Siegel, 2005), endocrine (Spiegel, Leproult, & Van Cauter, 1999), cardiovascular, and sympathetic nervous system functions (Kato et al, 2000). Furthermore, adverse sleep duration has been demonstrated to be associated with increased mortality (Kronholm et al., 2011; Patel et al., 2004), cardiovascular disease (Buxton & Marcelli, 2010), and obesity (Gangwisch et al., 2005; Lyytikainen et al., 2010; Patel et al., 2006; Watanabe et al., 2010). Because of the established adverse effects of short and long sleep duration, the National Sleep Foundation (2013) recommended 7-9 hours of sleep per day to aid in maintaining good health. Despite the established importance of attaining recommended amounts of sleep duration, adverse sleep duration among the U.S. population continues to increase. The prevalence of American adults sleeping fewer than 7 hours on average per day increased from 30% in 1998 to over 37% in 2005 (National Sleep Foundation, 2005).
4 Fewer information regarding long sleep duration trends exists. Nevertheless, the prevalence of long sleep duration has been reported to be between 5% and 9% among American adults (Buxton & Marcelli, 2010; Patel et al., 2006). Furthermore, several studies suggest that long sleep duration is associated with increased morbidity (Gottlieb et al., 2006; Patel et al., 2006) and mortality (Ferrie et al., 2007; Hublin et al., 2007). African American adults suffer the largest burden of short and long sleep duration among all American adults. African American adults have been reported to have average sleep duration 6.94 hours per night, compared to 7.18 hours for European American and 7.10 hours for Hispanics (Hicks et al., 1999). Additionally, it has been reported that African American adults have a long sleep prevalence of 10.1%, compared to long sleep prevalence of 9.5% among Hispanics, and 8.3% among European American adults (Krueger & Friedman, 2009). T2DM remains a significant problem among African American adults. Despite the established role of overweight/obesity (Hu et al., 2004; Krishnan et al., 2012; Wang et al., 2005; Weinstein et al., 2004), physical inactivity (Hu et al., 2001; Hu et al., 1999; James et al., 1998), smoking (Carlsson, Midthjell, & Grill, 2004; Will et al., 2001; Yeh et al., 2010), and low socioeconomic status (SES; Cunningham et al., 2008; Maskarinec et al., 2009; Maty et al., 2010; Ross et al., 2010; Sims et al., 2011) as significant predictors of T2DM, they have failed to explain the racial differences in T2DM observed among African American adults (Brancati et al., 2000; Cabassa et al., 2011; Shai et al., 2006). Previous research has suggested that sleep duration is associated with increased T2DM (Ayas et al., 2003; Beihl, Liese, & Haffner, 2009; Chao et al., 2011; Gottlieb et al., 2005;
5 Hayahsino et al., 2007; Kita et al., 2012; Lou et al., 2012; Mallon et al., 2005; Najafian et al., 2013; Tuomilehto et al., 2008; Yaggi et al., 2006). Because sleep duration has been implicated as a predictor of T2DM among various population groups, it was necessary to conduct a study to determine if this association could provide a possible explanation for the higher proportions of T2DM seen among African American adults. Problem Statement Adverse sleep duration is associated with increased risk of T2DM. A significant body of research indicates that sleep duration is an independent predictor of T2DM among women (Ayas et al., 2003; Gottlieb et al., 2005; Mallon et al., 2005; Tuomilehto et al., 2008), men (Ayas et al., 2003; Gottlieb et al., 2005; Tuomilehto et al., 2008; Yaggi et al., 2006), European American, Hispanics (Beihl et al., 2009), and other ethnic groups (Chao et al., 2011; Hayahsino et al., 2007; Kita et al., 2012; Lou et al., 2012; Najafian et al., 2013). Despite this demonstrated relationship in the current body literature, it remains unclear if sleep duration acts as an independent predictor of T2DM among African American adults. Specifically, few studies have examined the relationship between sleep duration and T2DM among African American adults (Beihl et al., 2009; Zizi et al., 2012). Of these few studies, none have found sleep duration to be a significant predictor of T2DM among African American adults. The small sample of African American adults used resulted in inadequate statistical power and thus, likely contributed to the insignificant results. Furthermore, Zizi et al. (2012) inclusion of 6 hours as normal sleep, which is inconsistent with most studies examining the sleep duration-T2DM relationship (Ayas et al., 2003; Beihl et al., 2009; Gangwisch et al., 2007; Gottlieb et al.,
