Hypertension prevalence in patients attending tertiary pain management services, a registry-based Australian cohort study.

Persistent pain and hypertension often co-occur, and share several biological and lifestyle risk factors. The present study aimed to provide insight into the prevalence of, and factors associated with, hypertension in the largest cohort of patients seeking treatment in 43 tertiary pain clinics in Australia. Adults aged > = 18 years registered to the electronic Persistent Pain Outcomes Collaboration registry between 2013 and 2018 were included if they had persistent non-cancer pain (N = 43,789). Risk Ratios (RRs) compared prevalence of self-reported hypertension with the general and primary care Australian populations, and logistic regression examined factors associated with hypertension. One in four (23.9%) patients had hypertension, which was higher than the Australian adult population (2014-15: RR = 5.86, 95%CI: 5.66, 6.06; 2017-18: RR = 9.40, 95%CI: 9.01, 9.80), and in primary care patients (2011-13: RR = 1.17, 95%CI: 1.15, 1.20). Adjusting for covariates, patients with higher odds of hypertension were older, lived in regions with higher socioeconomic disadvantage, had higher levels of BMI, were born outside the Oceania/Australasia region, and had comorbid arthritis, diabetes, or severe-extremely severe anxiety symptoms. Female patients and those with depression symptoms had lower adjusted odds. Unadjusted analyses showed an association between widespread pain, pain duration, pain severity and interference, and lower pain self-efficacy with hypertension; however, only pain severity remained significant in adjusted analyses. Hypertension was more prevalent in people with persistent pain than in the general community, was associated with more severe pain, and commonly co-occurred with pain-related impairments. Routine hypertension screening and treatment targeting shared mechanisms of hypertension and pain may improve treatment outcomes in the pain clinic setting.

Introduction

Several studies have shown higher prevalence of hypertension (i.e., systolic blood pressure >140mmHg) in people with persistent pain [1, 2]. People with persistent pain typically have higher blood pressure both at rest and when experiencing a painful stimulus (e.g., in the cold pressor test) [1] and prevalence is higher in people with more severe persistent pain [3].

A number of factors have been found to increase the risk of developing both persistent pain and other chronic diseases [4], especially cardiovascular diseases [1, 5–8]. These include psychological factors, such as depression, anxiety and stress), lifestyle factors (e.g., obesity, low physical inactivity, deconditioning) and social factors (e.g., isolation, unemployment, lower education) [9, 10]. Moreover, biological factors probably play a role given that people with persistent pain have lower baroreceptor sensitivity, and diminished inhibitory engagement of the parasympathetic nervous system at rest [11-13]. Some factors such as smoking, low levels of physical activity and obesity probably play a causal role in blood pressure and cardiovascular disease development, possibly independent of pain severity [14], due to their impact on vital homeostatic processes and organ function. Other factors like anxiety, however, probably have an indirect association via disruption in stress regulation systems [15] and decreased parasympathetic tone [13].

Understanding the prevalence of hypertension, and demographic, clinical and pain-related features associated with hypertension in patients with persistent pain will provide an important step in the future development and implementation of tailored interventions for patients seeking treatment for pain in outpatient pain management settings. While the co-occurrence of persistent pain and hypertension has been examined in several community [1, 2] and primary care cohorts [16], few studies have focused on patients with pain that is severe or disabling enough to warrant treatment in a multidisciplinary pain management service. In fact, only one study published nearly 15 years ago has examined the prevalence and clinical manifestation of hypertension in a small sample of 300 patients attending a tertiary pain management service in the the United States of America, which found that 39% of patients had hypertension [8]. The present study therefore aimed to provide an updated robust insight into the prevalence of hypertension in a large cohort of patients referred to outpatient pain management clinics in Australia, and to investigate the demographic, health and pain-related characteristics associated with having hypertension.

Methods

Setting

In Australia, pain management services are specialist sub-acute ambulatory care services located primarily in public and private hospital settings. Most clinics comprise medical staff (pain specialists, psychiatry, anaesthetists, rehabilitation medicine consultants, and general practitioners) and senior allied health and nursing staff (physiotherapists, occupational therapists, psychologists, and clinical and research nurses). In 2013, the Australian Health Services Research Institute, University of Wollongong, established the electronic Persistent Pain Outcomes Collaboration (ePPOC) to facilitate outcome assessment and service benchmarking. The design, procedures and characteristics of ePPOC are fully described elsewhere [17]. Under ePPOC, patients referred to specialist outpatient pain management clinics complete a standard battery of outcome measures at referral or prior to commencing treatment, and at set follow-up intervals according to a defined protocol.

Participants

All new patients entered in ePPOC up to September 2018, aged 18 years and over, who were seeking treatment for non-malignant persistent pain and responded to the primary outcome measure (patient reported hypertension status) were included. Data obtained in the initial patient questionnaire completed at referral or commencement of treatment were used for this study. The dataset was screened for multiple referrals for the same patient, and only data from the most recent referral were extracted for analysis.

Materials and procedures

Questionnaires were completed by patients and entered into the purpose-built software (epiCentre) by staff at the respective pain management clinic. The study had low risk approval from the Alfred Health Human Research Ethics Committee for the analysis of fully deidentified data. Participants contribute their data as part of their clinical episode, and while they do not provide written consent for the use of the data in individual studies they are informed that their data will be used for service benchmarking, and that deidentified data may be used for research.

Demographics

Patient age, sex, and country of birth were used to characterise individual-level demographics. Country of birth was classified into regions according to the Standard Australian Classification of Countries [18], and summarised as Oceania and Antarctica, North-West Europe, Southern and Eastern Europe, North Africa and the Middle East, South-East Asia, North-East Asia, Southern and Central Asia, Americas, and Sub-Saharan Africa. Area level socioeconomic position was determined from patient residential postcode using the Index of Relative Socio-Economic Advantage and Disadvantage (IRSAD) decile [19], which were summarised into quintiles with lower scores indicating higher disadvantage. The IRSAD rankings are determined from typical education, employment and family structure in all Australian postal codes from the National Census of Population and Housing in 2011, and each area is ranked nationally.

Health and comorbidities

Body Mass Index (BMI) was calculated from patient height (cm) and weight (kg), and classified into World Health Organization [20] ranges (kg/m2) of underweight (<18.5), normal weight (18.5 to <25), overweight (25 to <30), Obese Class I (30 to <35), Obese Class II (35 to <40) and Obese Class III (>40). Medical conditions and comorbidities were recorded in response to the question “do you have any of the following medical conditions?” which was followed by a list of conditions that included high blood pressure, arthritis (rheumatoid or osteoarthritis), and diabetes. The self-reported data are broadly consistent with population-based registry procedures for assessing the presence of comorbid conditions. Objective measures of blood pressure, or prescription antihypertensives were not available.

Pain

The Brief Pain Inventory (BPI) was used to measure pain severity and interference in the previous week [21]. Participants completed 11-point numerical rating scales for the intensity of pain (right now, on average, at worst and least; 0 “no pain”, 10 “Pain as bad as you can imagine”), and its impact on daily functions, including general activity, mood, walking ability, normal work, relationships with other people, sleep, and enjoyment of life (0 “does not interfere”, 10 “completely interferes”). Severity and interference scores were calculated according to the scoring guidelines [22], which showed high levels of internal consistency in this sample (Severity: Cronbach α = 0.88; Interference: Cronbach α = 0.89). Pain severity and interference were classified as low (0 to 3), moderate (4 to 6) and high (7 to 10) [23, 24].

