Prevalence of chronic pain and chronic widespread pain among subjects with heart failure in the general population: The HUNT study

Pain in chronic heart failure (HF) is a significant but often unrecognized symptom. Characteristics of pain in subjects with HF are largely undescribed. The study aimed to address a knowledge gap in the relationship between HF and pain by investigating the prevalence of chronic pain and chronic widespread pain (CWP) among subjects with HF.


| INTRODUCTION
Chronic heart failure (HF) is a clinical syndrome consisting of cardinal symptoms (e.g., breathlessness, ankle swelling, and fatigue) that may be accompanied by signs (e.g., elevated jugular venous pressure, pulmonary crackles, and peripheral oedema).It is due to a structural and/or functional abnormality of the heart that results in elevated intracardiac pressures and/or inadequate cardiac output at rest and/or during exercise (McDonagh et al., 2021).
HF affects 1%-2% of the adult population in developed countries, and worldwide, more than 26 million are affected (McDonagh et al., 2021).Due to the ageing of the population, the prevalence is increasing (Jones et al., 2018).One of the most frequently reported symptoms among patients with HF is pain (Alpert et al., 2017), yet the focus on pain is limited, both in research and clinical practice (Tesarz et al., 2019), where symptoms such as fatigue, dyspnoea, oedema and fluid retention are emphasized (Jurgens et al., 2022;McDonald et al., 2015;Yancy et al., 2017).
HF is the end stage of most cardiovascular disorders, and it is associated with significant morbidity and mortality (Goebel et al., 2009;Mamas et al., 2017;McIlvennan & Allen, 2016;Murray et al., 2005).Pain is one of the symptoms that contributes to the high symptom burden in HF.However, pain is often undertreated and not well understood, representing a knowledge gap and an opportunity for improvement in symptom control (Chen et al., 2019).
The relationship between HF and pain is rarely investigated in epidemiological studies, and studies of chronic pain in combination with a cardiovascular diagnosis like HF are scarce (Alpert et al., 2017;Blyth, 2008;Croft et al., 2010;Groenewegen et al., 2020).Further, methods for classifying chronic pain in population studies are highly variable, and prevalence estimates can differ from 11% to 64% (Steingrimsdottir et al., 2017).
The reported prevalence of pain in HF subjects varies significantly, from 23.1% to 85% (Mhesin et al., 2022;Rustoen et al., 2008) and the prevalence increases as the patients' functional status worsens (Johansson et al., 2006).It is still not established if the pain is related to HF itself, high age, or the burden of comorbidity.Studies suggest that HF shares several clinical features with fibromyalgia, a syndrome characterized by chronic widespread pain (CWP), often attributed to increased central sensitivity (Gist et al., 2017).Central sensitization is a complex phenomenon with amplified signalling and enhanced sensitivity of nociceptive neurons within the central nervous system.It is clinically manifested by allodynia, hyperalgesia, and referred pain across multiple segments (Valera-Calero et al., 2022;Woolf, 2011).However, studies of CWP and HF are almost non-existent.
This study aimed to address the knowledge gap in the relationship between HF and pain by investigating the prevalence of chronic pain and CWP among subjects with HF.We hypothesized that the prevalence of both chronic pain and CWP would be higher among subjects with HF than in the general population and higher among subjects with HF than among subjects with other cardiovascular diseases.
In HUNT3, 93,860 people qualified for participation, and 50,802 (54.1%) participated.All residents aged 20 years and older in the region of Nord-Trøndelag, Norway, were invited to participate, and a large and comprehensive study included interviews, self-administrated questionnaires, and clinical examinations (Krokstad et al., 2012).Those who completed the first baseline questionnaire also received the second baseline questionnaire.Those who completed both the first and second baseline questionnaires and qualified for the cardiovascular disease questionnaire (i.e., reported cardiovascular and/or renal disease in the first baseline questionnaire) received this as a final questionnaire; see Figure 1.This questionnaire included the following categories: hypertension, renal disease, cardiac diseases (i.e., myocardial infarction, angina pectoris, atrial fibrillation, heart failure), and stroke.
For comparison, the participants were divided into three groups.Those answering the first and second baseline questionnaires, represent the population in general (n = 41,198).Those with self-reported cardiovascular disease (those who completed the first and second baseline questionnaires and reported cardiovascular and/or renal disease in the first baseline questionnaire) representing the general population with cardiovascular diseases without HF (n = 10,808).In addition, those with self-reported HF (participants answering the cardiovascular disease questionnaire and reporting HF) (n = 696).These participants answered yes to the question 'Has a doctor said that you have heart failure (weak heart muscle, water in the lungs, swollen legs)?' and were defined as having HF.
The HUNT3 study and present project were approved by the Regional Norwegian Committee for Medical and Health Research Ethics Mid Norway (references 4.2006.250 and 2018/2142).All participants signed written informed consent, and the study adhered to the General Data Protection Regulation (GDPR).