6 2005; Kato et al., 2012; Najafian et al., 2013; Tuomilehto et al., 2008; Yaggi et al., 2006), most likely contributed to statistically insignificant results as a result of reduced variability in sleep categories among African American adult participants. This study examined short and long sleep duration’s role as a predictor of T2DM among a group of African American adults. Purpose of the Study The purpose of this study was to examine the relationship between sleep duration and T2DM prevalence, adjusting for established T2DM risk factors, among a sample of African American adults from the 2011 National Health Interview Survey (NHIS). The NHIS provides a large representative sample of African American adults. The independent variable for this study was sleep duration. The controlling variables for this study were BMI, physical activity, smoking status, education level, income level, and occupational status. Finally, the dependent variable for this study was diabetes prevalence. Research Questions and Hypotheses The following research questions and hypothesis were used to direct the course of the proposed study: 1. Is sleep duration associated with prevalent T2DM among African American adults? 1: Sleep duration is not associated with prevalent T2DM among African American adults.
7 1: Sleep duration is associated with prevalent T2DM among African American adults. 2. Do established T2DM risk factors mediate the association between sleep duration and prevalent T2DM among African American adults? 2: Established T2DM risk factors do not mediate the association between sleep duration and prevalent T2DM among African American adults. 2: Established T2DM risk factors mediate the association between sleep duration and prevalent T2DM among African American adults. Theoretical Framework The theoretical framework for this study was the social ecological model (SEM). The SEM, which was developed in the mid 1960s, views health outcomes as if they are interwoven in the fabric of society (Stokols, 1996; Whittemore, Melkus, & Grey, 2004). The SEM posits there is an interaction between the environment and health outcomes. Environment is said to consist of various cultural, social and physical dimensions (Stokols, 1996; Whittemore, Melkus, & Grey, 2004). Specifically, the SEM examines health behaviors and outcomes through one or more combinations of the following five levels of influences: individual; interpersonal; institutional; community; and public policy factors (Stokols, 1996; Whittemore et al., 2004). T2DM is influenced by several individual, social, and environment factors (Carlsson, Midthjell, & Grill, 2004; Cunningham et al., 2008; Demakakos et al., 2010; Howard et al., 2012; Hu et al., 2001; Hu et al., 2004; Krishnan et al., 2010; Krishnan et al., 2012; Maskarinec et al., 2009; Maty et al., 2010; Robbins et al., 2005; Ross et al.,
8 2010; Sims et al., 2011; Wang et al., 2005; Weinstein et al., 2004; Will et al., 2001; Yeh et al., 2010). These factors are also multifaceted and complex in nature (Whittemore et al., 2004). Therefore, the SEM provided a useful framework for exploring the ways that sleep duration could influence T2DM among African American adults. Sleep duration may be an additional individual factor that impacts T2DM risk among African American adults. Nature of the Study I conducted a cross-sectional quantitative secondary analysis of data from the 2011 NHIS. The 2011 NHIS is a cross-sectional health survey conducted by the National Center for Health Statistics (NCHS), CDC, for the purpose of collecting data regarding various health behaviors and illnesses among non-institutionalized members of the U.S. population (CDC, 2012). The 2011 NHIS utilized a multistage probability design. This design produced a population sample of 101,875 individuals from 40,496 families, yielded by 39,509 households (CDC, 2012). Participants for this study were selected from the 2011 NHIS Sample Adult file. The response rate for the 2011 NHIS Sample Adult component was 81.6%, with a final population consisting of 33,014 persons, ages 18 years and older (CDC, 2012). Of which, 5,086 were classified as African American adults. I examined the sleep duration-T2DM relationship using a sample of African American adults aged 18-85 years from the 2011 NHIS. Data for the 2011 NHIS were collected via face-to-face interviews using questionnaires (CDC, 2012). Specifically, questionnaires were used to collect
