Knowledge and Beliefs about Smoking among Urban African Americans with Type 2 Diabetes : A Qualitative Inquiry

Background. Despite elevated risks of cardiovascular disease among African Americans with diabetes, few investigators have examined beliefs about cigarette smoking and smoking cessation among this underserved population. Methods. Focus groups (n = 4) and a short survey were used to assess cigarette use patterns, perceived smoking health effects, preferences for treatment, and attitudes toward smoking cessation among low income, African American smokers with type 2 diabetes. Results. Participants (n = 25, M = 48.5 years [+10.23], 60% female) smoked 20.9 (+12.54) cigarettes per day (CPD) and had on average 3.2 (+6.12) 24-hour quit attempts in the past year. Few had used clinic support or pharmacotherapy in their quit attempts. Participants smoked to control health-related stress, keep their appetite down, and to control their glucose. Conclusion. The results suggested that African American smokers with diabetes may be an appropriate group to target for smoking cessation. Cessation recommendations are discussed. KJM 2009; 2(2):20-32.


Introduction
Cigarette smoking is the single most important preventable cause of morbidity, mortality, and excess health costs in the United States, accounting for approximately 427,000 deaths each year. 1,2Despite considerable prevention and intervention efforts, approximately 50 million US adults continue to smoke cigarettes, 3,4 including 5.2 million African Americans (AA). 5,6][9][10][11] While African Americans tend to smoke fewer cigarettes per day 12 and begin smoking later in life 13,14 than other groups in the US, they bear a disproportionate share of tobacco-related disease. 6,15When compared to European Americans, AAs have the highest incidence rates for all cancers combined, 6,15 are at elevated risk for cerebrovascular 16 and heart disease, 17,18 and have twice the rates of premature death attributable to cardiovascular disease. 4,8ccording to the National Institute of Health, the total prevalence of diabetes among non-hispanic whites is 8.7%.African Americans have disproportionately higher rates (i.e., 13.3%). 19Diabetes is associated with a substantial increase in the risk of cardiovascular disease (CVD) and this risk is increased further by cigarette smoking.Smokers with diabetes have an estimated two-to three-fold increase in the risk of cardiovascular complications compared to smokers without diabetes. 20,21he cardiovascular risks of smoking and diabetes in combination are nearly 14 times higher than the risk of either smoking or diabetes alone, much greater than expected from the simple combination of smoking and diabetes. 22Smoking also substantially increases the risk for additional diabetesrelated complications including the development of hypertension and microvascular complications (e.g., neuropathy, nephropathy, and microalbuminuria). 21,23,24espite these alarming statistics, the prevalence of smoking among persons with diabetes appears to mirror that among the general population. 257][28][29] Whether this is the result of concern that smoking cessation may compromise diabetes self-management efforts, the belief that patients with diabetes may not be interested in quitting, or limited awareness of the effectiveness of smoking cessation programs, is not clear.Research to date suggested that persons with diabetes may be less likely to quit smoking on their own than other smokers and may be more likely to fail in organized smoking cessation programs. 21,25ecently, Hokanson and colleagues 30 completed a randomized, controlled smoking cessation trial with 114 participants with type 2 diabetes.The results indicated that an intensive cessation intervention integrating face-to-face motivational interviewing plus telephone counseling into a standard diabetes self-management training program resulted in a trend toward greater abstinence at three months when compared to standard care.Although this trend was not observed at the six-month follow-up, the integration of smoking cessation into an existing diabetes education program did not impact diabetes management negatively, including A1c values.Further, our research team recently conducted a secondary analysis of three large, randomized clinical smoking intervention trials examining cessation rates among African American participants with self-reported diabetes.Our analyses concluded that there was a doubling of quit rates among persons with diabetes as compared to those with no diabetes. 31hus, results of these recent investigations suggested that persons with diabetes may be highly responsive to cessation efforts and that diabetes management may not be impacted negatively when cessation is targeted.
Because the knowledge, attitudes, beliefs, and preferences of smokers facilitate maximum receptivity to programs, these are important considerations when developing effective cessation interventions. 32ualitative research is well-suited to provide in-depth answers to these complex issues and may suggest ways to intervene with a target group successfully.Further, focus groups are well-suited for this study because the collaborative effort of the groups stimulated participant discussion of a topic about which they did not think about individually. 33Thus, this qualitative study employed focus groups and a short survey to assess cigarette use patterns, perceived health effects of smoking, preferences for treatment, and attitudes toward cessation interventions among urban, African American smokers with type 2 diabetes.

