Abstract
This study set out to explore the experience of food cravings and thirst, and their management strategies in patients on hemodialysis. Semi-structured interviews with N = 32 hemodialysis patients were analyzed thematically. Findings indicated that food cravings and thirst were common in everyday life of patients and resulted in different emotional responses. A combination of cognitive and behavioral strategies was employed, including avoidance, controlled consumption, and substitution. Self-monitoring and compensatory strategies were also used to prevent or compensate for lapses. These findings lay the groundwork for future work aimed to improve adherence in end-stage renal disease patients.
Introduction
With aging and increase in diabetes and hypertension, the numbers of patients developing renal complications (end-stage renal disease (ESRD)) and requiring renal replacement therapy (RRT), the most common being hemodialysis, are rising rapidly (Vathsala, 2007). ESRD, however, involves a complex and demanding behavioral regimen above and beyond dialysis therapy. This includes multiple medications and nutritional management (e.g. restriction of dietary intake of phosphate- and potassium-rich foods and salt) and reduction of fluid intake. As a result, the consumption of many palatable/pleasure foods and/or drinks is not generally encouraged. Patients are called to learn and assimilate dietary behaviors that contradict lay understandings of “healthy” living. For instance, the “renal” diet is typically incompatible with general healthy eating principles as it precludes a range of fruits and vegetables, and restricts fluid intake including water. The restricted consumption of fluids makes it difficult to quench one’s thirst and manage the intake of multiple medications that constitute the key element of renal regime.
Previous research has indicated food cravings (Cummings et al., 1982) and thirst (Mistiaen, 2001) to be among of the biggest obstacles in the diet and fluid-intake adherence of ESRD patients. Non-adherence to dietary and fluid-intake recommendations is a serious and rampant problem among ESRD patients and is estimated to be present in as much as 14 percent and 20 percent of the ESRD population around the world, respectively (Saran et al., 2003).
While there is a handful of studies on cravings in patients with diabetes (Strachan et al., 2004; Yu et al., 2013), a major cause and comorbid condition of ESRD (Singapore Renal Registry, 2010), most of the work related to food cravings has been undertaken on people with eating disorders/or obesity (Barnes et al., 2011; Barnes and Tantleff-Dunn, 2010). Little is known on how cravings operate or are managed in the context ESRD. Most notably, the evidence produced by this line of work on the effectiveness of specific strategies remains mixed and conflicting and hence cannot be used to inform research and practice in ESRD. For instance, some studies on the general and obese populations showed that control strategies such as suppression, albeit commonly used to cope with cravings do not always work well as they may lead to overconsumption (Hooper et al., 2012), increased binge eating and other maladaptive eating behaviors (Barnes and Tantleff-Dunn, 2010). In contrast, other studies have failed to show such rebound effects (Mann and Ward, 2001; May et al., 2010).
Even more bothersome and challenging than dietary restrictions is the need to restrict fluid intake (Fan et al., 2013) and the experience of thirst in hemodialysis patients. Feelings of thirst have been linked to poor adherence to fluid-intake recommendations (Bots et al., 2004; Porcu et al., 2007). As such managing thirst is a critical goal for patients on dialysis. Medication, substitution through chewing gums and strategies such as rinsing mouth, using ice cubes, or drinking from smaller caps have been shown to help relieve thirst and avoid overconsumption (Arai et al., 2013; Bots et al., 2005; Mistiaen, 2001). While there are more studies on thirst compared to food cravings in ESRD patients, these are usually intervention studies in which a specific thirst coping method is taught to the patients (Arai et al., 2013; Bots et al., 2005). The value of such work where strategies are identified using a “top-down” approach and then evaluated cannot be understated, yet it is of note that there is very little work on patients’ own practices and ways of dealing with thirst or food cravings. Generating “patient-driven” data might prove to be useful to better tailor patient education and/or interventions.
The broader Chinese cultural context that Singapore is largely based on has implications on managing food cravings. In the Chinese culture, meals are often consumed together in shared communal plates with family and friends on a daily basis. Furthermore, the greater emphasis on food in Chinese cultures, relative to others, also meant that many joyous social occasions are often celebrated with generous amounts of food and drinks (Warde and Martens, 2000). This social culture of eating together and celebratory feasting inevitably exposes patients to many craved yet not recommended food items and creates more pressure to consume for fear of being singled out (Gallant, 2003), which may make dietary management more challenging (Griva et al., 2013). In this setting where culture tends to favor a collectivistic approach, interpersonal expectations may become strong drivers of patients’ experience and related outcomes (Symister, 2011). Also within the Singaporean context, the moderately hot and sunny weather all year round makes thirst more salient, perhaps more so than in other non-tropical countries.