Participants reported which body regions were affected by pain on a body map [25]. Widespread pain was classified according to the Widespread Pain Index [26], which involves summing the number of sites of pain across the neck, upper back, lower back, chest, abdomen, and the left and right side of the: jaw, shoulder girdle, upper arm, lower arm, hip/buttock/trochanter, upper leg and lower leg. As the ePPOC body map coded for pain in the head, but not in the left/right cheek, the total possible score for the WPI was 18 instead of 19. Participants were classified as having widespread pain if they had a score > = 7, based on recommendations of the 2010 American College of Rheumatology criteria [27].

The Pain Self-Efficacy Questionnaire (PSEQ)[28] is a self-report measure of confidence in performing everyday tasks, despite being in pain. It comprises 10 items such as “I can enjoy things, despite the pain” that are rated on a scale from 0 to 6, with higher scores indicating greater confidence. Ratings are summed to produce a total score ranging from 0 to 60 with higher scores indicated better self-efficacy. Scores were classified as indicating severe impairment (<20), moderate impairment (20–30), mild impairment (31–40) and low impairment (>40) [28]. Cronbach’s α for this sample was 0.93.

The Pain Catastrophizing Scale (PCS) [29] is a self-report measure of catastrophic thoughts and feelings (e.g., rumination, magnification or helplessness) about pain, which comprises 13 items that are rated on a 5-point Likert scale. The PCS total score ranges from 0 to 52, and scores can be classified as clinically normal (0–19), high (20–29) and clinically severe (> = 30) [30]. Cronbach’s α for the total scale in this sample was 0.95.

Mood

The 21-item version of the Depression Anxiety Stress Scale (DASS-21) was used to characterise depression and anxiety symptom severity [31]. The DASS has been shown to validly measure the dimensions of depression (Cronbach α = 0.93, in this sample), and anxiety (Cronbach α = 0.86, in this sample) in large non-clinical samples [32], and in patients with persistent pain [33, 34]. Each subscale score was doubled to enable use of the DASS-42 severity classifications for depression: normal (scores: 0–9), mild (scores 10–13), moderate (scores 14–20), severe (21–27) and very severe (> = 28); and anxiety: normal (scores: 0–7), mild (scores 8–9), moderate (scores 10–14), severe (15–19) and very severe (> = 20).

Hypertension prevalence data sources

The prevalence rate of self-reported hypertension in the ePPOC cohort was compared with the Australian population (aged > = 18 years), and an Australian primary care population (aged > = 24 years). The prevalence in the Australian population was taken from the 2014 and 2017 Australian Bureau of Statistics (ABS) National Health Survey (NHS) data [35, 36]. The NHS is a purposely sampled survey intended to be representative of persons living in the general community in Australia; however, people living in institutions or in remote regions are not sampled. The NHSs recorded the prevalence of long-term health conditions (i.e., a condition that had lasted, or will last, for at least 6-months) in 2011–12, and 2017–18, respectively. We compared self-reported hypertension in the present ePPOC cohort with the prevalence of self-reported hypertension from the NHS interview. The NHS interview recorded hypertension status in response to the questions “Have you ever been told by a doctor or nurse that you have … condition?” and “Is this condition current and/or long term (>6 months)?”. The prevalence of hypertension in Australian primary care patients was taken from a sentinel study by Ghosh, Charlton [37], which examined the prevalence of chronic conditions diagnosed and recorded by a general practitioner from all patient interactions over a two year period (September 2011 to September 2013) from 17 general practices in a single health district in New South Wales, Australia (N = 103,917). The crude, and age-adjusted prevalence in the ePPOC cohort were compared with the crude and age-adjusted prevalence of hypertension in the study by Ghosh, Charlton [37].

Statistical analysis

The data were analysed using Stata Version 14.0. Significance was determined when the confidence interval (CI) for risk ratios (RRs), or odds ratio (ORs), did not include 1.00. BPI severity and interference subscales were calculated if patients completed all four severity ratings, and a minimum of four interference items. Subscale scores were calculated for responses given to items on the DASS anxiety subscale, DASS depression subscale, PSEQ or PCS; however, participants missing more than one item were coded as missing for that measure consistent with previous recommendations [38]. Responses to a single missing item were not imputed to generate summary scores. BMI was calculated after applying sex-stratified plausible weight and height values based on criteria published from a previous Australian epidemiological project, the Household, Income and Labour Dynamics in Australia (HILDA) study(height: men = 130 to 229cm; women = 110 to 210cm; weight: men = 35 to 300kg; women = 25 to 300kg) such that height and weight falling outside of those ranges were coded as missing due to a probable data entry error [39].

The RR and corresponding 95% CIs were calculated to determine the difference in risk of hypertension in the ePPOC cohort compared with the general population, and the primary care population.

The association between demographic, health and pain-related characteristics and hypertension were examined using logistic regression. The ORs and 95% CIs were calculated, including ORs without adjusting for covariates, and ORs covarying for all demographic, clinical and pain characteristics. The fully adjusted model estimated missing data from the covariates using multiple imputation by chained equations, which imputes one variable at a time, conditional on the other variables included in the multivariable model, through multiple iterations [40]. Twenty imputed datasets were produced and combined using Rubin’s rules [41]. Factors were retained in multivariable analyses if they had an unadjusted p-value of <0.20 in accordance with previous recommendations [42]. Factors were considered based on prior clinical studies examining factors associated with hypertension, and included age, sex, birth region, socioeconomic disadvantage (IRSAD quintiles), BMI, arthritis, diabetes, anxiety symptom severity, depression symptom severity, pain source, pain duration, presence of widespread pain, pain severity, pain interference pain self-efficacy and pain catastrophizing. Multicollinearity was assessed, and no issues were identified, with an overall Variance Inflation Factor of <2.50. Sensitivity tests were performed to examine whether the results from the logistic regression were robust. For the first sensitivity test, the cohort was randomly split into two groups (using the rbinomial stata command). The two groups were first compared to ensure that they did not differ in relation to the prevalence of hypertension. Then the fully adjusted logistic regression was re-executed in each group, and the coefficients in each sub-sample were statistically compared using the suest Stata function. For the second sensitivity test, ORs were visually compared between the imputed dataset and the same logistic regression that only included participants with complete datasets.

Finally given that hypertension is typically associated with disability and lower activity levels, differences in each pain interference domain between those with and without hypertension were examined using simple inferential statistics (t-tests and hedges g). Differences in interference levels were considered to show minimal clinically significant differences if they exceeded 10%, moderately different if they exceeded 30%, and substantially different if they exceeded 50% [43].

Results

Cohort overview

Data were provided by 43 pain management services in four Australian states. During the study period 52,873 patients at Australian pain management services were entered into epiCentre. A total of 45,272 (85.6%) patients completed initial questionnaires during this period, of whom 43,789 (96.7%) were eligible for inclusion, Fig 1.

Fig 1

STROBE recruitment chart.

Notes: 18 patients met multiple exclusion criteria (*). Age: mean = 52.17 years, SD = 15.59.

The cohort were, on average, 52 years of age (sd = 15.59, 95% CI: 27 to 79 years), and were predominantly female (n = 25,342, 57.9%). Most patients were seeking treatment for pain that commenced after an injury (n = 20,185, 48.0%); however, a large proportion reported that their pain had no obvious cause (n = 7,583, 18.0%), commenced after surgery (n = 4,139, 9.8%) or illness (n = 4,795, 11.4%), or had some other cause (n = 5,392, 12.8%); missing n = 1,695. Almost two thirds of patients reported that they had pain localised in at least one region of the back (n = 25,176, 57.5%). Most patients with pain in the arms or legs reported pain that affected both sides, and less than ten percent of the cohort reported pain that pain affected one side of specific body regions only, Fig 2. More than two thirds (n = 30,478, 69.6%) of patients reported three or more sites of pain, and 14,401 (32.9%) reported seven or more sites of pain. Additional demographic characteristics are reported in Table 1.

