COVID-19 and processes of adjustment in people with persistent pain: the role of psychological flexibility

Introduction

The appearance of COVID-19 has had a large and continuing impact on people’s lives [1]. This is particularly true for those in vulnerable circumstances, those with chronic health conditions, or already suffering restrictions in functioning [2], including those with chronic pain [3, 4]. It is apparent that the unprecedented circumstances of the COVID-19 pandemic will require new research directions and clinical approaches to address current priorities, future needs of those affected now, and to prepare for future pandemics [5, 6]. Because people with chronic pain appear to have been particularly affected, these priorities hold especially true for them [7], [8], [9].

Given that the COVID-19 pandemic remains recent history, the number of studies of related health impacts in people with chronic pain is remarkable. This includes studies conducted in Spain [10, 11] Canada [8, 12], the United Kingdom (UK) [3, 9] and the United States [13], to name a few. Quite consistently these studies document relatively poor health and functioning during the pandemic in the context of chronic pain and recommend better solutions to support people with chronic pain to improve their health and wellbeing. This is not to say that all people with chronic pain deteriorated in their health during the pandemic. Some will have clearly stayed the same, or even improved, in each of these cited studies, likely as a result of protective factors.

A set of psychological processes that might afford protection or facilitate recovery from the pandemic are those associated with psychological flexibility (PF) [14]. The PF model includes six facets: acceptance, cognitive defusion, contact with the present, self-as-context, values-based action, and committed action. These essentially entail, in turn, the ability to act with openness to unwanted experiences, without being adversely dominated by misleading thoughts, to be flexibly and purposely aware, to see experiences from a separate non-attached perspective, and know what one wants to do and how, and to build those actions into wider persisting patterns of behavior [14]. Psychological inflexibility (PI) facets, on the other hand, represent the opposing end of these dimensions or a lack of skills to do these things. Many of the PF and PI facets have been studied and applied in treatment for chronic pain outside of the context of COVID-19 [15] and have been recommended for application in this context [7].

In an earlier study conducted in the UK (n=555) three facets of PF, acceptance, self-as-context, and committed action, were examined and were shown to significantly correlate with depression, pain-related disability, and work and social adjustment [9]. Particularly acceptance and committed action made significant unique contributions in prediction of all outcomes, in multivariate prediction equations that included a wide array of background, pain-related, and COVID-19-related variables. No study has so far replicated these findings or investigated the full set of PF and PI factors in people with chronic pain during the time of COVID-19.

The purpose of the current study was to investigate the full range of six PF facets and six PI facets and their respective relations with depression, anxiety, and insomnia in people with chronic pain during the COVID-19 pandemic. We predicted that the full set of facets would significantly predict all three outcomes. We also explore the relative contributions of differing facets and overall PF and PI dimensions themselves without specific predictions regarding which would dominate.

Methods

The data for this study were collected in Sweden between 29st June to 23rd August 2021. It may be remembered that Sweden took a liberal approach to the pandemic at this time, without people being restricted to their homes, or facing other restrictions implemented in other countries. Most businesses remained open, but specialty pain services closed, many people chose to work from home, and face masks were used some of the time, particularly in health care settings. Participants were recruited via social media and through a regional healthcare website for a survey related to experiences of COVID-19 and health. All data were collected via online self-report. The study had ethics approval (Swedish national ethical board, dnr 2021-01647) and all participants provided consent. The analyses presented here are secondary analyses following the completion of primary planned analyses of depression, anxiety, and insomnia in Sweden 18 months after the start of the pandemic [16, 17]. These earlier studies did not focus at all on the subset of participants with persistent pain, and it was a pre-planned secondary study to specifically examine the health and potential health protective factors in this potentially vulnerable population.

Results

Mean and standard deviation values for the PHQ-9, GAD-7, and ISI were M=9.3, SD=6.3, M=6.8, SD=5.3, and M=12.5, SD=6.6, respectively. According to standard cutoff value of 10 for each of these, the percentages of participants with significant depression, anxiety, and insomnia were 43.1, 26.4, and 64.2%, respectively.

Discussion

While there have been many studies of the experience of COVID-19 in relation to people’s health and wellbeing, there are relatively few that have focused specifically on pain as was the main purpose of the current study. In this survey study conducted in Sweden during the summer of 2021 we show first that 38.2% of unselected participants reported pain most days in the past three months and 33.8% reported significant persistent pain, which included pain rated at a three out of 10 or higher in the past week. In the selected sample of people with significant persistent pain we found significant rates of depression, anxiety, and insomnia, at 43.1, 26.4, and 64.2%, respectively. In turn we showed that PF was negatively associated with these problems while PI was positively associated with them.