9 sociodemographic, health risk behavior, and health condition data from participants of the 2011 NHIS (CDC, 2012). Sociodemographic data collected from participants included age, race/ethnicity, gender, income, education level, and occupational status. Information collected on health behaviors included BMI, depression, physical activity, smoking status, and sleep duration (CDC, 2012). Finally, data collected on health outcome from study participants included various chronic diseases such as heart disease, hypertension, cancer, and diabetes (CDC, 2012). Chi-square analysis and binary logistic regression were used to analyze collected data. Operational Definitions The variables of interest for this study included sleep duration, T2DM, BMI, physical activity, smoking, education, income, and occupational status. Operational definitions for terms and variables used in this study are provided in the following: Sleep Duration: Sleep duration is a period of sustained inactivity and reduced sensitivity to external stimuli in human beings (Zepelin et al. 2005). For this study, sleep duration was the independent variable and was categorized into the following three groups: normal sleep is7-8 hours; short sleep duration is ≤6 hours; long sleep duration is ≥ 9 hours. Type 2 Diabetes Mellitus: Diabetes mellitus is defined as a cluster of metabolic disorders distinguished by chronic hyperglycemia (World Health Organization [WHO], 1999). Clinical diagnosis of diabetes is confirmed if an individual has a fasting blood glucose ≥ 126mg/dl, or a two-hour blood glucose ≥ 200mg/dl (WHO, 2006). The most common type of diabetes is T2DM, which is distinguished by problems with insulin
10 secretion and insulin reactivity (WHO, 1999). Type 2 diabetes mellitus was the dependent variable for this study. I categorized study participants based on self-reported diabetes status as follows: diabetes; no diabetes. Established T2DM Risk Factors: This term is used to identify all lifestyle risk factors including BMI, physical activity, smoking status, and individual SES. Body Mass Index: Body Mass Index is an index that determines a person’s fat mass based using the individual’s height and weight (CDC, 2011a). The unit of measure for body mass index (BMI) is kilograms per square meter (kg/ categorized as follows: 18.5 kg/ 25.0 kg/
to 29.9 kg/
to 24.9 kg/
). Body mass index is
is considered a normal weights status;
is considered to be overweight; 30 kg/
or higher is
considered to be obese (CDC, 2011a). For this study, I categorized BMI as follows: normal≤24.9 kg/
; overweight= 25.0 kg/
to 29.9 kg/
; obese ≥30 kg/
.
Physical Activity: Physical activity is the process of energy usage (in the form of calories) resulting from movement of the body (Caspersen, Powell, & Christeson, 1985). For this study, I categorized physical activity based on the frequency reported by the respondent. Smoking: Smoking is a lifestyle risk factor that continues to be one of the leading causes of morbidity and mortality in the United States. Smoking has been implicated in several diseases including heart disease (Lee et al., 2006), cancer (Agudo et al., 2012), and T2DM (Carlsson, Midthjell, & Grill, 2004; Will et al., 2001; Willi et al., 2007; Yeh et al., 2010). For this study, I categorized smoking status based on the reported frequency of smoking.
11 Individual Socioeconomic Status: Is a dynamic set of economic and a social variable that contribute to an individual’s or group’s position in society (Berkman & Kawachi, 2000; Halverson et al., 2004; Shavers, 2007). Education, income, and occupational status are the most common measures of individual SES (Berkman & Kawachi, 2000). Education: Education is a component of individual SES that provides future direction for occupational and income potential (Adler & Newman, 2002). Typical measures of education include: number completed years of education; highest level of education attained; and educational credential attained (Shavers, 2007). For this study, I measured education from the lowest reported level (Non-High School Graduate) to the highest reported level (Doctorate Degree). Income: Income is a component of individual SES that measure the degree to which one can obtain quality housing, healthy foods, recreation and quality education (Adler & Newman, 2002; Berkman & Kawachi, 2000). Income is typically determined by measuring annual individual income, annual household income, or family income (Shavers, 2007). For this study, I categorized total household income based on the amount of reported annual income in thousands of dollars. Occupational Status: Occupational status is an estimate of one’s working conditions (Shavers, 2007). For this study, I categorized occupational status based on reported employment status.