Fig 2

Percentage of all patients reporting pain in each body region.

Table 1

Association between demographic, clinical and pain characteristics with hypertension (N = 43,789) adjusting for all demographic and clinical characteristics.

	Total	No hypertension	Hypertension
	N (%)	N (%)	N (%)	OR (95% CI)	AOR (95% CI)
Sex c
Male	18,405 (42.1)	14,098 (42.4)	4,307 (41.2)	Reference	Reference
Female	25,342 (57.9)	19,191 (57.6)	6,151 (58.8)	1.05 (1.00, 1.10)	0.88 (0.83, 0.93)
Age d
18 to 24 years	1,458 (3.3)	1,428 (4.3)	30 (0.3)	Reference	Reference
25 to 34 years	4,733 (10.8)	4,523 (13.6)	210 (2.0)	2.21 (1.50, 3.25)	1.87 (1.27, 2.75)
35 to 44 years	7,991 (18.2)	7,219 (21.7)	772 (7.4)	5.09 (3.52, 7.36)	3.74 (2.58, 5.41)
45 to 54 years	10,777 (24.6)	8,706 (26.1)	2,071 (19.8)	11.32 (7.86, 16.31)	7.40 (5.13, 10.67)
55 to 64 years	9,340 (21.3)	6,298 (18.9)	3,042 (29.0)	22.99 (15.97, 33.09)	13.70 (9.50, 19.76)
65 to 74 years	5,516 (12.6)	3,116 (9.4)	2,400 (22.9)	36.66 (25.44, 52.84)	20.29 (14.04, 29.34)
75 to 84 years	3,181 (7.3)	1,605 (4.8)	1,576 (15.0)	46.74 (32.34, 67.55)	27.53 (18.96, 39.97)
> = 85 years	791 (1.8)	420 (1.3)	371 (3.5)	42.05 (28.54, 61.95)	30.65 (20.63, 45.55)
Birth Region e
Oceania and Antarctica	30,776 (73.7)	24,025 (75.8)	6,751 (67.2)	Reference	Reference
Other	10,979 (26.3)	7,691 (24.2)	3,288 (32.8)	1.52 (1.45, 1.60)	1.14 (1.08, 1.21)
IRSAD Quintiles f
5 (lowest disadvantage)	9,420 (22.8)	7,279 (23.2)	2,141 (21.5)	Reference	Reference
4	7,404 (17.9)	5,684 (18.1)	1,720 (17.3)	1.03 (0.96, 1.11)	1.05 (0.97, 1.14)
3	9,205 (22.3)	6,961 (22.2)	2,244 (22.5)	1.10 (1.02, 1.17)	1.08 (1.00, 1.16)
2	8,196 (19.8)	6,162 (19.6)	2,034 (20.4)	1.12 (1.05, 1.20)	1.10 (1.01, 1.19)
1 (highest disadvantage)	7,123 (17.2)	5,299 (16.9)	1,824 (18.3)	1.17 (1.09, 1.26)	1.12 (1.03, 1.22)
Body Mass Index g
Normal weight	8,751 (27.2)	680 (2.8)	1,098 (14.7)	Reference	Reference
Underweight	756 (2.4)	7,653 (31.0)	76 (1.0)	0.78 (0.61, 1.00)	0.81 (0.63, 1.04)
Overweight	9,975 (31.0)	7,828 (31.7)	2,147 (28.7)	1.91 (1.77, 2.07)	1.60 (1.47, 1.74)
Obese, Class I	6,711 (20.9)	4,741 (19.2)	1,970 (26.4)	2.90 (2.67, 3.14)	2.30 (2.09, 2.52)
Obese, Class II	3,343 (10.4)	2,196 (8.9)	1,147 (15.3)	3.64 (3.31, 4.00)	2.90 (2.62, 3.21)
Obese, Class III	2,601 (8.1)	1,566 (6.3)	1,035 (13.8)	4.61 (4.16, 5.10)	3.89 (3.47, 4.36)
Comorbidities
Arthritis	14,496 (33.1)	9,372 (28.1)	5,372 (51.3)	2.86 (2.72, 2.98)	1.50 (1.42, 1.59)
Diabetes h	4,881 (12.7)	2,274 (7.9)	2,607 (27.3)	4.38 (4.12, 4.66)	2.42 (2.26, 2.60)
Anxiety i
Normal/mild	17,030 (40.6)	12,898 (40.8)	3,731 (37.5)	Reference	Reference
Moderate	7,637 (18.2)	5,780 (18.3)	1,857 (18.7)	1.11 (1.04, 1.18)	1.07 (1.00, 1.16)
Severe/Extremely severe	17,266 (41.2)	12,912 (40.9)	4,354 (43.8)	1.17 (1.11, 1.23)	1.26 (1.17, 1.36)
Depression j
Normal/mild	14,598 (34.7)	10,741 (33.9)	3,462 (34.7)	Reference	Reference
Moderate	7,718 (18.3)	5,899 (18.6)	1,819 (18.2)	0.96 (0.90, 1.02)	0.91 (0.84, 0.98)
Severe/Extremely severe	19,777 (47.0)	15,073 (47.5)	4,704 (41.7)	0.97 (0.92, 1.02)	0.87 (0.80, 0.95)
Pain Source k
Post-injury	20,185 (48.0)	16,017 (50.1)	4,168 (41.2)	Reference
Post-surgery	4,139 (9.8)	2,975 (9.3)	1,164 (11.5)	1.50 (1.39, 1.62)	1.10 (1.01, 1.20)
Related to illness	4,795 (11.4)	3,500 (10.9)	1,295 (12.8)	1.42 (1.32, 1.53)	1.09 (1.00, 1.18)
No obvious cause	7,583 (18.0)	5,596 (17.5)	1,987 (19.6)	1.36 (1.28, 1.45)	1.06 (0.99, 1.14)
Other cause	5,392 (12.8)	3,891 (12.2)	1,501 (14.8)	1.48 (1.38, 1.59)	1.08 (1.00, 1.17)
Pain Duration l
3 to 12 months	5,681 (13.6)	4,499 (14.3)	1,182 (11.7)	Reference	Reference
12 to 24 months	6,410 (15.4)	5,074 (16.1)	1,336 (13.2)	1.00 (0.92, 1.09)	1.02 (0.93, 1.13)
2 to 5 years	10,095 (24.2)	7,860 (24.9)	2,235 (22.1)	1.08 (1.00, 1.17)	1.06 (0.97, 1.16)
> 5 years	19,508 (46.8)	14,136 (44.8)	5,372 (53.1)	1.45 (1.35, 1.55)	1.09 (1.00, 1.18)
Widespread Pain Index
< 3 sites	13,311 (30.4)	10,225 (30.7)	3,086 (29.5)	Reference	Reference
3–6 sites	16,077 (36.7)	12,275 (36.8)	3,802 (36.3)	1.23 (1.12, 1.35)	1.00 (0.94, 1.06)
7+ sites	14,401 (32.9)	10,817 (32.5)	3,584 (34.2)	1.48 (1.35, 1.63)	1.06 (0.99, 1.14)
Pain Severity m
Low	3,105 (7.5)	2,544 (7.9)	597 (6.1)	Reference	Reference
Moderate	21,259 (51.4)	16,730 (52.0)	4,827 (49.1)	1.23 (1.22, 1.36)	1.15 (1.03, 1.29)
High	16,962 (41.1)	12,873 (40.0)	4,410 (44.8)	1.47 (1.34, 1.62)	1.17 (1.04, 1.31)
Pain Interference n
Low	3,059 (7.2)	2,356 (7.3)	703 (6.9)	Reference	Reference
Moderate	14,159 (33.4)	10,761 (33.4)	3,398 (33.5)	1.03 (0.97, 1.08)	1.00 (0.89, 1.12)
High	25,128 (59.3)	19,074 (59.3)	6,054 (59.6)	1.10 (1.04, 1.16)	0.99 (0.88, 1.12)
Pain Self-Efficacy o
Low impairment	3,739 (8.9)	2,798 (8.7)	941 (9.3)	Reference	Reference
Mild impairment	5,393 (12.8)	4,012 (12.5)	1,381 (13.7)	1.02 (0.93, 1.13)	1.07 (0.96, 1.19)
Moderate impairment	10,945 (26.0)	8,378 (26.2)	2,567 (25.5)	0.91 (0.84, 0.99)	1.00 (0.90, 1.10)
Severe impairment	22,003 (52.3)	16,815 (52.5)	5,188 (51.5)	0.92 (0.85, 0.99)	1.02 (0.92, 1.13)
Pain Catastrophizing p
Clinically normal	10,589 (25.8)	8,036 (25.7)	2,553 (25.9)	Reference	Reference
High	8,604 (20.9)	6,616 (21.2)	1,988 (20.2)	0.95 (0.88, 1.01)	1.01 (0.93, 1.09)
Clinically elevated	21,918 (53.3)	16,611 (53.1)	5,307 (53.9)	1.01 (0.95, 1.06)	0.99 (0.91, 1.07)