Given the problems of food cravings and thirst in ESRD patients, and the inadequate attention devoted to managing such problems, this exploratory study aims to unravel patients’ experience with food cravings and thirst, and the management strategies employed to cope with these.
Method
This qualitative study involving semi-structured interviews was nested within a randomized controlled trial (RCT; Hemodialysis Self MAnagement intervention Randomised Trial (HED-SMART); Griva et al., 2011).
Participants
A total of 32 participants, consisting of both control and intervention participants from the parent RCT, were sequentially recruited. Participation in the RCT was staggered in batches, and all interviews were conducted during follow-up assessments at 3–6 months following baseline assessment and randomization. Participants were recruited based on these criteria:
ESRD patients who have been receiving hemodialysis for at least 6 months;
Aged 21 years and over;
Willing and able to give informed consent;
Able to speak English and/or Mandarin;
No previous diagnosis of functional psychosis or organic brain disorder;
No major visual, hearing, motor, or cognitive impairments that may prohibit completion of the scheduled interviews.
These participants attended regular dialysis sessions at the National Kidney Foundation Singapore dialysis centers.
Procedures
Following informed written consent, arrangement for interviews (scheduling and language, that is, Mandarin or English) was made as per patients’ preference. All interviews were conducted during a dialysis session. An interview guide developed for the purposes of this study was used to explore the following:
Experience of food cravings/thirst, for example, “When/how often do you experience food cravings/thirst?”
Emotions related to coping with food cravings, for example, “How does it feel to resist your food cravings/thirst?”
Coping strategies for cravings and thirst, for example, “How do you convince yourself not to eat/drink that?”
Other coping strategies in general, for example, “What other methods do you use to prevent yourself from eating/drinking more than recommended?”
All interviews were audio-recorded and transcribed verbatim. Interviews conducted in Mandarin were first translated into English. No personal identifiers were used, and recordings were promptly deleted upon transcription. All respondents received a nominal reimbursement (SGD$5) for their participation.
Qualitative analysis
The study was stopped when thematic saturation was achieved from concurrent analysis, that is, when no new concepts emerged in at least four interviews. Transcribed interviews were imported into NVivo 9 software (QSR International™) for coding, searching, and organization of data. The data were analyzed thematically, via the systematic search of themes and repetitions emerging from the data (Braun and Clarke, 2006). Themes were identified using an iterative process. Two coders worked independently on each transcript using open coding approach to identify themes and tag them to facilitate analyses. This was followed by a consensus meeting to arrive at a preliminary codebook. This codebook was applied to transcripts from subsequent interviews and was updated as new codes emerged. The final codebook was then applied to recode all transcripts. At this stage, the inter-rater reliability between both coders was high, kappa coefficient = .89. Coders then discussed the discrepancies and make necessary modifications to create the final dataset.
Study characteristics
All of the patients approached (N = 32 dialysis) agreed to participate (100% response rate). Most of them were ethnic Chinese (N = 22), 19 were conversant in Mandarin, and 13 in English. The age of these patients ranged from 21 to 76 years, and the mean was 52 years (standard deviation (SD) = 11.9 years). The mean dialysis vintage was 6 years (range: 1–15 years). The most common cause of ESRD was diabetic nephropathy (N = 10) (Table 1). The ethnic and gender ratio, mean age, and distribution of ESRD causes in our sample were comparable to the general hemodialysis population in Singapore (Singapore Renal Registry, 2010).
Demographic profile and primary kidney diagnosis of participants.
Results
Overview of themes
The analysis yielded 11 themes that were grouped into emotional responses, cognitive and behavioral strategies to manage food cravings and thirst, as well as other strategies employed to regulate diet and fluid intake in general (Figure 1). Strategies and emotions experienced in relation to food cravings or thirst were not differentiated, so the themes are reported jointly for food cravings and thirst.
Overview of diet and fluid-intake regulation management strategies.
The emotional aftermath of cravings
Most participants (N = 27) discussed their everyday frustrations and challenges of managing recommendations related to diet and fluid intake.
I know I should not eat it, but I still want to eat it. (P20) Especially in a hot day, you want to drink but you know you have to control your intake. (P16)
In tandem with recounting occasions of cravings, participants shared their emotional reactions. A total of 22 participants experienced negative emotions that revolved around feelings of frustration with the diet or fluid restrictions.
