Parental fear of progression in juvenile idiopathic arthritis and type 1 diabetes

Abstract

Background

This cross-sectional study examined the relevance and impact of fear of disease progression (FoP) to parents of youth with two common childhood chronic conditions—juvenile idiopathic arthritis (JIA) and type 1 diabetes mellitus (T1DM).

Methods

124 parents (57 JIA, 67 T1DM; 107 mothers, 17 fathers) of youth aged 2–18 years (M=11.36, SD=3.90) were recruited through social media and patient organisation groups. Parents completed online questionnaires assessing fear of their child’s disease progression (FoP), responses to their children's symptoms and their own psychopathology. In addition, we measured parent reports of their child’s pain, health-related quality of life (HRQoL), anxiety and depressive symptoms. Bivariate correlations were used to assess the associations between parental FoP and parent and parent-reported youth psychosocial outcomes. A hierarchical regression was conducted to identify the unique contribution of parental FoP to youth HRQoL. Secondary exploratory analyses estimated the proportion of parents with elevated levels of FoP.

Results

Parental FoP was associated with poorer parent and parent-reported youth psychosocial outcomes. After controlling for covariates, parental FoP explained a small amount of additional variance (1.6%) in parent-reported youth HRQoL. Secondary analyses revealed that 76% of parents had elevated levels of FoP.

Conclusions

As FoP appears to be a common concern among parents of youth with JIA and T1DM, which is consistently associated with poorer psychosocial outcomes, future research should be directed towards understanding and treating FoP in parents of children with a chronic condition.

WHAT IS ALREADY KNOWN ON THIS TOPIC.

WHAT THIS STUDY ADDS

• This study revealed that fears about disease progression were a common concern in parents of youth with juvenile idiopathic arthritis or type 1 diabetes, which were associated with a range of poorer parent and youth psychosocial outcomes, including explaining a small amount of additional variance in parent-reported youth quality of life.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• These findings highlight the need to better screen for and support parents in their concerns about their child’s disease progression.

Juvenile idiopathic arthritis (JIA) is a common condition in children, affecting 3.8–400 per 100 000 children.¹ As can be seen from this estimate, reported rates of JIA are often highly variable and dependent on the population, classification criteria and methodology used across studies. It is diagnosed in youth under 16 years who present with inflammatory arthritis which persists for at least 6 weeks.² Type 1 diabetes mellitus (T1DM) is a similarly prevalent condition, estimated to impact 11.07 per 100 000 adolescents and young people aged 10–24 years.³ T1DM is an autoimmune condition which targets the insulin-producing cells in the pancreas, resulting in an inability to regulate blood glucose levels.⁴ Alongside their physical impacts, JIA and T1DM have profound impacts on children’s psychosocial functioning.^{5 6} These conditions also impact the broader family, with parents usually taking on the responsibility and associated stress of disease management, often to the detriment of their mental health.⁷

One construct that may contribute to parents’ distress, which is yet to be investigated, is fear of disease progression (FoP). FoP originates largely from the fear of cancer recurrence literature and is defined as being ‘the fear, worry or concern relating to the possibility that cancer will come back or progress’.⁸ Research suggests FoP is also prevalent among those with non-cancer chronic illnesses⁹ and is also relevant to caregivers as well as patients. For example, a meta-analysis found that 48% of caregivers of patients with cancer have clinical levels of FoP.¹⁰ There are relatively few studies in parents of children with cancer, but research has supported equivalent or even higher levels of FoP than found in adults with cancer or their caregivers.^{11 12} However, research is yet to investigate whether parents of youth with other chronic illnesses also worry about their child’s condition worsening.

But why might FoP be relevant to JIA and T1DM? First, disease activity in JIA is not confined to childhood or adolescence. Rather, a systematic review found that less than half of those with JIA will achieve drug-free remission a decade after their diagnosis.¹³ Research also suggests that 20–47% of adults with JIA have a mild disability and another 3–11% will be severely disabled.^14–16 These high rates of disability may be the result of current disease activity, but likely also encompass the long-term impacts of JIA on the body. As JIA affects the body when it is still developing, it can cause joint damage and deformity.¹⁷ Youth with JIA have also been found to have poorer bone density, which places them at increased risk of osteoporosis.¹⁸ JIA can also cause chronic inflammation of the eyes and increase the risk of cataracts and vision loss.¹⁹ Lastly, there is also emerging evidence to suggest that youth with JIA have more cardiac abnormalities and may be at greater risk of cardiovascular events.²⁰ Given this, parents of youth with JIA may be understandably concerned about the long-term impacts and possible progression of their child’s condition.