Abbreviations: AOR = Adjusted Odds Ratio, IRSAD = Index of Relative Social Advantage and Disadvantage.

Missing data

c n = 42

d n = 2

e n = 2,034

f n = 2,441

g n = 11,652

h n = 5,341

i n = 2,257

j n = 2,091

k n = 1,695

l n = 2095

m n = 2,463

n n = 1,443

o n = 1,709

p n = 2,678.

Notes: AOR adjusted for all demographic and clinical characteristics associated with hypertension (p<0.20), with multiple imputation with chained equations to estimate missing covariate data.

Prevalence of hypertension in the ePPOC cohort

A total of 10,472 (23.9% crude prevalence; 20.0% age-adjusted prevalence) patients registered to ePPOC reported having high blood pressure in their referral questionnaires. Hypertension was significantly more prevalent in the ePPOC cohort than the general population in 2014–15 (RR = 5.86, 95%CI: 5.66 to 6.06), 2017–18 (RR = 9.40, 95%CI: 9.01 to 9.80), and compared with patients attending primary care practices in a single health district in NSW, Australia (2011–13: RR = 1.17, 95%CI: 1.15 to 1.20). The age-adjusted rates of hypertension were higher than the unadjusted rates in the ePPOC cohort relative to primary care attendees (RR: 1.68, 95%CI: 1.64, 1.72).

Patient characteristics associated with hypertension

Older age was consistently associated with higher odds of hypertension, whereas women had 12% lower adjusted odds of having hypertension. People living in neighbourhoods with higher levels of socioeconomic disadvantage had 10–12% higher adjusted odds of having hypertension, and people who were born outside of the Oceania region had 14% higher adjusted odds of hypertension.

Increasing BMI levels had markedly higher adjusted odds of having hypertension, from 60% higher odds for overweight patients up to 3.89-fold higher odds for those with Class III obesity. The adjusted odds of hypertension was 50% higher in people with arthritis, and 2.42-fold higher in people with diabetes. Severe and extremely severe anxiety symptoms increased the adjusted odds of hypertension by 26%; however, depression symptoms were associated with lower odds of having hypertension by 9–13%, but only when adjusting for all other demographic and clinical covariates.

Hypertension was more prevalent in people whose pain was began after surgery, with 10% higher adjusted odds compared with people whose pain began after injury. While widespread pain had a crude association with hypertension, with 23% higher odds in people with 3–6 sites of pain and 48% higher odds in those with seven or more sites of pain, this association was no longer significant when adjusting for other demographic, clinical and pain-related characteristics. Moderate and high levels of pain severity increased the odds of having hypertension by 23–47%; however the magnitude of these associations was reduced to 15–17% when adjusting for all demographic, clinical and pain characteristics.

Sensitivity analyses

Sensitivity analyses showed that the overall results were consistent between two randomly generated sub-groups, χ2(42) = 27.86, p = 0.95. The findings were consistent for most variables when comparing the AOR for the analyses that included multiple imputation (N = 43,789) and those that had complete datasets (N = 21,653), S1 Table. The exceptions were that the complete dataset showed a significantly increased odds of hypertension only in the most disadvantaged neighbourhood versus the two most disadvantaged neighbourhoods in the imputed dataset, and in people who had no obvious cause for their pain, which did not significantly differ from the reference group (pain after injury) in the imputed dataset. On the contrary, there was only reduced odds of having hypertension for people with severe/extremely severe depression in the complete case analysis versus reduced odds for both moderate and severe/extremely severe depression groups in the imputed dataset. There was reduced odds of hypertension for people in the underweight BMI range, which was not significant in the imputed dataset. While the direction of the associations between hypertension and pain severity, age, and BMI were in the same direction in both analyses, the magnitude of the associations were slightly stronger in the complete case analysis than in the imputed dataset.

Hypertension and interference of pain with activity, mood and enjoyment of life

Very small differences in pain interference were reported by patients with hypertension compared with those without hypertension with the exception of walking ability which had a mean difference of 0.69; however, this difference was not clinically significant, Fig 3, and S2 Table.

Fig 3

Mean (standard deviation) levels of pain interference between patients with hypertension and without hypertension, raw values are available in S2 Table.

Discussion

This large pain registry-based study found that one in four patients attending pain clinics had hypertension. Hypertension prevalence in these patients was 5.6 to 9.8 times higher than the Australian general population, and 1.2 times higher than the Australian primary care setting, and 1.7 times higher when adjusting for age in the clinical samples. While the direction of these findings is consistent with previous studies, the prevalence of 24% in this large and clinically representative cohort was much lower than the 39% prevalence found in a much smaller cohort of 300 patients attending an American pain clinic [8]. This suggests that previous studies may have overstated the prevalence of hypertension in people living with persistent, disabling pain who present to a pain clinic for treatment. However, it is possible that differences in the demographic and clinical characteristics between Australian and American patients may explain the differences in the hypertension prevalence. Hypertension was associated with advancing age, higher BMI, greater socioeconomic disadvantage, being an immigrant (i.e., being born outside of the Oceania region), having arthritis or diabetes, and moderate to extremely severe anxiety symptoms. Depression, however, was not associated with hypertension in unadjusted analyses, and appeared to be protective in the adjusted analyses.

The relative risk of having hypertension in the pain clinic sample was higher when compared with the sample representative of the Australian population and a primary care cohort, although the hypertension prevalence in the pain clinic sample was much more similar to the clinical sample relative to the general population. These observations are understandable given that people seeking treatment in both a primary care clinic and a pain clinic are likely to be older and to have more comorbid conditions, especially conditions requiring treatment with medications like hypertension, than people drawn from the general population. While we could not examine covariate adjusted prevalence of hypertension between the pain clinic cohort and the NHS and primary care samples, we could contrast the age-adjusted prevalence of hypertension between the clinical samples. These analyses showed that people attending a pain clinic had an even higher prevelance of hypertension compared with patients in a primary care setting when adjusting for patient age. This highlights the importance of accounting for age when examining the prevalence of conditions like hypertension that are known to increase in prevalence across the lifespan.