12 Assumptions Several assumptions were made for this study. One assumption was that the sample African American study participants of the 2011 NHIS were a nationally representative of the African American population in the United States. A second assumption was that all questions regarding sleep duration, physical activity, smoking status, individual SES, and T2DM status were answered accurately and honestly by all participants of the 2011 NHIS. A third and final assumption was that bias resulting from the use of self-reported information in the 2011 NHIS was low. Scope and Delimitations This study was limited to African American adults from the 2011 NHIS. African American adults, ages 18-85 years, from the 2011 NHIS with complete sleep duration, established T2DM risk, individual SES, and T2DM status data were included in this study. This study was also limited to exploring sleep duration only. Other aspects of sleep such as sleep quality were not included in this inquiry. Finally, the results of this study were generalizable to African American adults of this study only. Limitations Several limitations were identified for this study. The limitations of this study concerned themselves with data source and study design. One limitation was that data used for this study came from the 2011 NHIS, which is a secondary data source (CDC, 2012). The use of a secondary data source could result in the lack of accurate and complete data (Sorensen, Sabroe, & Olsen, 1996). A second limitation was that bias and misclassification of study participant status could have resulted from intentional or
13 unintentional misreport of information. Finally, this study used a cross-sectional design. A major limitation of a cross-sectional study design is its inability to ascertain a temporal association between the disease and exposure if the exposure varies (Aschengrau & Seage, 2008). Significance of the Study This study was unique because it addressed an area of diabetes research that has received little attention. Furthermore, this investigation of the relationship between sleep duration and T2DM among African American adults has significant implications for research, clinical practice, policy, and positive social change. This study’s results contributed to the existing body of literature by providing a much-needed better understanding of how sleep duration may contribute to the high rates of T2DM observed among African American adults. By having a better understanding of the sleep durationT2DM relationship, strategies can be developed to identify and prevent individual, social, and environmental factors contributing to adverse sleep duration, thus reducing the T2DM burden among African American adults. In addition, this better understanding of the sleep duration-T2DM relationship can be used to better inform existing and future policy, clinical practice, and health promotion programs aimed at preventing or reducing T2DM in the African American adult community. Finally, the results of this study can bring about social change by empowering African American communities suffering from T2DM through the provision of additional knowledge to combat this deadly disease.
14 Summary and Transition This chapter has provided a summary of how African American adults are disproportionately affected by T2DM. This chapter also briefly discussed the current established risk factors such as obesity, physical inactivity, smoking, and low SES, as well as their inability to explain the racial differences in T2DM outcomes observed in African American adults. This chapter additionally discussed the role of sleep duration as a potential significant predictor of T2DM among African American adults. Finally, this chapter outlined the purpose, nature, research questions, hypotheses, theoretical framework, potential limitations, delimitations, and potential significance of this study. Chapter 2 will provide systematic evaluation of the existing body of literature to gain insight to the existing gap of knowledge regarding sleep duration and the T2DM burden in African American adults. Chapter 3 will provide an in-depth characterization of the study methodology and design that was employed. Chapter 4 will present the findings of this study. Finally, an interpretation of the findings will be presented in Chapter 5.