T1DM is also associated with acute and long-term complications. When severe, hypoglycaemia can result in loss of consciousness, seizures and even death, making it a major fear for parents.^21–23 To mitigate this risk, parents may maintain their child’s blood glucose higher than recommended.^{24 25} This is a double-edged sword, as poorer glycaemic control is a risk factor for future diabetes-related complications and mortality.²⁶ This includes vascular complications, such as coronary heart disease and stroke, of which adults with T1DM are more than twice as likely to die from compared with healthy controls.²⁷ Other non-fatal, but debilitating complications include neuropathy, lower-limb amputation and retinopathy, which can cause vision loss.^{28 29} Despite qualitative research indicating that parents of youth with T1DM worry about long-term complications and reduced lifespan,²³ no quantitative research has investigated the relevance of FoP to parents in this group or its relationship to important psychosocial outcomes.

Heightened parent FoP may negatively impact parent and child alike. In the cancer literature, caregiver FoP is associated with greater general anxiety and depression symptoms.¹⁰ Adults with rheumatic conditions and clinical levels of FoP have also been found to have more pain, depression and anxiety and worse health-related quality of life (HRQoL).³⁰ Parent FoP may also negatively impact on patient outcomes; for instance, it has been found to predict worse HRQoL in adults with head and neck cancer³¹ and children with tumours.³² However, research has not examined whether parental FoP is associated with poorer HRQoL in youth with chronic conditions.

The aim of this study is to investigate whether FoP is relevant to parents of youth with JIA and T1DM. To do so, we will examine the associations between parental FoP and both parent psychosocial outcomes and parent-reported youth psychosocial outcomes. Following this, we will then conduct a regression analysis to determine whether parental FoP accounts for unique variance in youth HRQoL. Lastly, as a secondary exploratory analysis, we will calculate the proportion of parents with elevated levels of FoP.

We hypothesised that:

1. Higher levels of parental FoP will be associated with poorer parent and parent-reported youth psychosocial outcomes.

2. Greater parental FoP will independently predict worse parent-reported youth HRQoL over and above pain variables, parent-reported child anxiety and depression and parents’ psychopathology.

Methods

Participants

This cross-sectional study recruited parents of youth with JIA or T1DM through social media and patient organisations (Juvenile Arthritis Foundation Australia, Diabetes Australia). Recruitment took place from August 2024 until January 2025. Participants were eligible for the study provided they were a parent of a child (aged ≤18), who self-reported that their child had a diagnosis of JIA or T1DM and resided in Australia. Patients and the public were not involved in the development of this study.

A total of 182 participants accessed the study link. Of these, 160 consented to taking part in the study. Eight participants were excluded from the study due to their child being older than 18 years. One other participant was excluded as their child had a diagnosis of type 2 diabetes. Of the remaining participants, 27 did not complete any measures. This resulted in a final sample of 124 participants who completed at least one study measure and were included in the analyses. Of the 124, 108 completed the study in full (completion rate of 87%). Consequently, sample sizes varied across analyses due to missing data on some variables. See figure 1 for the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) participant flow diagram.

Figure 1. STROBE participant flow diagram. JIA, juvenile idiopathic arthritis; STROBE, Strengthening the Reporting of Observational Studies in Epidemiology; T1DM, type 1 diabetes mellitus.

Procedure

Participants accessed the study via Qualtrics, an online survey platform. They were provided with a participant information statement and consent form to read. After obtaining informed consent, participants then completed a series of questionnaires. The questionnaires collected demographic information about themselves as well as their child (age, gender, ethnicity, income, employment, education, sick leave taken in last 12 months). Characteristics about their child’s condition were also collected (type of JIA/diabetes, disease duration, delay to diagnosis, medications, number of hospitalisations, missed days of school and comorbidities). Parents then completed measures asking them about their child’s pain, HRQoL, anxiety and depression and about their own anxiety, depression, stress and FoP.