Hypertension was associated with several pain-related characteristics, particularly having pain that was related to illness, having pain for more than five years, pain of moderate or high severity, widespread pain, higher pain interference, lower pain self-efficacy and higher pain catastrophizing. However, when adjusting for all covariates hypertension was only associated with moderate to high pain severity. The other pain-related factors probably did not remain uniquely associated with hypertension when considered together given that they covary with each other and with other pertinent demographic characteristics. Patients with hypertension reported higher levels of interference of pain with their walking ability, general activity, and work; however, the magnitude of these differences was not clinically significant, and other factors probably have a stronger association with pain interference than blood pressure. A recent study found that the association between pain and blood pressure appeared to act via shared risk factors for both pain and cardiovascular disease, especially obesity [14], which may explain why most pain-related characteristics were no longer associated with hypertension in the fully adjusted analyses.

The present finding that hypertension is positively associated with the severity of persistent pain is consistent with previous studies in people attending primary care services in the United Kingdom [3], and patients attending an American pain clinic [8]. Not surprisingly, factors that had the strongest association with hypertension (i.e., the greatest increase in odds) were attributable to life stage (i.e., age), and health and lifestyle factors (i.e., BMI, diabetes), which are known to have causal biological impacts on blood pressure [44]. Given that additional conditions like arthritis and diabetes, and poorer general health with high BMI, were associated with higher odds of having hypertension, multi-morbidity may be especially problematic in patients with both persistent pain and hypertension.

The protective effect of depression was not expected given that previous studies have shown that depression is positively associated with cardiovascular disease [45]. However, several studies have shown that there is limited evidence of a causal association between depression and hypertension [46], and even some evidence of a negative association between depression and blood pressure in community samples [47, 48]. One prospective study found that low blood pressure increased the incidence of later depression suggesting that blood pressure may play a causal role in depression [49], perhaps via enhanced fatigue and somatic symptoms. Further research is required to replicate the present findings in people living with persistent pain.

Biological models suggest that the higher prevalence of hypertension in people with hypertension may be due to insufficient inhibition and heightened facilitation in shared ascending/descending networks that govern both nociception and blood pressure [50]. In particular, people with persistent pain have been found to have lower baroreceptor sensitivity [12], suggesting that baroreflexes have reduced control over cardiovascular functions in people with persistent pain. Recent meta-analyses have shown that people with persistent pain, especially those with conditions involving widespread pain [51], have diminished capacity to engage inhibitory parasympathetic cardiovascular activity [13]. More recently, diminished activity in the parasympathetic nervous system has been found to mediate the association between persistent pain and hypertension [11].

As patients who have comorbid persistent pain and hypertension are at much higher risk of other poor health outcomes, including premature mortality [5, 52], routine screening and tailored treatments may be required to address the common biological and lifestyle risk factors for both conditions. The most effective management of these chronic conditions probably requires interdisciplinary care to ensure that patients do not only receive specialist pain management, but concurrent and integrated treatment of their comorbid conditions. This poses a challenge in many health systems that often have “siloed” approaches to chronic disease management.

Future research

Further research is required to develop and evaluate the impact of interventions that target pain, hypertension, and the common lifestyle and health risk factors for both conditions, on clinical outcomes. Collaborative care has been identified as an emerging priority that may lead to better management of shared health and lifestyle factors associated with a range of chronic conditions. Moreover, specific lifestyle interventions and pharmacological treatments for hypertension [53], or transdisciplinary interventions targeting lifestyle to improve diet, activity, general health and psychological wellbeing [54] may improve both pain and cardiovascular outcomes. Cognitive-behavioural and lifestyle-focussed interventions can successfully improve activity capacity and fitness, and reduce activity avoidance [55], while also reducing excess body weight [56], through gradual increases in physical therapy and activity pacing. These treatments are generally safe for patients with persistent pain cardiovascular disease risks like hypertension, and are likely to have wide-ranging health benefits including improved exercise tolerance, reduced systolic blood pressure, improved quality of life [57], and reduced pain-related disability [58, 59]. Several complementary rehabilitation therapies, that are available in the community and in some pain clinics, may also be beneficial for patients with comorbid hypertension and pain, including yoga, mindfulness, meditation, relaxation, Feldenkrais, or tai chi [60, 61]. While isolated interventions tend to only have a small impact on either blood pressure or pain, the greatest benefits have been found when implementing a multipronged approach comprising behavioural and pharmacological treatments together with bolstering social and clinical support [62]. Moreover, given the strong association between hypertension and other chronic conditions (i.e., diabetes and arthritis) providing a range of education and supportive care resources, action plans, psychological strategies and lifestyle advice may enhance treatment effectiveness by fostering greater self-management [62].

The present study provides some insight into the characteristics associated with hypertension in people living with persistent pain. However, further research is required to better understand the direction of the relationship between pain and hypertension using prospective study designs. In particular, we must examine whether particular pain-related illnesses or lifestyle behaviours play a causal or mediational role in the development and resolution of hypertension in the context of persistent pain. Given the association between pain severity and diseases like diabetes and hypertension, it may be that these conditions arise in part due to limited capacity of the heart to meet metabolic demands of everyday activities. This is particularly important given that cardiac reserve has important implications for the ability to respond to stress [63], mechanisms that are known to be impaired in several persistent pain conditions [13]. Moreover, cluster or network analyses may provide further insight into the patterns of demographic, lifestyle, illness and pain-related characteristics associated with hypertension. These analyses could be especially helpful in guiding the development of targeted interventions to better manage both persistent pain and hypertension.

Limitations

Some limitations of this study should be considered. First, patients were only asked whether they had “high blood pressure”, among a list of comorbid conditions, and clinical diagnoses or pharmacological management of those comorbidities were not recorded. While previous studies have shown that self-report of hypertension or high blood pressure is typically consistent with clinical diagnoses or treatment for hypertension based on electronic medical record review [64], some studies have found under-reporting of hypertension status in 8 to 10 percent of cases [65, 66]. Moreover, a recent meta-analysis showed that self-reported hypertension had only 42% sensitivity and 90% specificity for objectively assessed hypertension in epidemiological studies [67]. It is notable, however, that the data in that meta-analysis were heterogeneous. Given that the majority of studies under-estimated hypertension in self-reported data we speculate that the present study underestimates the prevalence of hypertension in pain clinic attendees compared with the general population, especially for patients who may not have been diagnosed with hypertension, forgot to report their diagnosis, were unwilling to disclose their diagnosis, or who had not had a recent blood pressure assessment [68]. To ensure better comparability with other epidemiological data sources such as the NHS it may be useful to modify the questions about comorbidities. Moreover, the associations between hypertension and persistent pain may be better understood through evaluation of blood pressure across the spectrum (i.e., as a continuous variable based on blood pressure recordings), rather than as a dichotomised characteristic of “normal” vs “high” blood pressure. This is particularly pertinent given that the curvilinear relationship between blood pressure and risk of major cardiovascular events [69]. The present findings should be replicated in studies that also objectively record blood pressure to confirm the presence of probable hypertension.