| Variables and the definition of pain
Four questions were used to describe and determine chronic pain, CWP, and pain intensity: (1) 'In the last year, have you had pain or stiffness in muscles or joints that has lasted at least 3 consecutive months?' (Response: yes-no), (2) If yes, 'Where have you had this pain or stiffness?' (Response: neck, shoulders, upper part of the back, elbows, lumbar regions, wrists/hands, hips, knees, ankles/ feet), (3) If yes, 'Have you had this pain/stiffness on both the right and left side of your body?' (Response: yes or no) and (4) 'How much bodily pain have you had during the last 4 weeks?' (Response: No pain; very mild; mild; moderate; severe; very severe).
Chronic musculoskeletal pain information and information on pain localization were obtained based on the Standardized Nordic Questionnaire (SNQ) (Kuorinka et al., 1987).Information on pain intensity was measured by one question from the Medical Outcomes Study Short Form health survey (SF-36) (Ware et al., 2001).The pain intensity question has been validated as a single item as part of the SF-8 health survey (Lang et al., 2018) and divided into two categories with a cutoff between no to mild and moderate to high pain, treated as a dichotomous variable: low pain intensity and moderate to high pain intensity.This has been shown to adequately identify subjects with complex and long-lasting pain in population samples (Landmark et al., 2012).
Chronic pain was defined as pain that persisted or recurred for more than 3 months (Treede et al., 2019), and CWP was defined in line with the definition accounted for in 1990 by the American College of Rheumatology (ACR) and further by (Butler et al. 2016).Hence, CWP was defined as having chronic pain, bilateral pain, and pain in all three major body areas: upper limbs (i.e., at least one of the following: shoulders, elbow, and hand), lower limbs (i.e., at least one of the following: ankles, knees, and hips) and trunk (i.e., at last one of the following: neck, upper back, and lumbar).Answering yes to all five criteria was needed to qualify for CWP.
Potential confounders included in this study were age, sex, body mass index (BMI), and comorbid status.These variables were chosen based on results from previous studies on pain in HF and chronic pain in general (Croft et al., 2010;Goodlin et al., 2012).BMI was divided into three categories: underweight (BMI < 18.5), normal/overweight (BMI 18.5-30.0),and obese (BMI > 30.0) (Åsvold et al., 2023).Comorbid status was categorized into no diseases, one disease, two diseases, and three or more diseases.Diseases included in comorbid status were all variables available for the study participants (from the first baseline questionnaire): stroke or brain haemorrhage, kidney disease, asthma, chronic bronchitis, emphysema, or chronic obstructive pulmonary disease (COPD), diabetes, psoriasis, cancer, and epilepsy.Age, BMI, and comorbidity were given as continuous variables used in regression analysis, but divided in categories for sample characteristics.

| Statistical analysis
Descriptive statistics of frequencies, percentages, means, and standard deviations (SD) were used to characterize the study population.The prevalence of CWP and chronic pain with their respective confidence intervals (CI) were estimated for the population with HF, cardiovascular disease without HF, and the general population separately.Two separate logistic regression analyses were conducted with the aim of investigating the association between HF and CWP and HF and moderate to high pain intensity among the participants with cardiovascular diseases, respectively.The models included age, sex, BMI and comorbidity as potential confounders.The predictors were presented with their respective odds ratios (OR) and 95% CI.The reference groups used in the logistic models were age 20-49 years, woman's sex, BMI normal or overweight, no comorbidity, and no HF.Results were considered statistically significant if p < 0.05.Results are presented from both unadjusted and adjusted models.Analyses were performed using the IBM SPSS 25 statistical package for Macintosh (released 2017; IBM Corp.).a higher BMI (mean 29.8, SD 5.2), and had more comorbidities (mean 1.05, SD 5.2) compared to the participants with cardiovascular diseases without HF and the baseline participants.
Figure 2 illustrates the prevalence of pain at each pain location site.Overall, the distribution of pain locations was similar among the three groups.For every location, a higher prevalence of pain was seen among the HF-population, most markedly for the lower back and the lower extremities, that is hips, knees, and ankle and feet.