15 Chapter 2: Literature Review Introduction The primary objective of this chapter is to provide a systematic evaluation of the existing body of literature related to sleep duration and T2DM. This chapter also systematically examines the current body of literature related to the sleep duration-T2DM relationship in order to gain a better understanding of what the current research says about sleep’s role as an independent predictor among African American adults. This chapter begins with a focus on the current burden of T2DM seen in African American adults. This review will then focus on the existing body of literature regarding sleep duration, its burden on the U.S. population, sleep duration among African American adults, and a review of existing studies that have examined the sleep duration-T2DM relationship. This review will then transition to a discussion of the common covariates that are used in sleep duration-T2DM research,. Finally, measures used to ascertain sleep duration, covariates, and diabetes status in sleep duration-T2DM research will be discussed. This review was conducted using various databases including CINAHL, EBSCO Host, Google Scholar MEDLINE, PubMed, SAGE Premier, and ScienceDirect. To locate relevant articles for this literature review, keywords and phrases such as type 2 diabetes, type 2 diabetes and African Americans, type 2 diabetes risk factors, obesity and type 2 diabetes, physical activity and type 2 diabetes, smoking and type 2 diabetes, dietary intake and type 2 diabetes, sleep duration, sleep duration and type 2 diabetes, socioeconomic status and type 2 diabetes, sleep duration and African Americans, were
16 used. Over 100 relevant articles were identified and included in this review. The publication dates of these articles range from 2003 to 2013. T2DM and its Burden in African American Adults Over a century ago, European Americans were thought to be at highest risk of T2DM (Leopold, 1931; Tuchman, 2011). In this same time period, African Americans were viewed as the populations with the lowest risk of, and in some cases were thought to be genetically immune to, T2DM (Leopold, 1931; Tuchman, 2011). Over time, epidemiological evidence has painted a completely different picture of the observed racial disparities in T2DM. Currently, African American adults bear a significant burden of T2DM. Over the last 3 decades, the prevalence of diagnosed T2DM among African American adults increased from 4.5% in 1980 to nearly 9.5% by the end of 2010, compared to the T2DM prevalence of 2.6% in 1980 to 6.0% in 2010 among their European American counterparts (CDC, 2011b). Diabetes incidence among African American adults in 2010 was 13.0 per 1,000, compared to 7.7 per 1,000 among European Americans (CDC, 2011b). African American adults additionally face disproportionate rates of diabetes related complications that include End-Stage Renal Disease (Collins et al., 2010), Hyperglycemic Crisis (Smiley, Chandra, & Umpierrez, 2011), amputations, and death (Lanting et al., 2005). Increased rates of overweight/obesity (Hu et al., 2004; Krishnan et al., 2007; Wang et al., 2005; Weinstein et al., 2004), physical inactivity (Hu et al., 2001; Hu et al., 1999; James et al., 1998), smoking (Carlsson, Midthjell, & Grill, 2004;Will et al., 2001; Yeh et al., 2010), and low SES (Cunningham et al., 2008; Maskarinec et al.,
17 2009; Maty et al., 2010; Ross et al., 2010; Sims et al., 2011) have been implicated as the most common factors contributing to the increased risk of T2DM. However, African American adults continue to demonstrate excess T2DM risk when compared to their European American counterparts even after controlling for these risk factors (Brancati et al., 2000; Cabassa et al., 2011; Shai et al., 2006). Brancati et al. (2000) sought to assess the racial differences in T2DM incidence, as well as characterize the risk factors contributing to increased risk of T2DM among 2,646 African American and 9,461 European American adults, aged 45-64 years, from the Atherosclerosis Risk in communities (ARIC) Study. In a model adjusted for age, body mass index (BMI), waistto-hip ratio, family history of diabetes, physical activity, alcohol consumption, smoking, dietary