It is possible that hypertension prevalence between our sample and the comparison cohorts differed due to time-varying differences in population characteristics. Moreover, demographic and clinical differences between the samples may partially explain these results, particularly given the increased risk ratios when accounting for age in the respective clinical populations. In ePPOC, several factors associated with hypertension are not recorded including smoking, alcohol consumption, physical activity levels, specific medications taken, or individual level socioeconomic measures such as education, occupation skill level, household income, or marital status. Nonetheless, neighbourhood level socioeconomic disadvantage was fairly consistently associated with higher odds of hypertension highlighting the potential role of social determinants in clinical characteristics. To enable analysis of health profiles and outcomes in people with persistent pain future studies could examine the role of smoking, physical activity or education.

We calculated BMI using previously recommended data cleaning criteria [39], and classified BMI according to the World Health Organization [20] cut-off ranges. While these are considered to be robust methods for cleaning large datasets and identifying people who were overweight or obese, we may have omitted some genuine BMI data for people in the extremely low or high weight and height ranges. Moreover, we acknowledge that BMI is not necessarily the best measure of overweight and obesity given that people with high lean mass could be classified in the same way as someone with high levels of body fat [70], and different cut-off points have been recommended for Asian populations [71]. Unfortunately implementation of alternate thresholds was not possible as our dataset only recorded country of birth and not ethnicity. Moreover, indices like percentage body fat, waist-to-hip ratio or waist circumference were not available. It should be noted that our sensitivity analyses revealed that the inclusion of all cases with imputation of missing data generated slightly more conservative odds ratios than if we had only included cases with complete data; however, these analyses did not have a substantial effect on which clinical or pain-related characteristics were associated with hypertension. Finally, the cross-sectional design of this study prevents us from making causal assumptions. The association between blood pressure and moderate to severe pain may be bidirectional [7] given that longitudinal studies show that cardiovascular disease is predictive of developing persistent pain [72, 73]. Prospective population cohort studies with objective assessment of blood pressure are required to gain insight into long-term causal mechanisms involved in comorbid pain and hypertension.

Conclusions

For the first time we have documented the prevalence of hypertension in a large clinically representative cohort of patients attending tertiary pain management services in Australia. Several aspects of pain were associated with hypertension, particularly moderate to high pain severity. Most other characteristics associated with hypertension were consistent with life-stage and health-related mechanisms (i.e., advancing age, higher BMI and other comorbid conditions like arthritis and diabetes). Future studies should evaluate the effectiveness of treatments targeting factors associated with hypertension in patients with both pain and hypertension. Considering the population prevalence of persistent pain and cardiovascular diseases is expected to increase markedly over the next 30–40 years [74-76], the development of treatment protocols and guidelines for collaborative management of common comorbid chronic conditions will be essential to reducing the burden of those diseases.

Comparison of adjusted ORs for the imputed dataset (N = 43,789) and the cohort with complete data only (N = 21,653).

(DOCX)

Click here for additional data file.

Levels of pain interference in patients with and without hypertension.

(DOCX)

Click here for additional data file.

19 Nov 2019

PONE-D-19-21769

Hypertension prevalence in patients attending tertiary pain management services, a registry-based Australian cohort study

PLOS ONE

Dear Dr. Giummarra,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

==============================

The reviewers address some points that should be considered. Please provide also a reply regarding availability of the data.

==============================

We would appreciate receiving your revised manuscript by Jan 03 2020 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Hans-Peter Brunner-La Rocca, M.D.

Academic Editor

PLOS ONE

Journal Requirements:

1. When submitting your revision, we need you to address these additional requirements.

Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

http://www.journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and http://www.journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2.  Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.

3. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially identifying or sensitive patient information) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. Please see http://www.bmj.com/content/340/bmj.c181.long for guidelines on how to de-identify and prepare clinical data for publication. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: Yes

Reviewer #3: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Nice and clear paper, reads well and well explained. The finding of increased prevalence of hypertension in those visiting tertiary pain centers is interesting, but as this is the main finding, I feel it needs a bit more context where the comparison with the general population and GP patients is concerned;

In the methods section, the authors state that hypertension in the ePPOC was based on self-report. The NHS that was used for the general population was also based on self-report, but it remains unclear what the exact question was used for the NHS - some light is shed on it in the discussion, but that should be done much earlier. Similarly for the GP cohort, it is not clear from the methods on p7 how hypertension yes/no was defined here. Please specifiy this in the paper.

Furthermore, what is the explanation for the fact that hypertension in pain patients is much higher in the general population but not so much higher than in the GP group? Could this be just the fact that GP and pain clinic patients are generally less healthy, older etc? Nothing is said about this, it is only stated that there is another study with even higher prevalence - but we don't know whether this was self report as well, or not. The comparison with the other cohorts is relevant and has added value, but if it will be part of this paper it needs to be properly discussed.

- page 7 statistical analysis - the weight criteria (and height, to lesser extent) used to define whether the data entered was valid or not seem rather extreme, would a man said to weigh 35 kg really be included as such in the BMI calculations? Probably there were no such cases in the dataset but this sentence makes one wonder what could have gone wrong in the analyses.

- the protective effect of depressive symptoms is rather surprising. Has this ever been described before? Would there be an explanation?

Minor textual revisions:

- Abstract second last sentence: 'Hypertension was more prevalent in people with persistant pain, is associated severe pain' - this should probably be 'associated WITH severe pain'

- p13 - below middle - 'a recent study recent found' - the last recent should be deleted

- p16 'this particularly pertinent' - this IS particularly pertinent?

- p17 conlusion last sentence 'will be integral' - I would expect something like vital or essential instead of integral but that may not be what was intended.

Reviewer #2: Giummarra and colleagues aimed to evaluate the prevalence of hypertension in a population suffering from chronic pain. For this purpose they evaluated the electronic Persistent Pain Outcomes Collaboration Registry between 2013 and 2018. They show that patients with chronic pain suffer more frequently from hypertension. Moreover, pain severity was independently associated with hypertension after adjustment for covariates.

The main limitation of the study relies on the self-reporting methods used to evaluate medical conditions and comorbidities. The fact that patients are not always aware of the medical conditions they suffer from, may introduce an important bias in the analysis.

Furthermore, the covariates that has been used to calculate the adjusted model should be mentioned in the statistical analysis section.

The authors have chosen to impute covariates. Why perform imputation? Is there any baseline difference between the imputed and non-imputed data?

Baseline characteristics show total and hypertensive population, why is there the non-hypertensive population not included? Supplementary table 1 shows this difference for a very limited amount of variables, but would be interesting to evaluate the differences in all studies variables and to add it in an additional supplementary table.

A visual evaluation of the main variables (i.e. pain severity and other comorbidities like Diabetes Mellitus and arthritis) and hypertension would be very illustrative and would be helpful to understand a quite complex statistical analysis in a single shoot. For this purpose, I would suggest to include a Network analysis figure.

It is remarkable that specially illness related pain and pain severity are related with hypertension. Could be related with underlying diseases and therefor poorer cardiovascular reserve? In this that this issue should be broader discussed in the discussion. Moreover, a mediation analysis could be very helpful to elucidate the direction of this relationships.

Reviewer #3: The paper of Guimmarra et al investigates 1) the prevalence of hypertension and 2) the factors associated with hypertension in patients in tertiary pain clinics. To this end, the Persistent Pain Outcomes Collaboration registry was used (n=43,789). 23.9% had self-reported hypertension. Factors associated with self-reported hypertension were higher age, lower socioeconomic status, higher BMI, born outside Australasia, comorbid arthritis, diabetes or severe anxiety symptoms. Protective factors were female sex and depressive symptoms. The authors suggest that screening for hypertension in pain clinics may improve treatment outcomes.