Quite frustrating. (P7) Sometimes you feel so disappointed, you feel so regretful. (P32)
Upon successful control of food cravings and thirst, many patients (N = 13) experienced positive emotions. Overcoming temptation was accompanied with contentment and pride that served to reinforce patients’ resolve to self-manage.
Yes, definitely (I’m happy), since I’m able to resist the temptation. (P20) I’ll definitely feel a bit proud that I’m able to control myself. (P30)
Managing food cravings and thirst
In recounting occasions of temptation in everyday life, patients went on to highlight the skills and ongoing commitment needed to deal with cravings. They reported using a combination of different strategies (cognitive, behavioral, or other) applied proactively to pre-empt “high risk” situations, overcome temptation(s) and avoid lapses, or used reactively so as to compensate when lapses occurred.
Cognitive strategies
Thought suppression
A total of 10 patients attempted to ignore or deny the existence of the craved food or fluid.
I try to control my mind. I won’t touch it. (P2) Don’t look at it, don’t think about it. (P11)
Thinking of consequences
Most of the patients (N = 24) reminded themselves of the potential consequences related to poor adherence to dietary and fluid-intake restrictions upon experiencing cravings or thirst. These consequences mainly were related to physical side effects and likely complications and hospitalization. Hinged upon these were thoughts of the financial costs which can be considerable in settings where copayment models of health care are in effect.
When I will think of the discomfort I will experience later, I won’t touch it. (P4) I will think of the spasms I will suffer later and then tell myself not to drink too much. (P11)
Self-perception “me as a dialysis patient”
Eight patients identified themselves with certain identities/profiles that facilitate adherence behaviors. They made references to seeing themselves as dialysis patients while distancing themselves from their past pre-ESRD selves or differentiating themselves from healthy individuals to justify not giving into their cravings. Internalizing their new identity meant accepting boundaries related to food and drinking practices.
We can’t be like them (other healthy people), we have to eat less and we cannot eat too much. (P26) Now that we have this condition, we have to do something about it, we can’t remain like in the past (eat without any restraint), unless you’re not afraid of dying. (P17)
Psychological devaluation
Six participants revised their representations/perceptions of the craved items in order to defend against cravings and thirst. In some cases, patients attempted to devalue the craved items as unimportant or view them as no longer a pleasure, but an adversary that needs to be tamed. This essentially made it less painful for them to resist or reject the craved item: It’s not like these are really precious stuff like abalone, bird’s nest or shark’s fin, where you should eat them because it is very expensive. (P17) I will think of water as my enemy then I won’t drink too much. (P29)
Behavioral methods
These behaviors encompassed both avoidance as well as a range of actions toward the craved objects. The actions ranged from minor/light touch revisions such as controlled intake consumption to more bold and drastic actions such as lifestyle modifications.
Behavioral avoidance
A total of 13 patients attempted to reduce the cravings by physically avoiding the craved item: When I see it I will run away. (P3)
Controlled and paced “giving in”
The majority of the patients (N = 29) used a range of eating and drinking techniques that reflect controlled consumption to deal with cravings. This purposeful yet controlled “giving in” to cravings involved tasting or sampling the desired item(s) so as to satisfy cravings but sticking to small portions. Essential to this were patients’ ability to indulge in favorite flavors/dishes yet maintain good restraint/control: I’ll just give it a try. Just to put something in my mouth, so that I can taste the food. I won’t sit down and eat. (P24) Maybe I’ll just take a bit, maybe just a small cut of it. (P8)
To manage thirst, patients opted for regular yet slowed down intake by using ice cubes or taking small sips. While thirst was not fully satisfied, patients were in a better position to tolerate the discomfort of thirst and keep to recommended levels of fluid consumption: You can suck on an ice cube and then spit it out. (P26) Sip on a drink, but don’t swallow it right away, play with your tongue and then you swallow it. (P32)
Substitution
Eight patients attempted to resist the temptation to drink by substituting drinking liquids with eating other food items that appear to quench their thirst (e.g. sweets, lozenges, or fruits): I’ll take this sweet and sour candy, and other sour items. They’re able to draw out my saliva a lot. (P3)
Lifestyle changes
Some patients (N = 10) undertook broad lifestyle changes so as to better manage their cravings around food or fluid intake. They modified their dietary intake to better control thirst or adapted social and outdoor activities to minimize exposure to hot weather and hence thirst or minimize temptation and social pressure to food or fluid intake. These new routines were seen as important in order to maintain a consistent pattern of dealing with food cravings and thirst: I try not to eat spicy stuff. I just eat plain rice, so that I’ll be less thirsty. (P32) Just try as much as possible to avoid going out when it’s hot, then you can to control it (the thirst). (P6) I try my best to avoid eating out with my friends. (P4)
Thus far, the strategies outlined above are primarily targeted at mitigating cravings and thirst. They are usually used while consuming food and drinks or when faced with cravings. In addition to these, participants also highlighted strategies, which may be helpful to their adherence/self-management goals and not necessarily in response to food cravings or thirst.