energy intake, and education, they found that African American adult women maintained a 1.8-fold increased risk of T2DM (RR 1.85, 95% CI= 1.55-2.21), compared to European American women. Adjusting for similar covariates, they found that African American adult men had a 1.6-fold increased risk of T2DM (RR 1.62, 95% CI=1.321.99), compared to European American men. Shai et al. (2006) also assessed the racial differences in T2DM among 78,419 European American, Asian, Hispanic, and African American adult women aged 30 to 55-years-old, from the Nurses’ Health Study. After adjusting for age, BMI, alcohol consumption, family history of diabetes, physical activity, smoking status, and dietary score, they found that African American women had a 1.3-fold increased risk of T2DM (RR 1.38, 95% CI= 1.15-1.66), compared to European American women. Finally, among 34,653 European American, Hispanic, American Indian, and African American adults, it was also reported that African American adults
18 had a 1.4-fold increased risk of T2DM (OR 1.49, 95% CI=1.22-1.83), compared to their European American counterparts after adjusting for the same covariates (Cabassa et al., 2011). The results of these studies suggest the need to perform an epidemiological investigation to evaluate other possible risk factors that could be responsible for observed racial differences in T2DM risk among African American adults. A growing body of evidence suggests that sleep duration is a significant predictor of T2DM (Ayas et al., 2003; Beihl et al., 2009; Chao et al., 2011; Gangwisch et al., 2007; Gottlieb et al., 2005; Kita et al., 2012; Lou et al., 2011; Mallon, Broman, & Hetta, 2005; Najafian et al., 2013; Yaggi, Araujo, & McKinlay, 2006). An Overview of Sleep Until the mid-20th century, the majority of society conceptualized sleep as a temporarily inactive period of daily living (National Institute of Neurological Disorders and Stroke [NINDS] & National Institutes of Health [NIH], 2007). Since this time, scholars have been able to gain a better understanding of the sleep process. From these advances, it is recognized today that the human brain is very active during the period known as sleep. Sleep in human beings is generally characterized by the period of sustained inactivity and lowered sensitivity to external stimuli (Zepelin et al. 2005). The process of sleep in human beings further characterized by cycles of non-rapid eye movement (nonREM) and rapid eye movement (REM) stages (Hobson, 2005; NINDS & NIH, 2007).
19 For adult human being, 80% of the sleep process is spent in non-REM stages while 20% is spent in the REM stage (NINDS & NIH, 2007). Despite the general definition of sleep, an exact purpose remains an elusive and controversial subject (Krueger, Obal, & Frank, 1999; Mignot, 2008; Rail et al., 2007; Zepelin, Siegel, & Tobler, 2005). For example, the process of sleep has been theorized as a means for the human body to conserve energy (Zepelin et al., 2005), a means to aid memory and learning in the human brain (Born, Rasch, & Gais, 2005; Karni et al., 1994; Stickgold, 2005; Yoo et al., 2007), and macromolecule biosynthesis cellular component restoration (Mignont, 2008). Despite these differences in the conceptualization of sleep, it is universally recognized that sleep is an actively controlled process that is essential to human health (Hobson, 2005; Mignot, 2008). Sleep Duration and Health Outcomes Sleep duration is a significant contributor to health outcomes in human beings. Adverse sleep patterns have been demonstrated to reduce neuro-cognitive function and increase daytime sleepiness (Alvarez & Ayas, 2004). A significant body of literature also indicates that short and long sleep duration are associated with increased mortality (Patel et al, 2004), obesity (Gangwisch et al., 2005; Lyytikainen et al., 2010; Patel et al., 2006; Watanabe et al., 2010) and T2DM (Ayas et al., 2003; Beihl et al., 2009; Chao et al., 2011; Gangwisch et al., 2007; Gottlieb et al., 2005; Kita et al., 2012; Lou et al., 201; Mallon et al., 2005; Tuomilehto et al.,2008; Yaggi et al., 2006). Several physiological mechanisms, by which sleep duration affects human health, have been suggested.