Measures

Parent FoP

An adapted version of the Worries about Recurrence or Progression Scale (WARPS)³³ was used to measure FoP in parents (see online supplemental material for the adapted measure). The WARPS is an 18-item scale which assesses concerns about the progression of illness. Participants responded to each statement on a 5-point Likert scale from 1 (Strongly disagree) to 5 (Strongly agree). The original wording of the WARPS was adjusted to assess parents’ perspectives on their child’s disease progression, as has been done previously to adapt the Fear of Progression Short Form³⁴ for use with parents of children with cancer.³⁵ For example, items that in the original WARPS referred to ‘my illness’ were changed to ‘my child’s illness’. No cognitive pretesting was undertaken as the items closely aligned to those used in the original WARPS validation study and similar wording has been used previously with other carer groups. Higher scores on the WARPS suggest greater concerns about disease progression. A clinical cut-off score of greater than 65 was previously established in a validation study in adults diagnosed with cancer.³⁶ In this study, the cut-off was developed from two previously gold standard measures with validated cut-offs for patients with cancer. In both analyses, the cut-off of 66 and above was optimal, according to receiver operating curve analyses. As such, for our secondary analyses, parents scoring 66 and above were classified as having elevated FoP. The WARPS demonstrated good psychometric properties in adults across several chronic illnesses.³³ The confirmatory factor analysis confirmed a single-factor explanation, with factor loadings showing that all items loaded significantly onto a single latent factor (standardised loadings=0.51–0.80, ps<0.001). Internal consistency was excellent with a Cronbach’s alpha of 0.95.

Parent-reported child pain intensity

A single item numeric scale was administered to parents to assess their estimate of their child’s pain intensity over the last week. Responses range from 0 (No pain) to 10 (Worst pain you can think of).

Parent-reported child pain interference

The pain interference subscale of the Paediatric Patient-Reported Outcomes Measurement Information System 25 Proxy Scale³⁷ was administered to parents. The 8-item scale asks parents the frequency with which pain interferes in their child’s daily activities. Responses range from 1 (Never) to 5 (Almost always). The scale has been well validated among parents of youth with chronic illnesses.³⁷ A Cronbach’s alpha of 0.96 was found in this study.

Parent-reported child HRQoL

The Pediatric Quality of Life Inventory Proxy Report (PedsQL)³⁷ is a 23-item measure of HRQoL. Parents were asked to indicate how much their child encountered problems in their physical, emotional, social and school functioning in the past month. Items are responded to using a 5-point Likert scale, ranging from 0 (Never a problem) to 4 (Almost always a problem). Items were then reverse coded and linearly transformed to a 0–100 scale, where higher scores are indicative of better HRQoL. This inventory has been well-validated among parents of children with chronic illnesses.³⁸ In this study, the PedsQL returned a Cronbach’s alpha of 0.96.

Parent-reported child anxiety and depression

The Revised Children’s Anxiety and Depression Scale 25—Parent Version (RCADS-25)³⁹ was completed by parents to assess anxiety and depression experienced by their child. The RCADS-25 consists of 25 items which contribute to two anxiety and depression subscales. Responses are given on a 4-point Likert scale where 0 (Never) and 3 (Always). Scores on the RCADS subscales were then summed and converted into T-scores. Research has found good psychometric properties for the RCADS-25 parent version.³⁹ A Cronbach’s alpha of 0.94 was found in this study.

Parent depression, anxiety and stress

The Depression Anxiety Stress Scales-21 (DASS-21)⁴⁰ was administered to parents to assess their depression, anxiety and stress symptoms. The DASS-21 requires the participant to indicate how much each statement applied to them on a 4-point Likert scale ranging from 0 (Did not apply to me at all) to 3 (Applied to me very much or most of the time). The DASS-21 has been extensively validated in adult populations.^{40 41} In the present study, a Cronbach’s alpha of 0.96 was returned, indicating good reliability.