Methods
Participants.Eligible participants were 18 years or older, African American, selfreported as having diabetes and smoked five or more cigarettes per day.Exclusion criteria included variables that would hinder participation in a group discussion or ability to be contacted.These criteria included homelessness, marked inappropriate affect or behavior, or impaired cognition.Participants also were excluded if they had previously participated in any prior formal smoking cessation program.
Procedures.The research protocol was approved by the University of Kansas Medical Center's Human Subjects Committee prior to implementation.Participants were recruited at a community health center (Swope Health Services) that serves under-and uninsured, predominantly African American, patients.Flyers were posted in several clinics including Internal Medicine, Nutrition, Podiatry, and the patient pharmacy.In addition, a research assistant sat at a booth and handed out flyers to interested participants in the lobby of the health center.The one-page recruitment flyer was designed to recruit African American adults who had diabetes, smoked cigarettes, and were interested in participating in a group discussion about smoking and diabetes.Potential participants who responded to the flyer were screened for eligibility by phone or in the health center lobby.
Four focus groups were held over a twoweek period.A clinical psychologist, with focus group facilitation training, moderated the groups.Three research assistants, one of whom was a representative of the Swope community, greeted participants, offered snacks, completed forms, video-and audiorecorded the groups, and distributed incentives.All participants provided both written and verbal informed consent to be audio-and video-recorded.
Following the consent procedure, participants completed a brief survey documenting demographic, tobacco-related and diabetes-specific health information.The assistant moderator read all survey questions aloud while a research assistant circulated to assist individual participants as needed.The focus groups were led using a moderator's guide developed specifically for this study.Researchers with experience in smoking cessation, diabetes, and focus group methodology developed the guide. 34,35he moderator followed a semi-structured interview format using open-ended questions to stimulate discussion about knowledge, attitudes, and beliefs about smoking and diabetes.Table 1 displays example questions from the moderators guide.
During each focus group, the moderator probed participants' responses and encouraged all members to participate.The group discussions lasted approximately 90 minutes.
After the groups ended, the research assistant collected a list of participants' medications in an attempt to corroborate their self-reported diagnosis of diabetes.After medications were recorded, participants received a $30 Wal-Mart ® gift card as compensation for their travel cost, time, and effort.Data collection stopped after data saturation had occurred for the majority of our topic areas (i.e., no new data would be found by conducting further focus groups). 33ata analysis.Survey data were doubleentered and range checks were performed.Descriptive statistics were computed using SPSS Version 13.0 (SPSS Inc., Chicago, IL).Audio-recordings of the focus groups were transcribed verbatim by a contracted professional transcription service.The focus group moderator proofread each transcript and compared them to the video-recordings to check for completeness and accuracy.Three independent coders deductively categorized transcripts by hand into six major topic areas using initial codes developed by the research team based on the focus group moderator's guide.Coders then coded transcripts by hand within each major Table 1.Sample questions from moderators guide.topic area using an inductive approach whereby categories and concepts emerge from the text and are linked together. 33This approach allowed the data to represent itself.
A fourth independent researcher crosschecked inductive codes and identified minor discrepancies in the coding and varying terminology used by each coder to describe the same content.Cross-checking codes provided a measure of how well the data were indexed and, thus, gave a qualitative measure of inter-coder reliability. 36Overall, the independent researcher found high inter-coder reliability and identified major themes within the codes.The research team then met as a group to discuss the major themes and to address any discrepancies.
Ten major saturated themes emerged across coders, as well as several unsaturated topics and themes that provided avenues for future research.

Results
Participants.Of the 59 people who responded to the study flyers, 50 were eligible to participate.Reasons for exclusion were not having diabetes (n = 2), being homeless or living in a treatment facility (n = 4), and previous participation in smoking cessation groups (n = 3).Of those eligible, 25 did not show for their scheduled focus group and efforts to reschedule were not successful.Thus, the final sample included 25 participants.