| Association between heart failure and chronic widespread pain
Logistic regression analyses were also used to investigate the association between HF and CWP among subjects with self-reported cardiovascular disease; see Table 2.In the unadjusted model, the OR for CWP was 1.61 (CI 1.33-1.96)among those with HF.In the adjusted analysis, OR was 1.62 (CI 1.30-2.01),p < 0.001 Hence, it could not be explained by the other variables in the model.

| Association between heart failure and moderate to high pain intensity
Logistic regression was also used to investigate the association between HF and moderate to high pain intensity among subjects with self-reported cardiovascular disease.The OR for moderate to high pain intensity was 1.73 (CI 1.47-2.05)among those having HF in the unadjusted model and 1.54 (CI 1.29-1.85) in the adjusted model, p < 0.001, when compared to the reference group without HF, presented in Table 3.

| Descriptive analysis of missing data
Statistical analysis revealed that there was a noticeable amount of missing data in the diagnostic HF question from Cardiovascular Disease Questionnaire (14.3%, n = 1645).This missing group is descriptively described in Table S1.Among the group of participants missing the HF question, 59.2% reported chronic pain, 12.5% reported CWP and 54.6% reported moderate to high pain intensity (see Table S1).For comparison, the amount of information missing in other, comparable diagnostic questions is presented in Table S2.

| DISCUSSION
In this study, we found a high prevalence of pain in the HF population when measuring chronic pain, CWP, and pain intensity.To summarize, associations between HF, CWP, and moderate to high pain intensity could not be explained by the sociodemographic factors age, sex, BMI, or comorbidity.These results contribute to an increased understanding of pain in HF.Therefore, assessment of pain should not be restricted to cardiac-related pain, but chronic pain in general, as widespread pain adds to the symptom burden.

| Prevalence of pain in heart failure
The prevalence of chronic pain among subjects with HF was 67.8%, the prevalence of CWP was 20.7%, and the prevalence of moderate to high pain intensity was 58.8%.
In previous studies, the reported prevalence of pain in HF has varied significantly (Mhesin et al., 2022;Rustoen et al., 2008).Our results confirm the high prevalence in previous studies (Conley et al., 2015;Godfrey et al., 2006;Goodlin et al., 2008Goodlin et al., , 2012).In the general population, the prevalence of chronic pain was 51.2%, the prevalence of CWP was 10.3%, and the prevalence of moderate to high pain intensity was 38.8%.This was lower than among HF subjects, but on the upper end compared to the previously estimated prevalence of chronic pain.Also here, the estimates fall within the upper range of what has been reported in previous studies (Steingrimsdottir et al., 2017).A total of 15.5% of the participants in the cardiovascular disease study reported HF.The HF-population was older and had higher comorbidity than both the population in general and the other participants with cardiovascular disease.Previous studies have shown that HF is more prevalent in older people with serious comorbidities (Johansson et al., 2004;McIlvennan & Allen, 2016).Age and comorbidity have often been assumed to explain the high prevalence of pain in HF.However, results from our study showed that age and comorbidity alone did not explain the high odds of CWP or moderate to high pain intensity.This is an important finding, as pain is often undertreated.Our findings also point to pain as being more relevant in HF when compared to other cardiovascular diseases.
The most common pain sites among all groups were the lower back, shoulders, hips, and knees.However, the HF population reported a higher prevalence of pain in all nine sites of the body diagram.High prevalence of pain in the lower extremities has also been reported by Goodlin et al. (2012), where the most common site of pain was the legs below the knees (32.3%).Chen et al. (2021) found prevalent back pain to be associated with an increased level of fatigue, depressive symptoms, and decreased physical function among older adults with HF, illustrating the impact of noncardiac pain on the symptom burden.The primary cause of HF might be important for the location and other features of pain.(Clark & Goode, 2013), investigated the prevalence of chest pain as a possible cause of chronic pain among 1786 patients with HF.They found that it was unlikely that the pain was due to angina, even with underlying coronary heart disease.However, other studies have found chest pain to be more prevalent in HF (Alemzadeh-Ansari et al., 2017).Several studies have found that the number of pain sites is a strong prognostic factor on its own for the further course of disease, cardiovascular disease included (Andersson, 2004;Bergman et al., 2002;Larsson et al., 2012;Staud, 2005).Despite this, there is sparse knowledge of the consequences of CWP on physical health and mortality.Research has suggested that disabling pain, which may be relevant for many subjects with HF, leads to a lifestyle characterized by factors associated with increased mortality risk and comprising self-care, such as lower levels of physical activity, sleeping problems, and unfavourable nutritional habits (Chen et al., 2021;Fayaz et al., 2016;Vandenkerkhof et al., 2011).Further, pain may negatively affect medication adherence, an important clinical issue highlighted in HF-guidelines (McDonagh et al., 2021).Research has also indicated a possible shared genetic basis between cardiovascular disease and pain (van Hecke et al., 2017;Winsvold et al., 2017).