20 Hypothesized mechanisms by which sleep duration affects health include altered metabolic and endocrine function (Redwine et al., 2000; Spiegel, Leproult, & Van Cauter, 1999), altered cardiovascular function and increased activity of the sympathetic nervous system (Kato et al, 2000). In a randomized controlled trail, Spiegel et al. (1999) found that 11 young men experienced lower glucose tolerance and increase cortisol concentration after being subjected 4 or fewer hours of sleep. In another randomized controlled trail, Redwine et al. (2000) found that 31 healthy men experienced increased levels of pro-inflammatory cytokines after being subjected to sleep deprivation. Yet in another randomized controlled trail, Kato et al. (2000) found that eight healthy men and women experienced an increase in mean blood pressure from 82mm Hg to 86mm Hg (P= 0.01), after being deprived of one night of sleep. Sleep Duration in the U.S. Population Sufficient sleep is generally determined by the amount of sleep an individual requires. Determining the amount of sleep one needs is complex, as it depends on several variables such as lifestyle and environmental factors (Bonnet & Arand, n.d.; National Sleep Foundation, 2013). Despite the complexities of sleep requirements, the National Sleep Foundation (2013) recommended 7-9 hours of sleep for a healthy adult. Short and long sleep duration among the U.S. population is a rising problem. Despite the levels of sleep recommended by the National Sleep Foundation (2013), the amount of U.S. adults who obtain fewer than 7 hours of sleep continues to rise (National Center for Health Statistics, 2005; National Sleep Foundation, 2013). Over the last 50 years, the average amount of sleep reported by U.S. adults has gone from an average of 8.5 hours in 1960
21 (Kripke et al., 1979), to 7 hours in 2005 (National Sleep Foundation, 2005). Furthermore, the U.S. adult population, reporting fewer than 7 hours of sleep, has increased from 30% in 1998 to 40% in 2005 (National Sleep Foundation, 2005). Despite the demonstrated effect of longer sleep duration on health, few data exist regarding its proportions and trends among the U.S. population (Krueger & Friedman, 2009; Patel et al., 2006). It has been reported that 8.5% of 110,441 U.S. adults from the National Health Interview Survey reported sleeping 9 hours or more (Krueger & Friedman, 2009). Over 5% of the population of 85,700 middle-aged women from the Nurses Health II Study reported sleeping 9 hours or more (Patel et al., 2006). Sleep Duration among African American Adults African American adults have reported lower mean sleep duration than their European American counterparts (Hicks et al., 1999; Lauterdale et al., 2006). Studies evaluating sleep duration also suggested that African American adults are more likely to exhibit lower sleep rates, compared to their European American counterparts (Hale & Do, 2007; Krueger & Friedman, 2009). It has been reported that African American adults have average sleep duration 6.94 hours per night, compared to 7.18 hours for European Americans and 7.10 hours for Hispanics (Hicks et al., 1999). Examining sleep duration differences by gender, it has been reported that African American women and men have average sleep duration of 6.4 hours and 6.0 hours respectively, compared to 7.0 hours for European American women and 6.7 hours for European American men (Lauterdale et al., 2006).
22 Some studies also suggest that African American adults are at higher risk of both short and long sleep duration, compared to European Americans. Among a study population of 110,441 U.S. adults from the 2004-2007 NHIS, African American adults have been reported to have a 2-fold increased risk of reporting sleeping fewer than 5 hours (OR 2.00, 95%, CI=1.84-2.17, p < 0.001), when compared to European Americans (Krueger & Friedman, 2009). Similarly in this population sample, African American adults have been reported to have 1.7-fold increased risk of reporting sleeping 9 or more hours (OR 1.72, 95% CI=1.59-1.87, p < 0.001), when compared to European Americans (Krueger & Friedman, 2009). Studies Examining the Sleep Duration and T2DM Relationship Sleep duration’s role as a predictor of T2DM remains an understudied area in diabetes research. It is only within the last decade that large-scale epidemiologic studies, examining this relationship, have been conducted (Ayas et al., 2003; Beihl et al., 2009; Bjorkelund et al., 2005; Chao et al., 2011; Gangwisch et al., 2007; Gottlieb et al., 2005; Hayahsino et al., 2007; Kita et al., 2012; Lou et al., 2012; Mallon et al., 2005; Najafian et al., 2013; Tuomilehto et al., 2008; Yaggi et al., 2006; Vgonztas et al., 2009; Zizi et al., 2012). In a recent meta-analysis of sleep duration-T2DM research, Cappuccio et al. (2010) identified only seven studies examining this relationship. Of the currently published literature examining the sleep duration-T2DM relationship, only two studies examining this relationship among African American adults have been published to date (Beihl et al., 2009; Zizi et al., 2012).