Analysis plan

All analyses were undertaken using SPSS (V.29). Missing data were handled using pairwise deletion for correlation analyses and listwise deletion for all other analyses. As a result, sample sizes differed slightly across analyses depending on data availability.

Bivariate correlations were completed to investigate the relationship between FoP and other important parent and parent-reported youth variables in the total sample. Following this, a hierarchical regression analysis was completed with parent-reported child total HRQoL as the dependent variable. Demographic information (child and parent age, child and parent gender and diagnosis) was entered in the first step. In the second step, parent-reported child pain intensity and pain interference, as well as total RCADS-25 and DASS-21 scores were entered. Total RCADS-25 and DASS-21 scores were used to avoid multicollinearity and ensure the regression was sufficiently powered. Parent FoP was then entered in the last step to assess its unique contribution to parent-reported child HRQoL.

Several post hoc sensitivity analyses were also conducted for the regression model. First, the model was repeated with the number of hospitalisations reported by parents entered in the first step as a proxy measure of disease severity. Next, the model was repeated with the addition of a diagnosis x parental FoP interaction term in the final step, to determine whether the association between youth HRQoL and parental FoP differed according to illness group. Finally, the regression was repeated excluding the RCADS-25 and DASS-21 to examine the contribution of parental FoP without controlling for broader parent and child psychological symptoms.

Lastly, as secondary and exploratory analysis, the proportion of parents with elevated levels of FoP was calculated using the clinical cut-off previously validated in adults with cancer (scores of ≥66).³⁶

Results

Demographics and illness variables

124 parents (107 mothers; 17 fathers) participated in the study, 57 of whom had a child with JIA and 67 with a child with T1DM (see table 1). Parents were aged 42 years on average (SD=6.64), with no significant difference in age between the groups, p=0.361. However, children in the JIA group were significantly older (M=12.19 years, SD=3.91) than in T1DM (M=10.66 years, SD=3.75), p=0.028, d=0.40. As such, analyses of covariance were conducted to compare the two groups, controlling for child age. Parents who completed the study were primarily female (86.3%), as were their children (59.7%). Most parents described their ethnicity as being white/European (92.7%) and university educated (67%). 24 youth with JIA (42.1%) and 20 (29.4%) youth with T1DM were reported to have another health condition. The most frequently reported comorbid conditions were attention deficit hyperactivity disorder (n=12), asthma (n=12), coeliac disease (n=8) and autism spectrum disorder (n=7). Illness-specific characteristics are reported in online supplemental table S1.

Table 1. Demographic characteristics of the JIA and T1DM group.

Variable	JIA (n=57) Mean (SD)	T1DM (n=67) Mean (SD)
Parent age	42.80 (6.66)	41.70 (6.62)
Child age	12.19 (3.91)	10.66 (3.75)
Parent gender
Female (%)	50 (87.7)	57 (85.1)
Male (%)	7 (12.3)	10 (14.9)
Child gender
Female (%)	37 (64.9)	37 (55.2)
Male (%)	19 (33.3)	30 (44.8)
Non-binary (%)	1 (1.8)	0 (0)
Area of residence
Major city (%)	33 (57.9)	36 (53.7)
Regional (%)	24 (42.1)	29 (43.3)
Remote (%)	0 (0)	2 (2.9)
Ethnicity
White/European (%)	53 (93)	62 (92.5)
Aboriginal/Torres	1 (1.8)	1 (1.5)
Strait Islander (%)
Black/African American (%)	1 (1.8)	1 (1.5)
Other (%)	2 (3.5)	3 (4.5)
Household income
<$A30 000	1 (1.8%)	3 (4.5%)
$A30 000–49 999	0 (0%)	0 (0%)
$A50 000–79 999	4 (7%)	3 (4.5%)
$A80 000–99 999	5 (8.8%)	6 (9%)
$A100 000–149 999	20 (35.1%)	26 (38.8%)
$A150 000–199999	13 (22.8%)	14 (20.9%)
>$A200 000	14 (24.6%)	15 (22.4%)
Education
Did not complete high school	0 (0%)	1 (1.5%)
High school	5 (8.8%)	3 (4.5%)
Trade certificate	14 (24.6%)	18 (26.9%)
Undergraduate degree	23 (40.4%)	29 (43.3%)
Postgraduate degree	15 (26.3%)	16 (23.9%)
Employment
Unemployed (%)	0 (0)	0 (0)
Casual (%)	3 (5.3)	4 (6)
Part-time (%)	25 (43.9)	23 (34.3)
Full-time (%)	26 (45.6)	35 (52.2)
Homemaker (%)	3 (5.3)	5 (7.5)
Additional health condition (%)	24 (42.1)	20 (29.4)