| Chronic widespread pain and heart failure
The odds of CWP were more than twice as high among the HF group compared to the population in general.This study extends the literature, as CWP affects one in 10 individuals globally within the general population (Andrews et al., 2018;Tesarz et al., 2019).The prevalence of CWP in the general population in this study was equal to previous studies, indicating a correct choice of definition.Several similarities exist between subjects with CWP and those with HF.Female sex, depression and anxiety, smoking, sleeping problems, overweight or obesity, and a family history of pain are known risk factors for CWP (Bergman et al., 2002;Kvalheim et al., 2013;Mundal et al., 2014).A high prevalence of sleeping problems, overweight, and depression also characterize patients with HF (Alpert et al., 2017;Gullvåg et al., 2019).Typically, subjects with advanced HF have marked limitations of physical activity (Woolf, 2011) and also subjects with CWP experience difficulties with physical activity (Sylwander et al., 2020).This suggests that it might be useful to consider some of the recommendations for treatment of CWP for patients with HF as well.For example, non-pharmacological treatment such as physical activity and exercise is recommended and found effective in CWP (Mannerkorpi et al., 2009).Studies suggest that patients with advanced HF can also benefit from physical activity, including high-intensity training (Kobat & karatas, 2021;Rognmo et al., 2012).Treating patients with HF and pain is a clinical challenge given the cardiovascular risks associated with common pain medications like NSAIDs (Cavanagh et al., 2021).Therefore, understanding pain and pain treatment, including both non-pharmacological and pharmacological options, is essential to reducing symptom burden in HF patients.
This study examined the relationship between HF and CWP beyond previous studies.The causes of pain can be many, are often complex, and include physiological, sociocultural, sensory, affective, cognitive, and behavioural components (Goebel et al., 2009;McDonald et al., 2015).The causes may be associated with multiple negative consequences, including depression and lower physical activity, both of which increase symptom burden and reduce quality of life (McDonald et al., 2015).However, one previous study reported that widespread pain was a risk factor for cardiovascular mortality (Tesarz et al., 2019).Another study investigated the several clinical features between HF and fibromyalgia and reported a prevalence of 22.8% of the studied cohort fulfilling both fibromyalgia diagnosis criteria and HF (Gist et al., 2017).This underscores the relationship between CWP and HF as relevant.Providing a new angle to understand pain in HF.The number of studies on HF and CWP is still scarce.Further studies are needed, emphasizing the clinical impact and possible interventions.
The pain among HF subjects could not be explained by either age, sex, BMI, or comorbidity.This should raise our awareness of chronic pain in HF.Which, despite its high prevalence, often goes unrecognized by health care providers and is underreported by patients (Godfrey et al., 2007).Further, pain assessment and treatment are only briefly mentioned in clinical guidelines (McDonagh et al., 2021).A total of 58.8% of the HF population reported moderate to high-pain intensity during the last 4 weeks, and 20.7% reported CWP.This underpins that pain is present and needs attention and adequate management both in the clinic and in further studies.