As detailed in Table 2, participants were on average middle-aged (M = 48.5+10.2years) and female (60%).Most had at least a high school level education (32%) or some college (36%), were divorced (36%) or single (24%), and most did not have any health insurance (73%).Participants had been diagnosed with diabetes for an average of 12.2 years, and the average age at diabetes diagnoses was 36.8 years (SD+13.33).Forty-four percent used insulin Tobacco-related variables.As detailed in Table 3, many participants (68%) were considering to quit smoking in the next six months and 60% had cut down the number of cigarettes smoked to lower their health risks.
Further, they moderately were motivated (M=6.04,SD+2.59) and confident regarding quitting (M=5.60,SD+3.4).Also, most participants moderately were concerned about the effects of smoking (M=6.23,SD+3.73) and weight gain (M=5.16,SD+4.0) on their diabetes complications should they quit smoking.When asked about prior quit attempts, participants reported that they relied on their own will power (76%), spirituality (62%), and support from family and friends to help them quit (32%).Fewer had used formal treatment (4%) or nicotine replacement therapy (32%) or buproprion (4%) in their prior quit attempts.Thematic Analyses By Topics Ten unique themes emerged across focus groups and reached saturation (see Table 4).
Attitudes toward smoking.The most commonly reported reason to continue smoking was to control weight through appetite suppression.Many participants also expressed the belief that smoking helped them to control their blood sugar level.They also reported that smoking helped them cope with high levels of stress related to their diabetes self-care and life stress (e.g., financial difficulties, child care).Alternately, participants reported that the most important reasons to quit smoking were the cost and the known health consequences.
Examples of participant comments were: "As long as I am smoking, I won't be feeding myself all the time."(male participant) "Sometimes if I get upset at somebody, like one of my kids, I light a cigarettethat's how I calm down."(female participant) "If I have five dollars and I know I need gas, I think, 'well, maybe I could put three dollars in the gas tank and take this two dollars and buy some cigarettes'…I want to quit to save money."(female participant) Beliefs and knowledge about smoking and diabetes.Most participants endorsed the belief that smoking would increase their risk of diabetes-related negative health outcomes.However, few participants were able to articulate their understanding of the mechanism by which smoking and diabetes might interact to elevate diabetes-related risk factors.
Examples of participant comments were: "I'm not clear on the connection between smoking and diabetes, but they say it makes diabetes worse."(female participant) "Diabetes already decreases our circulation, but smoking makes that even worse because we are reducing the amount of oxygen that flows through our bodies.I think that is why we start to lose feeling in our fingers and toes."(female participant) "I've got diabetes, heart disease, high cholesterol, and I smoke.I shouldn't smoke and the doctors even say I need to quit smoking cigarettes, but I keep telling them I don't want to." (male participant) Participants believed that smoking decreased their appetite and that quitting smoking would result in weight gain, a known hazard to their diabetes management.Comments included: "When I found out that I had diabetes, I lost weight, but then once the medicine kicked in, I gained weight.Then, I started smoking again so that I wouldn't eat as much."(female participant) "Since I'm a diabetic I feel that I have to watch my weight, so I don't eat that much.That is why I'm scared to stop smoking.I'll gain weight."(female participant).Change in smoking since diagnosed with diabetes.
Most participants endorsed making a smoking quit attempt since being diagnosed with diabetes.However, they explained that cessation resulted in the experience of additional stress and difficulty focusing on their diabetes self-care.Because our inclusion criteria included being a current smoker, it is unknown whether other persons with diabetes who have had success in quitting met with these same barriers yet were able to overcome them.
Examples of participant comments were: "When I was diagnosed with diabetes, I had already been smoking for 30 years.I know that I should have quit before then, but now I am under so much more "I'm more aware now since I've been diagnosed with diabetes, I'm much more aware of how much I smoke."(male participant) Topics Themes Attitudes toward smoking (positives and negatives).
Most common reasons for smoking were to cope with stress or negative emotions (calming mechanism) or to control or lose weight.Most common reasons to stop smoking were the health consequences and cost.Beliefs and knowledge about smoking and diabetes.
Participants believed that smoking increased their risk for all health outcomes, though there was not a clear understanding of how (some discussion of ties to insulin levels in blood, but this is unclear).Participants believed smoking decreased their appetite.They also thought that quitting makes you gain weight and that it would negatively affect diabetes.Change in smoking since diagnosed with diabetes.
Since diagnosis, many participants have attempted to quit, but with no success.Many participants voiced concern that quitting will result in worsened diabetes self-management.Prior experiences with quitting.