Other strategies
Self-monitoring
Five patients set limits in the consumption of certain food items and water or used anthropometric indices to systematically track their behavior during the day to ensure that one does not over-consume.
I’ll measure the amount. If I exceeded the limit, I won’t drink anymore. (P21) The whole day I’ll be checking my weight to see that I don’t over-drink. (P16)
Such strategies seemed to have been ingrained into daily routines and thereby were executed effortlessly and automatically.
Having done it (self-monitoring) for so many years you become so used to it, it becomes a routine. (P30)
The routinization of these behaviors can perhaps be explained by the fact that most of these patients had been on dialysis for a long time (mean dialysis vintage = 6 years).
Compensatory methods
Upon realizing that they had over-consumed (food or fluid), many patients (N = 19) tried to compensate by restricting subsequent intake. This supposedly involves some self-monitoring, but the focus of such strategies is on restoration rather than prevention. The adjustment and modifications undertaken after lapses seem to reflect self-regulatory processes by which patients recognize the lapses and take redressing actions to realign behavior with set goals. Participants described several compensatory methods. Some patients responded to these lapses by attempting to restrict subsequent consumption of the craved item. Others exercised to rid excess fluid in their body via perspiration. A few patients sought to make arrangements for earlier hence longer dialysis sessions so as to flush out excess fluid or nutrients, in response to excessive dietary and fluid intake: I will try to control myself and not to eat them anymore for the next month. (P20) What I usually do is to go for strolls, or exercise to reduce the water. (P16) I will usually call NKF to give me the first shift. (P19)
Compensatory responses were performed whenever necessary, yet seemed to be employed with certain regularity as lapses were common: I will try to exercise every Tuesday to get rid of the extra weight. (P10)
Discussion
Dealing with food cravings and thirst is ingrained into the lives of hemodialysis patients. Patients recounted numerous occasions of temptations in everyday life, yet not all of these were described in terms of deprivation as patients discussed practices around food and fluid control as means to better health and to avoid physical suffering. This was also evident in participants’ discussion of emotional responses to cravings and thirst. Although many shared their disappointment and frustration in that they can no longer indulge, positive feelings, for example, pride were also reported upon successfully resisting temptations and exercising self-control. While excessive negative emotions may undermine motivation and adherence by increasing cravings and subsequent food consumption (Macht, 2008), positive emotions such as pride can facilitate adherence (Katzir et al., 2010).
Our findings indicate that management of food cravings and thirst is intertwined with cognitive strategies and shifts in personal identity as well as a range of behavioral techniques. Most prominent in participants’ accounts were “thinking of consequences,” “avoidance,” and “paced controlled giving in.”
Patients managed to exert cognitive restraint when faced with desired yet not recommended food items or thirst by purposefully reminding themselves of the adverse consequences of excess fluid and poor dietary control, as shown in patients with heart failure (Malpass et al., 2009; Riegel and Carlson, 2002). This involved recalling previously experienced physical symptoms/complications and likely hospitalization. Linked to these were concerns about financial costs of poor health and hospitalization, which may be more accentuated in settings like Singapore where medical care is fee for service. The focus on health was also reflected in shifts in personal identity, internalization of self as dialysis patients, and distancing from pre-ESRD identity. They also devalued desired items so as to better align behavior to treatment guidelines, as similarly reported in heart failure patients (Piamjariyakul et al., 2012).
Avoidance strategies were also commonly employed. These involved thought suppression, in line with findings with obese populations (Barnes et al., 2011; Barnes and Tantleff-Dunn, 2010), or range of actions aimed at minimizing exposure, such as not buying craved items for household, avoiding places where such items are readily available, or by asking family and friends to refrain from consuming such items in front of them.
As avoidance may not be always feasible or preferred, participants also reported engaging in purposeful yet controlled lapses. Key to this was moderation through tasting and savoring fluid or foods in small quantities. This “controlled and paced giving in” approach allows patients the satisfaction of indulging while exercising restraint. Several of these techniques, namely, ice cubes or cold water, have been described in Mistiaen (2001)’s review. Thirst-quenching substitutes such as consuming sour plums and/or lozenges also help participants to relieve their thirst.