23 Existing epidemiologic studies, examining the sleep duration-T2DM relationship, have used a range of study designs, sample sizes, and results. Most studies evaluating the sleep duration-T2DM relationship have used a longitudinal study design (Ayas et al., 2003; Beihl et al., 2009; Bjorkelund et al., 2005; Gangwisch et al., 2007; Hayahsino et al., 2007; Kita et al., 2012; Lou et al., 2012; Mallon et al., 2005;Yaggi et al., 2006). Fewer studies have evaluated this relationship using a cross-sectional study design (Chao et al., 2011; Gottlieb et al., 2005; Lou et al., 2012; Najafian et al., 2013; Tuomilehto et al., 2008; Vgonztas et al., 2009; Zizi et al., 2012). All designs used in these studies were appropriate for the variables of interest measured. Some studies have indicated that sleep duration is a significant independent predictor of T2DM (Ayas et al., 2003; Beihl et al., 2009; Chao et al., 2011; Gangwisch et al., 2007; Gottlieb et al., 2005; Kita et al., 2012; Lou et al., 2012; Mallon, Broman, & Hetta, 2005; Yaggi et al., 2006). This significant relationship has been demonstrated among women (Ayas et al., 2003; Gottlieb et al., 2005; Mallon et al., 2005; Tuomilehto et al., 2008), men (Ayas et al., 2003; Gottlieb et al., 2005; Tuomilehto et al., 2008; Yaggi et al., 2006), European American, Hispanics (Beihl et al., 2009), and other ethnic groups (Chao et al., 2011; Hayahsino et al., 2007; Kita et al., 2012; Lou et al., 2012; Najafian et al., 2013). To the contrary, it remains unclear if sleep duration acts as an independent predictor of T2DM among African American adults (Beihl et al., 2009; Zizi et al., 2012). Specifically, few studies have examined the sleep duration-T2DM relationship among African American adults (Beihl et al., 2009; Zizi et al., 2012). Of these few, none suggest significant relationship between aberrant sleep duration and T2DM among African
24 American adults. The following is a discussion of the current body of literature addressing the sleep duration-T2DM relationship. The Sleep Duration-T2DM Relationship by Gender Women. Among the epidemiological studies examining the sleep duration-T2DM relationship by gender, most suggest that this association is significant among women (Ayas et al., 2003; Gottlieb et al., 2005; Tuomilehto et al., 2008). It has been reported among 70,026 American women, aged 30-55 years, from the Nurses’ Health Study, that those self-reporting a sleep duration of 5 or fewer hours per day maintained a 1.3-fold increased risk of T2DM (RR 1.37, 95% CI= 1.07-1.77), compared to women reporting 8 hours of sleep per day, after controlling for hypertension, high cholesterol, smoking, physical activity, snoring, depression, family history of diabetes, postmenopausal hormone use, and shift working. Controlling for the same covariates among this same group of women, those sleeping in excess of 9 hours had a 1.3-fold increased risk of T2DM (RR 1.36, 95% CI= 1.04-1.73), compared to women reporting 8 hours of sleep per day. Similar to the findings of Ayas et al. (2003), a study in a sample of 1,434 Finnish women, aged 45–74 years from the Finnish type 2 diabetes (FIN-D2D) survey, found that those self-reporting a sleep duration of 6 or fewer hours per day had a 2.5-fold increased odds of T2DM (OR 2.55, 95% CI= 1.21-5.35), compared to women reporting a sleep duration of 7 hours per day, after adjusting for the covariates of age, BMI, smoking, probability score, Central Nervous System (CNS)-affecting medication, sleep apnea, and leisure-time physical activity (Tuomilehto et al., 2008). In this same group of women,
25 those sleeping 8 or more hours per day, had a 1.7-fold increase odds of T2DM (OR 1.76, 95% CI= 1.12-2.61), compared to women reporting a sleep duration of 7 hours, after adjusting for the same covariates (Tuomilehto et al., 2008). Yet in another outcome similar to those of Ayas et al. (2003) and Tuomilehto et al. (2008), researchers discovered in a sample of 764 American women, aged 53-93 years from the Sleep Heart Health Study (SHHS), that those reporting a sleep duration of 5 or fewer hours per day had a 1.8-fold increased odds of T2DM (OR 1.83, 95% CI= 1.07-3.11, p