JIA, juvenile idiopathic arthritis; T1DM, type 1 diabetes mellitus.

Correlational analyses

The results of the correlational analyses are presented in table 2. Parental FoP shared small to large correlations with the DASS-21, RCADS-25 and PedsQL. However, pain intensity was not significantly associated with parent FoP in either illness group. The PedsQL demonstrated moderate to large negative associations with all psychosocial outcomes across the groups.

Table 2. Pearson correlations between parental FoP and psychosocial variables across total sample.

	WARPS	Pain intensity	Pain interference	DASS-21	RCADS-25	PedsQL
WARPS	–	0.16	0.28**	0.41**	0.33**	−0.43**
Pain intensity		–	0.32**	0.17	0.27**	−0.30**
Pain interference			–	0.11	0.44**	−0.58**
DASS-21				–	0.53**	−0.41**
RCADS-25					–	−0.72**
PedsQL						–

Sample sizes varied across correlations depending on data availability (n=110–120).

*p<0.05, **p<0.01.

DASS-21, Depression Anxiety Stress Scales-21; FoP, fear of disease progression; PedsQL, Pediatric Quality of Life Scale; RCADS-25, Revised Child Anxiety and Depression Scale-25; WARPS, Worries about Recurrence and Progression Scale.

Hierarchical linear regression analysis

A hierarchical regression analysis was conducted to assess the unique contribution of parental FoP to total youth HRQoL. Demographic variables (both youth and parent age and gender, and illness group) were entered first and the model was significant, accounting for 15% of variance in HRQoL, F(5,102)=3.59, p=0.005 (see table 3). Parent gender (β=0.23, p=0.016) and youth diagnosis (β=0.24, p=0.012) were significant predictors of HRQoL.

Table 3. Hierarchical regression analysis predicting total HRQoL.

Variable	Step 1	Step 2	Step 3
	β (SE)
Parent age	0.12 (0.41) p=0.309	0.02 (0.29) p=0.863	0.04 (0.29) p=0.655
Parent gender	0.23 (5.94) p=0.016*	0.00 (4.41) p=0.976	0.03 (4.42) p=0.681
Child age	−0.18 (0.73) p=0.130	−0.01 (0.51) p=0.918	−0.02 (0.51) p=0.829
Child gender	0.08 (4.49) p=0.403	0.02 (3.13) p=0.788	0.02 (3.08) p=0.787
Diagnosis	0.24 (4.37) p=0.012*	0.16 (3.54) p=0.040*	0.12 (3.58) p=0.113
RCADS-25	–	−0.53 (0.10) p<0.001***	−0.51 (0.10) p≤0.001***
DASS-21	–	−0.11 (0.07) p=0.192	−0.03 (0.08) p=0.710
Pain intensity	–	−0.05 (0.48) p=0.466	−0.05 (0.48) p=0.502
Pain interference	–	−0.24 (0.19) p=0.004**	−0.23 (0.19) p=0.005**
WARPS			−0.15 (0.14) p=0.042*
R2	0.150 p=0.005	0.612 p<0.001	0.629 p<0.001
R2$∆$	0.150 p=0.005	0.463 p<0.001	0.016 p=0.042

Analyses were based on the 108 participants with complete data for all variables included in the regression model.

*p<0.05, **p<0.01, ***p<0.001.

DASS-21, Depression Anxiety Stress Scales-21; HRQoL, health-related quality of life; RCADS-25, Revised Child Anxiety and Depression Scale-25; WARPS, Worries about Recurrence and Progression Scale.