Several points may be addressed:

1. What was the rationale to categorize variables such as age, BMI and other scores?

In addition, for BMI, there may be some misclassification, as the cutoff point for overweight is different for Asians.

2. Have the authors considered multiple testing and ways to correct for this (e.g. false discovery rate?)

3. Can authors reflect on what the potential consequence may be of using only self-reported hypertension (e.g. underestimation or overestimation of the effect sizes?)

4. Authors correctly state in the limitations that the higher hypertension prevalence in the ePOCC sample compared with the general population/primary care samples may be due to demographic and clinical differences. Do the authors have information about covariates in the general population sample and primary care sample? Otherwise, authors adjust for such differences and present adjusted RRs, which is more informative.

5. Many previous studies suggest that higher blood pressure is associated with depression, as opposed to ‘no hypertension’. Was this finding a false positive finding? Or is there a biological mechanism to this phenomenon?

6. Table 1 may be more informative if the ‘no hypertension’ group was described as well (authors may consider to omit the ‘total’ group). I understand that it can be calculated by the reader, but it is a hassle.

7. On page 15, first line, BP was used as an abbreviation, but was not defined earlier. Please check all abbreviations if they were defined.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: Yes: Tan Lai Zhou

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

19 Dec 2019

Dear Professor Hans-Peter Brunner-La Rocca,

We thank you for the opportunity to revise and resubmit our manuscript. The reviewers have made some excellent suggestions, which we have taken on board in our revision of the manuscript. We respond below, point by point, to each comment and indicate where the respective changes have been made to the track change version of the manuscript.

On behalf of my co-authors,

Dr Melita Giummarra

Reviewer #1: Nice and clear paper, reads well and well explained.

The finding of increased prevalence of hypertension in those visiting tertiary pain centers is interesting, but as this is the main finding, I feel it needs a bit more context where the comparison with the general population and GP patients is concerned. In the methods section, the authors state that hypertension in the ePPOC was based on self-report. The NHS that was used for the general population was also based on self-report, but it remains unclear what the exact question was used for the NHS - some light is shed on it in the discussion, but that should be done much earlier. Similarly for the GP cohort, it is not clear from the methods on p7 how hypertension yes/no was defined here. Please specify this in the paper.

RESPONSE: Thank you for this suggestion. We have described the data collection procedures for the NHS and primary care study in the Methods (Page 6-7). To reduce duplication we have moved the NHS interview questions from the Discussion to the Methods section.

Furthermore, what is the explanation for the fact that hypertension in pain patients is much higher in the general population but not so much higher than in the GP group? Could this be just the fact that GP and pain clinic patients are generally less healthy, older etc? Nothing is said about this, it is only stated that there is another study with even higher prevalence - but we don't know whether this was self report as well, or not. The comparison with the other cohorts is relevant and has added value, but if it will be part of this paper it needs to be properly discussed.

RESPONSE: This is an excellent point, which we now discuss further in the Discussion on page 14.

Page 7 statistical analysis - the weight criteria (and height, to lesser extent) used to define whether the data entered was valid or not seem rather extreme, would a man said to weigh 35 kg really be included as such in the BMI calculations? Probably there were no such cases in the dataset but this sentence makes one wonder what could have gone wrong in the analyses.

RESPONSE: We understand the reviewers concern; however, we were not aware of any other criteria to apply during data cleaning to ensure that our BMI data were valid. As ePPOC relies on patients and clinicians to enter data, it appears that in some cases height and weight data have been entered erroneously. Upon screening the data we identified that the most common errors were probably that these data were entered into the incorrect field (i.e., weight in the height field, and vice versa), or that height and weight were entered in a metric other than kilograms or metres.

Given that this was such a large dataset we needed to apply a systematic rule to omit these data entry errors and probable invalid height and weight combinations. We therefore elected to use criteria that had previously been used in the Australian epidemiological study. We acknowledge this as a limitation on Page 18.

The protective effect of depressive symptoms is rather surprising. Has this ever been described before? Would there be an explanation?

RESPONSE: We have further discussed this finding in the context of the literature on the association between depression and hypertension on Page 15.

Minor textual revisions:

Abstract second last sentence: 'Hypertension was more prevalent in people with persistent pain, is associated severe pain' - this should probably be 'associated WITH severe pain'.

RESPONSE: This part of the abstract was highlighting that the prevalence was higher in our cohort with persistent pain than in the general community. To make this clearer we’ve revised the wording as follows:

“Hypertension was more prevalent in people with persistent pain than in the general community, was associated with more severe pain, and commonly co-occurred with pain-related impairments”

p13 - below middle - 'a recent study recent found' - the last recent should be deleted

RESPONSE: Thank you, we have deleted the second “recent”.

p16 'this particularly pertinent' - this IS particularly pertinent?

RESPONSE: Thank you, we have made this correction.

p17 conclusion last sentence 'will be integral' - I would expect something like vital or essential instead of integral but that may not be what was intended.

RESPONSE: This is a great suggestion. We have rephrased this expression to “will be essential to reducing...” as suggested.

Reviewer #2

Giummarra and colleagues aimed to evaluate the prevalence of hypertension in a population suffering from chronic pain. For this purpose they evaluated the electronic Persistent Pain Outcomes Collaboration Registry between 2013 and 2018. They show that patients with chronic pain suffer more frequently from hypertension. Moreover, pain severity was independently associated with hypertension after adjustment for covariates.

The main limitation of the study relies on the self-reporting methods used to evaluate medical conditions and comorbidities. The fact that patients are not always aware of the medical conditions they suffer from, may introduce an important bias in the analysis.

RESPONSE: We agree that this is the key limitation of the study, which is why we discuss it extensively in the discussion on Page 17-18. We are not sure what else we can add to further highlight this cautionary point.

Furthermore, the covariates that has been used to calculate the adjusted model should be mentioned in the statistical analysis section.

RESPONSE: The covariates are now listed in the statistical analysis section on Page 8.

The authors have chosen to impute covariates. Why perform imputation? Is there any baseline difference between the imputed and non-imputed data?

RESPONSE: We chose to impute missing data for the multivariable analysis so that the analyses included the total cohort, and not the subset of patients with 100% complete data.

For most variables missing data was present for approximately 4-5% of the cohort. The exceptions were that 12% did not indicate whether they had diabetes, and 26% did not have a BMI due to failure to report height or weight, or due to data entry errors resulting in omission of the BMI from analyses. The levels of missing data are reported in the footnote of Table 1.

To examine whether the effects were different in the whole cohort versus those with complete data we now include a “complete cases” analysis in supplementary materials. The imputed dataset did not yield substantially different effects to the complete cases cohort, but we do note in the results that the imputed dataset yielded smaller AORs than the complete dataset (Page 12-13). We feel that it is more appropriate to err on the side of caution and to not overstate the potential relationship between pain and hypertension, and therefore retain the imputed dataset for our primary analyses. We acknowledge these sensitivity analyses in the discussion on Page 19 with the following sentence:

“It should be noted that our sensitivity analyses revealed that the inclusion of all cases with imputation of missing data generated slightly more conservative odds ratios than if we had only included cases with complete data; however, these analyses did not have a substantial effect on which clinical or pain-related characteristics were associated with hypertension.”

Baseline characteristics show total and hypertensive population, why is there the non-hypertensive population not included? Supplementary table 1 shows this difference for a very limited amount of variables, but would be interesting to evaluate the differences in all studies variables and to add it in an additional supplementary table.

RESPONSE: We have added the number (%) of people in the non-hypertensive group with each of the “predictor” characteristics, see Table 1 column 3.