Most participants had tried to quit, with the most common method being "cold turkey."Few pharmacotherapy products were discussed, but participants agreed that insurance benefits should cover cessation products.Beliefs and opinions about quitting.
Most participants wanted to quit and believed it was important to quit, but were not motivated to quit or confident they could achieve cessation.Treatment preferences.
Participants were undecided on whether persons with diabetes should have a separate smoking cessation program from other smokers or if diabetes care should be kept separate.
Ideas about what should be included in a smoking cessation program and barriers to that program are mixed and mirror a heterogeneous group of smokers.
"Well, I smoke more because it seems like there's an urge to smoke more since I'm taking pills.It [diabetes] is kind of threatening."(male participant) "As soon as I leave here, I am going to smoke a cigarette.What am I going to stop smoking for?I am already messed up anyway."(male participant) Prior experience with quitting.Most participants had tried to quit in the past (most using the "cold turkey" method) and believed that it is important for them to quit.Few had tried nicotine replacement therapy or bupropion to quit.The few who had tried found it helpful, therefore, they focused their discussion on their disappointment that insurance did not cover these cessation aids or that they did not have insurance.Finally, participants were unaware of the existence of state-funded cessation services (e.g., quit lines).
Examples of participant comments were: "The pill (bupropion) mellows you out.Someone could say something very negative to me and it wouldn't bother me.I was more pleasant.So, yeah, to me, if I had it to do all over again and I could ever get that pill, yes, I would go back."(female participant) "The last time I was in the hospital, I was in a diabetic coma.They gave me the nicotine patch, but because I do not have insurance, I could not keep using it."(female participant) "I have made an honest effort to quit smoking, but I tried to get the patch along with my medicine.Since I do not have insurance and I am already using the hospital discount to get my medicine, I can't get the patch."(female participant) Beliefs and opinions about quitting.Most subjects endorsed the belief that it is important for them to quit for their health, particularly given their diabetes, but were not ready to make that change nor confident that they could do so. Examples

Discussion
Our qualitative findings offer insights for understanding and developing intervention strategies for smoking cessation among African American adults with diabetes.In general, our participants expressed a desire to quit smoking, but voiced concerns regarding the potential impact of quitting on their diabetes self-management.Specifically, participants were concerned about the impact of quitting on their stress level and cessation-related weight gain.In addition, several participants noted that they had increased their smoking level after receiving the diagnosis of diabetes.
The association between diabetes diagnosis and increased smoking behavior and decreased motivation to quit needs to be determined through further investigation.However, if smokers are at risk for an increase in smoking following diabetes diagnoses, then using the diagnostic visit as a potentially "teachable moment" to introduce the importance of cessation and providing treatment options may be an important preemptive strategy.
Using cigarette smoking to cope with stress and weight gain are consistent themes found in the smoking literature.However, the stress management properties of smoking may be particularly important in our sample given their substantial sources of life stress (e.g., poverty status, ethnic minority, no health insurance, and diabetes).As such, participants overwhelmingly agreed with one participant when she mentioned that "…without stress there would be no need for smoking."Previous research has focused on the stress inherent with the self-management requirements of diabetes and the negative impact of stress on diabetes care.Thus, although most diabetes programs include instruction on the importance of stress management in diabetes self-care, it is clear that if we are to address the devastating health effects of the combination of diabetes and smoking, providers must offer patients alternative stress management strategies.
Our participants frequently mentioned continuing to smoke to prevent weight gain and a few mentioned a noticeable association between smoking and glycemic control (in both "positive" and "negative" directions).Although participants recognize that weight gain may have a negative impact on their diabetes self-care, they did not seem to have the same level of awareness regarding the negative consequences of the combination of diabetes and smoking.One participant reported that she quit smoking but that when she gained weight, she returned to smoking to decrease her appetite so that she might be in more control of her eating habits.
Weight gain following cessation is indeed a risk.Iino and colleagues 37 found that in a sample of smokers with diabetes who quit, body weight at six months increased by approximately three pounds.This increase is less of a health risk than the alternative of continuing smoking.However, the weight gain is a risk with which individuals with diabetes appear to be concerned, possibly because it is physically noticeable.

Table 4 :
Focus group topics and themes.
stress, and now I have gained more weight too.I know I need to quit, but now it is even more of a crutch for me."(female participant)