Adding in total RCADS-25 scores, total DASS-21 scores, pain intensity and pain interference explained an additional 46.3% of variance in HRQoL. This was a significant contribution, F(9,98)=29.26, p<0.001. After controlling for these variables, diagnosis remained a significant predictor (β=0.16, p=0.040) alongside total RCADS-25 scores (β=−0.53, p<0.001) and pain interference (β=−0.24, p=0.004).

In the final step, parent FoP contributed an additional 1.6% of variance in HRQoL, which was a small but significant contribution, F(1,97)=4.25, p=0.042. Following this, diagnosis no longer significantly predicted HRQoL (p=0.113). However, total RCADS-25 scores (β=−0.51, p<0.001) and pain interference remained significant predictors (β=−0.23, p=0.005). Parental FoP was also found to significantly predict HRQoL (β=−0.15, p=0.042).

The same hierarchical regression was repeated post hoc with the addition of the number of hospitalisations to the first step to investigate whether the association between parental FoP and youth HRQoL persisted even after controlling for a proxy measure of disease severity. In this model, parental FoP explained an additional 1.4% of variance in HRQoL, which was a significant increase (F(1,96)=3.95, p=0.050). Parental FoP was also associated with youth HRQoL (β=−0.14, p=0.05), as number of hospitalisations remained a significant predictor in the final step of the model (β=−0.20, p=0.002).

An additional sensitivity analysis was completed to determine whether the association between parental FoP and youth HRQoL differed according to illness group. To do so, the same hierarchical regression was conducted post hoc, with the addition of an interaction term (FoP × diagnosis) in the final step. The addition of this interaction variable did not explain significantly more variance in HRQoL (F(1,96)=0.39, p=0.536), and the interaction was not a significant predictor (β=0.080, p=0.536), suggesting that the association between FoP and youth HRQoL did not differ according to illness group.

Lastly, the hierarchical regression analysis was repeated excluding the RCADS-25 and DASS-21 to identify the contribution of parental FoP without controlling for these variables. In the final step of the regression, parental FoP explained an additional 7.1% of variance in youth HRQoL, which was a significant increase (F(1,100)=13.33, p<0.001). Parental FoP was also a significant independent predictor of youth HRQoL in this final step (β=−0.28, p<0.001), alongside pain interference (β=−0.45, p<0.001) and parent gender (β=0.17, p=0.024).

As a secondary exploratory analysis, 76% (n=84) of the 111 parents who completed the WARPS scored in the elevated range based on the clinical cut-off validated in adults with a cancer diagnosis.³⁶

Discussion

This study examined the relevance and impact of FoP among parents of youth with JIA or T1DM. In line with our predictions, parental FoP was consistently associated with poorer parent and parent-reported youth outcomes for both youth with JIA and T1DM. Importantly, parental FoP was also found to explain a small but significant amount of the unique variance in parent-reported youth HRQoL. Post hoc sensitivity analyses also indicated that this association was consistent across both illness groups. Our secondary exploratory analyses revealed that more than three quarters of parents scored above the cut-off for clinically significant FoP derived from adults with cancer.³⁶ Cumulatively, these findings emphasise the significance of FoP to parents of youth with JIA and T1DM, and its potential psychosocial consequences for parents and their children.

As hypothesised, parental FoP was consistently related to poorer psychosocial outcomes. This includes greater parental distress as well as more parent-reported youth pain, anxiety, depression and poorer HRQoL. These findings are congruent with research which has supported the relationship between parent FoP and increased depression and anxiety in parents.^{11 12} They also align with studies which have shown an association between FoP and both pain and psychological distress in adults with rheumatoid arthritis⁴² and diabetes.⁴³ Additionally, these findings are consistent with meta-analytical research that has found strong associations between caregiver FoP, anxiety and depression.¹⁰ However, expanding on this, we also demonstrated that parents with greater levels of FoP have children for whom they report poorer outcomes. Unfortunately, due to the cross-sectional nature of this study, it is unclear whether parental FoP contributes to poorer outcomes in youth, or if parental FoP is simply higher in parents of children with more impaired psychosocial functioning. Future research is needed to clarify the direction of this relationship. Nonetheless, the fact that our results demonstrated that higher parental FoP was associated with poorer parent and parent-reported youth psychosocial outcomes highlights the need for services to place a greater emphasis on supporting parents and addressing their concerns for their children’s future. This includes implementing regular screening measures to identify parents who have high levels of FoP, so they can be supported. It may also be beneficial to provide parents with some brief resources or information sessions about FoP and coping with the uncertainty of these two childhood illnesses.