| Strengths and limitations
This study was cross-sectional and based on HUNT3 data.The HUNT surveys cover a broad range of health-related topics (Åsvold et al., 2023).There were 41,198 subjects included in this study, or approximately 81.1% of the total number of subjects invited, giving a large and sufficient sample size.The study population's demographic and clinical characteristics are quite like those of the population in general, hence being representative of an approximation of the Norwegian population (Krokstad et al., 2013).The use of well-established, reliable questionnaires, SNQ and SF-36, is also considered a strength of the study.Clinical examinations were not performed, and a diagnosis of pain was not However, self-reported pain is in line with the definition of pain by the International Association for the Study of Pain (IASP) (Raja et al., 2020).
The main limitation of the study is that subjects with HF were selected based on a self-reported diagnosis.Hence, the HF population could be prone to misclassification.Selfreport is likely to underestimate the true HF prevalence, especially in subgroups with low health literacy (Rethy et al., 2020).Camplain et al. (2017)) found that the sensitivity of self-report was low and specificity was high.However, the HF prevalence of 1.7% (696/41198) corresponds well to the prevalence of 1%-2% reported before.In the National Health and Nutrition Examination Survey (NHANES), selfreported physician diagnosis of HF were used to estimate the prevalence of HF in the United States.This was the same question as in the HUNT study (Rethy et al., 2020).
For the HF diagnosis question, the proportion of missing responses was 14.3%.To explore possible reasons for and consequences of missing, descriptive analysis was conducted.The prevalence of pain and CWP was higher among subjects missing on the HF diagnosis question, meaning that the prevalence of pain was most likely not overestimated due to missing data.
Pain was also based on self-report, which is considered standard as pain is a subjective experience (Dworkin et al., 2005;Steingrimsdottir et al., 2017).However, questions are subject to interpretation, and even small changes in wording may have a noticeable impact on the estimates (Landmark et al., 2012;Steingrimsdottir et al., 2017).We used questions from two different instruments, one asking about chronic pain or stiffness in muscles and joints (Kuorinka et al., 1987), and the other about bodily pain (Ware et al., 2001).As the phrasing in these two measures deviates slightly and they have not previously been combined, we chose to handle them separately in the analyses.However, using a cut of at moderate pain or more in the SF-36 bodily pain question have shown to identify most subjects with chronic pain (Landmark et al., 2012).Moreover, in our study, the prevalence of chronic pain was higher among subjects with HF than in the general population.Thus, the analyses conducted on moderate pain or more, to a large degree reflects the association between HF and chronic pain of moderate intensity or more.
We used the 1990 ACR criteria to define CWP (Wolfe et al., 1990).These have been criticized for being ambiguous, and different interpretations may give widely differing prevalence estimates (Butler et al., 2016).More recent criteria such as those proposed by ACR in 2010 and revised in 2016, and those proposed by CWP in the 11th version of the International Classification, ICD-11, may in the future become less of a problem by agreeing on one standard definition.However, many studies still use the ACR 1990 criteria, and as in our study, this may be due to the measurements being used.However, we used a strict approach to the ACR 1990 criteria with the intention of increasing the reliability.This is particularly relevant to the HF population, as this group may have pain of diverse origins.It is also likely to increase comparability with the ICD-11 definition and the 2010 /2016 ACR criteria (Landmark et al., 2019).

| CONCLUSION
This population-based study revealed a high prevalence of both chronic pain and CWP among subjects with HF compared with the population in general and with other cardiovascular diseases.The high prevalence of both moderate to high pain intensity and CWP in HF patients could not be explained by higher age, BMI, comorbidity, or sex distribution.Measuring CWP gave useful information on pain among subjects with HF, and these findings underpin the need to identify and assess pain among individuals with HF and tailor their pain management.Assessment of pain should not be restricted to cardiac-related pain but to chronic pain in general, as widespread pain adds to the symptom burden.This is relevant and important both for clinical and research purposes.
with a cardiovascular disease Participants with heart failure

.1 | Sample characteristics and pain prevalence
Demographic data and baseline characteristics categorized by groups are presented in Table1.The participants with HF were older (mean age 72.4 years, SD 11.2), had Sample characteristics of the study population.
T A B L E 1Abbreviations: BMI, body mass index; CWP, chronic widespread pain; N, number of participants; SD, standard deviation.aInall three groups for comparison there are some missing in terms of BMI and comorbid status.bThesum of moderate to high and low-pain intensity does not sum up to 100%, due to missing data in the questionnaire.c Low pain intensity (no pain-very mild-mild).d Moderate to high pain intensity (moderate-severe-very severe).

Adjusted Total, n Prevalence a moderate to high pain intensity (%) OR [95% CI] p-value OR [95% CI] p-value
Logistic regression for pain intensity.
T A B L E 3Abbreviations: BMI, body mass index; CI, confidence interval; CWP, chronic widespread pain; N, Number of patients; OR, odds ratio.aNumber of participants in statistical analysis.This number may deviate from Table1because of missing data.Those numbered (N) in this table are all registered with low or moderate to high-pain intensity in the pain intensity-question.