Besides strategies employed in response to food craving or thirst, participants shared more general self-regulatory strategies used to manage diet and fluid intake. These involved both pre-emptive strategies, such as self-monitoring to prevent overconsumption as well as compensatory actions to counter/rectify lapses.
Self-monitoring of intake/consumption was ingrained in the daily lives of patients. Self-monitoring was deployed mainly as a preventative measure to ensure that patients stayed in line with dietary and fluid-intake recommendations. This involved measuring consumption by using containers, preparing daily menu, or weighing themselves to track behavior and take action so as to avoid excessive consumption. Although only used by handful of patients, it is important to note that such self-monitoring strategies appeared to have evolved into routinized set of practices that helped patients to both remember and enact self-management behaviors by making day-to-day decisions around food and water intake easy and effortless.
Lapses, however, in the context of any chronic illness are unavoidable; hence, patients’ ability to be able to recognize these and engage in appropriate self-regulatory actions to bounce back onto right track are paramount for effective self-regulation. As such, many patients spontaneously shared their compensatory behaviors. These were in the main reactive responses to address dietary/fluid-intake lapses and involved restricting their subsequent intake of food and fluid, prolonging dialysis duration, and/or exercising so as to get rid of the excess fluid. A similar parallel can be drawn between these behaviors and those of diabetic patients compensating for their high blood sugar levels (Collins et al., 2009; Malpass et al., 2009).
Whereas in the context of dieting, compensatory beliefs and intentions have been shown to be associated with non-adherence (Miquelon et al., 2012) and increased calorie intake (Kronick et al., 2011), the ability to combine pre-emptive with compensatory strategies may be a more realistic goal for patients with long-term conditions. There is robust evidence for the value of relapse prevention approach in diet/weight reduction programs (Perri et al., 1993; Turk et al., 2009) but less so in the context of chronic disease management. Study findings indicate that such strategies are the norm rather than the exception, underscoring the need for more work on the implications in terms of clinical management for ESRD patients, and the “safe” margins for such compensatory actions. More attention should also be devoted to their role in patient education and training. Our findings would suggest a need for education or self-management programs traditionally structured around pre-emptive behaviors to include some training in compensatory skills.
Overall, study findings revealed that patients combine a range of cognitive and behavioral strategies either proactively to avoid lapses or compensate when slips occur. While many of these strategies share many similarities among patients with other chronic illnesses, some of the strategies, such as self-perception modification, paced and controlled “giving in,” and consumption of thirst-quenching substitutes, are unique to the ESRD population.
Contrary to what we anticipated, we did not find any culture-specific themes and strategies. Participants discussed the struggle to deal with the hot weather in Singapore, but issues related to social sharing of food or social pressure were not endorsed. It is likely that paced controlled consumption and compensation—the most commonly used strategies—alleviate such pressures as patients behave in line with cultural norms while exercising some restraint or compensation. Nevertheless, as the study was based on patients from an Asian background, more work is needed to explore the generalizability in other contexts. Another important direction for future research pertains to the effectiveness of the different strategies in terms of clinical management and behavioral endpoints such as biochemical markers or self-report adherence.
This is the first study to explore how patients on hemodialysis manage both food cravings and thirst, and serves as an important first step toward understanding their experience and needs in order to guide interventions and programs of support. This study has a few notable strengths and weaknesses. First, this was a cross-sectional investigation; hence, we were not able to capture how strategies may evolve or adapt over time. Second, due to logistical constraints, we have excluded study participants (from the parent study) who could not speak either English or Mandarin, as such patients who spoke only Malay, Tamil, or other Chinese dialects were not represented. Despite this, we have managed to obtain a robust sample that is representative of the general hemodialysis population in terms of ethnicity, gender ratio, and age (Singapore Renal Registry, 2010). Third, since participants consisted of both control and intervention patients from the parent RCT, it is possible that participation in the intervention may have an impact on their responses. Nevertheless, it is unlikely that the intervention participants would differ significantly from controls in their responses associated with food cravings and thirst, since the intervention’s focus on food cravings and thirst was minimal. Strengths of this work include an excellent response rate and the considerably large number of interviews which provided the study with rich material.
Footnotes
Acknowledgements
The authors thank Dr Tonia Griva, the NKF Health Care Professionals, and patients for their support in this study.
Funding
This research was supported by a grant from National Kidney Foundation Research Fund (NKFRC2008/07/24) and Ministry of Education—National University of Singapore (NUS) Academic Research Fund (start-up) (FY2007-FRC5-006)—which are gratefully acknowledged.