The rates for depression, anxiety, and insomnia found here are substantially higher than levels previously found in a general sample of Sweden [2] and in the larger sample from which these data were selected, where the rates were 27, 16, and 45% [16]. This suggests a particular vulnerability toward impacts on wellbeing in those with persistent pain during the COVID-19 pandemic.

It is notable to see that pain severity itself obtained small correlations with the health outcomes here, none higher than r=0.28. On the other hand, the overall flexibility dimension score correlated at a similar or higher level with outcomes while the inflexibility score clearly surpassed pain as a correlate with these outcomes. Multiple facets of PF and PI achieved medium or large correlations, especially with depression and anxiety. For insomnia the PI facets reached this level but the PF facets did not. These results seem to show that it is not just the pain itself that leads to difficultly but rather the combination of pain and the context of the individual’s capacity to respond with flexibility or inflexibility to whatever difficulties life presents. The demonstrated potential role of PF and PI in relation to health outcomes during the pandemic is consistent with a previous study in people with chronic pain in the UK [9] and large prospective general population studies in the UK [23] and Italy [24].

Results of the regression analyses once again demonstrate the superiority of PF and PI facets in predicting health outcomes, beyond contributions of background, pain, and even COVID-19 infection. The PF facets accounted for significant variance in depression, anxiety, and insomnia, 17, 17, and 8%, respectively. For the PI facets variance accounted for was significantly greater for these outcomes, 32, 35, and 16%, respectively. The greater variance accounted for by PI is consistent with findings from other studies during the pandemic in a general population, using the same measure [24]. As well as highlighting the greater prediction entailed by the PI facets, particular facets that make strong unique contributions in explained variances included cognitive defusion from among the PF facets, and cognitive fusion and self as content among the PI facets. These results reinforce the well-known role of thoughts and beliefs in human suffering and behavior problems, and highlight a particularly functional contextual way to consider these. According to the processes highlighted, it is both what we think AND how we respond to what we think that matters.

A particularly notable result here is the much better prediction of outcome obtained by PI rather than PF. In some ways this is also consistent with a long history in psychology and chronic pain where “negative” variables such as catastrophizing and fear and avoidance have always performed well in studies while the list of positive variables that achieve a similar status has always been small [25]. We speculate that there may be at least two different reasons for superiority of PI here. One is that the outcomes are negatively framed and so this might create the greater association with the negatively framed PI variables. Second, it may be that the negative instances of behavior patterns and related problems may simply be more salient and assessible than the, in some ways, more subtle aspects of openness, awareness, and engagement. Unless one is aware or is clear on their values, for example, it is difficult to report on how aware one is or how well one is following one’s values.

There are several limitations in the methods of this study. This study employed online and social media-based recruitment methods. This is therefore a selected sample that excludes all those who cannot be recruited this way. The sample is mainly women and well-educated. It is also the case that we cannot independently verify any of the information respondents provide. Because the study was first designed as a study of COVID-19 in Sweden and only secondarily as a study of those with persistent pain, we were limited in gathering some of the information that one might be used to see in pain studies. This is not a clinical sample. We did not collect information on medication or diagnoses per se. Although we defined persistent pain in a conventional way, and even excluded from our analyses those with low levels of pain, we do not know the history of the persistent pain, nor how many precisely had persistent pain that preceded the pandemic and was not brought on in some way by COVID-19 or the pandemic circumstances. So, the population is clearly those with significant persistent pain now, not only those with pre-existing persistent pain. Other standard caveats apply. We note that the results in the regression models in particular are vulnerable to influence of variables selected for inclusion as well as characteristics of the current sample, and should be replicated. These are one-time cross-sectional observation data and we cannot claim direction of relations or cause and effect. Longitudinal and treatment outcome studies are recommended [e.g., 23, 24]. Likewise, we have no pre-pandemic data on the same people so we cannot call the results a change or claim that results represent a direct effect of the pandemic.

In summary, PF and especially PI may play a role in relation to health outcomes in people with persistent pain during the COVID-19 pandemic. This is relevant as this group of people appears to be especially vulnerable to the impacts of the pandemic. This study should motivate further investigation and development of treatment approaches, possibly focusing on training PF, for people with persistent pain in the current pandemic context and for the future.