A visual evaluation of the main variables (i.e. pain severity and other comorbidities like Diabetes Mellitus and arthritis) and hypertension would be very illustrative and would be helpful to understand a quite complex statistical analysis in a single shoot. For this purpose, I would suggest to include a Network analysis figure.

RESPONSE: Conducting a network or cluster analysis on these data does sound like a very interesting idea. However, we do not believe that this is necessary to address the aims of the present study: to identify which demographic, clinical and pain-related characteristics are associated with having hypertension in people living with persistent and disabling pain. Moreover, as multicollinearity was not problematic in our data, and we are not concerned about the relationship between the main “predictor” variables.

That said, we do wish to highlight the potential utility of these approaches in future research to allow us to better understand factors associated with having hypertension in the context of persistent pain. Therefore, we have added further discussion about the potential use of cluster and network analyses in future research in the Discussion (see the bottom of the new paragraph on Page 17).

It is remarkable that specially illness related pain and pain severity are related with hypertension. Could be related with underlying diseases and therefor poorer cardiovascular reserve? In this that this issue should be broader discussed in the discussion. Moreover, a mediation analysis could be very helpful to elucidate the direction of this relationships.

RESPONSE: We really like the suggestion that cardiac reserve may be impaired in these patients, which we now briefly discuss on Page 17. As the study data are cross-sectional we are cautious about the fact that a mediation analysis might provide misleading insight into potential directionality of the relationship between hypertension and pain severity. However, future prospective research could certainly address this potential question, which we have now suggested in the discussion on Page 17.

Reviewer #3

The paper of Guimmarra et al investigates 1) the prevalence of hypertension and 2) the factors associated with hypertension in patients in tertiary pain clinics. To this end, the Persistent Pain Outcomes Collaboration registry was used (n=43,789). 23.9% had self-reported hypertension. Factors associated with self-reported hypertension were higher age, lower socioeconomic status, higher BMI, born outside Australasia, comorbid arthritis, diabetes or severe anxiety symptoms. Protective factors were female sex and depressive symptoms. The authors suggest that screening for hypertension in pain clinics may improve treatment outcomes.

Several points may be addressed:

1. What was the rationale to categorize variables such as age, BMI and other scores? In addition, for BMI, there may be some misclassification, as the cutoff point for overweight is different for Asians.

RESPONSE: Thank you for these comments. We chose to use categorical predictors to allow us to identify the differences in odds of having hypertension according to the level of the predictor variable, which does not always vary linearly across the range of possible values. We used recommended or validated cut-points for BMI (i.e., the WHO recommended cu-points) and all questionnaires, and commonly used categories for age, to allow comparison with other studies. We now acknowledge the potential misclassification of some people on the BMI weight ranges based on lean mass or ethnicity in the Discussion (see Page 18-19).

2. Have the authors considered multiple testing and ways to correct for this (e.g. false discovery rate?)

RESPONSE: Thank you for this comment. For the analyses examining the relative risk of having hypertension there were only three analyses. While we report the 95% CIs in the paper, we can report that the p-values were <0.0001 for each of these contrasts, and so they would hold up if adjusting for multiple comparisons (e.g., using the Benjamini-Hochberg procedure) for false discovery rates. Likewise, the results from the t-tests examining differences in pain interference also hold up even with correction for multiple comparisons when applying corrections for false-discovery rate as the p-values are already very low, but we do note in the paper that these are probably not clinically meaningful differences given the small effect sizes.

The primary analyses are the multivariable logistic regression it is not necessary to make corrections for multiple testing when examining factors associated with hypertension. It is routine to report the unadjusted odds when conducting multivariable analyses, but these are not the focus of the analyses.

As a result of these considerations we have not made any changes to the manuscript regarding corrections for false discovery rate.

3. Can authors reflect on what the potential consequence may be of using only self-reported hypertension (e.g. underestimation or overestimation of the effect sizes?)

RESPONSE: We have suggested that our study probably indicates an underestimation of the prevalence of hypertension (page 17, “therefore we speculate that the present study underestimates the prevalence of hypertension in pain clinic attendees compared with the general population”). Our data probably underestimate the relative risk of pain clinic attendees having hypertension compared with the general population, and includes some measurement error in the magnitude of the relationship between demographic, health and pain-related characteristics and hypertension. However, as we cannot be sure which patients underreported their hypertension status to err on the side of caution we do not speculate on the effect of under-reporting on these latter relationships.

4. Authors correctly state in the limitations that the higher hypertension prevalence in the ePOCC sample compared with the general population/primary care samples may be due to demographic and clinical differences. Do the authors have information about covariates in the general population sample and primary care sample? Otherwise, authors adjust for such differences and present adjusted RRs, which is more informative.

RESPONSE: Unfortunately, the NHS reports did not provide age- or covariate-adjusted prevalence rates; however, the sentinel study did provided crude and age-adjusted prevalence of hypertension. We have compared these rates with the age-adjusted prevalence of hypertension in the ePPOC cohort and found an even greater relative risk of hypertension in the ePPOC cohort when accounting for age. These new analyses are reported on Page 11-12, and noted in the first paragraph of the discussion on Page 14.

5. Many previous studies suggest that higher blood pressure is associated with depression, as opposed to ‘no hypertension’. Was this finding a false positive finding? Or is there a biological mechanism to this phenomenon?

RESPONSE: We have further discussed this finding in the context of the literature on the association between depression and hypertension on Page 15.

6. Table 1 may be more informative if the ‘no hypertension’ group was described as well (authors may consider to omit the ‘total’ group). I understand that it can be calculated by the reader, but it is a hassle.

RESPONSE: We have added this column to the Table.

7. On page 15, first line, BP was used as an abbreviation, but was not defined earlier. Please check all abbreviations if they were defined.

RESPONSE: We have changed this to the full expression given that we do not use the same abbreviation again. We have also checked that all other abbreviations are defined appropriately.

9 Jan 2020

Hypertension prevalence in patients attending tertiary pain management services, a registry-based Australian cohort study

PONE-D-19-21769R1

Dear Dr. Giummarra,

We are pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it complies with all outstanding technical requirements.

Within one week, you will receive an e-mail containing information on the amendments required prior to publication. When all required modifications have been addressed, you will receive a formal acceptance letter and your manuscript will proceed to our production department and be scheduled for publication.

Shortly after the formal acceptance letter is sent, an invoice for payment will follow. To ensure an efficient production and billing process, please log into Editorial Manager at https://www.editorialmanager.com/pone/, click the "Update My Information" link at the top of the page, and update your user information. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, you must inform our press team as soon as possible and no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

With kind regards,

Hans-Peter Brunner-La Rocca, M.D.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #2: All comments have been addressed

Reviewer #3: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #2: Yes

Reviewer #3: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #2: Yes

Reviewer #3: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #2: Yes

Reviewer #3: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #2: The authors addressed correctly all comments and improved the manuscript considerably. Supplemental material has been added and the main statistical and methodological issues have been tackled.

Reviewer #3: The authors have sufficiently addressed the comments and the manuscript has ameliorated, I have no further suggestions.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #2: Yes: Arantxa Barandiaran Aizpurua

Reviewer #3: Yes: Tan Lai Zhou

15 Jan 2020

PONE-D-19-21769R1

Hypertension prevalence in patients attending tertiary pain management services, a registry-based Australian cohort study

Dear Dr. Giummarra:

I am pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

For any other questions or concerns, please email plosone@plos.org.

Thank you for submitting your work to PLOS ONE.

With kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Hans-Peter Brunner-La Rocca

Academic Editor

PLOS ONE