Importantly, parental FoP was found to account for a small, yet significant amount of additional variance in parent-reported youth HRQoL over these covariates. This finding is consistent with others who found that parent FoP predicted worse HRQoL in child³² and adult patients with cancer.³¹ As parental FoP only contributed a small amount of additional variance (1.6%), this finding should be interpreted cautiously. Nevertheless, this was after accounting for diagnosis, pain variables, total DASS-21 and RCADS-25 scores, which were each found to strongly correlate with HRQoL. As such, parental FoP may capture variance in HRQoL not otherwise explained by general internalising symptoms of the child or parent psychopathology. However, as the PedsQL also assesses psychosocial functioning, it may overlap conceptually with the RCADS-25. Hence, the strong association between these two constructs may at least partially be the product of this conceptual overlap. This possibility should be borne in mind when considering the contribution of parental FoP to youth HRQoL over and above the RCADS-25, particularly given the small amount of additional variance explained by parental FoP. In support of this, our post hoc analyses demonstrated that parental FoP accounted for substantially greater unique variance in youth HRQoL (7.1%) after removing the RCADS-25 and DASS-21 from the model. This suggests that the modest contribution of parental FoP in the primary regression model may reflect the shared variance between parental FoP, these psychological variables and HRQoL.

It is also worth noting that our regression data were entirely derived from parent report and are thus vulnerable to the influence of common-method and negative-affectivity bias. As such, one potential explanation for these findings could be that parents who are highly fearful of their child’s disease progression may also perceive their child as having more impaired HRQoL. Additional follow-up studies are needed to confirm these findings using parental fear of progression with youth self-report data to minimise the influence of shared-rater bias.

Furthermore, although we controlled for demographics, pain and psychosocial variables in our regression model, we did not include a measure of disease severity or activity. Disease severity is likely to be associated with youth HRQoL and may similarly contribute to parents’ levels of FoP. Parental FoP nevertheless continued to account for unique variance in youth HRQoL during our sensitivity analyses when we also controlled for hospitalisations as an indicator of disease severity. Likewise, the fact that the association between FoP and parent-reported youth HRQoL was not found to differ in strength between the illness groups suggests that this relationship is consistent across the groups. Although the association between parental FoP and HRQoL was similar when controlling for hospitalisations, we recognise that hospitalisations cannot be considered a direct or objective measure of disease severity. Unlike our self-report measures, objective indicators of disease severity could be one way to obtain data that is not subject to the influence of parents’ perceptions of their child’s health. Thus, collecting and controlling for objective indicators of disease severity in future studies could help further clarify the unique contribution of parental FoP to youth HRQoL.

Our secondary exploratory analyses also suggest that FoP may be highly relevant to both parent groups. The scores that parents reported on this measure indicated that they were very frequently worried about the future of their children’s illness. Indeed, if one were to use the clinical cut-off recommended and validated in adults with cancer, 76% of parents would have scored above this threshold. While this cut-off has not been validated in parents of children with chronic illness, it nevertheless suggests that parents are endorsing items known to be associated with severe FoP in other illnesses, confirming that fear of progression is worthy of further investigation. In many ways, this high level of FoP is understandable considering the risks of serious long-term health complications in both JIA and T1DM. However, our sample might have been particularly high in FoP concerns considering our method of recruitment. As recruitment was completed through patient organisation groups and social media, this may have resulted in a sample of more worried and engaged parents. Considering this, it could be that the proportion of parents with elevated FoP in the present study is higher than what would be found in the broader community. Nevertheless, these findings suggest that FoP is worthy of further investigation in parents. In particular, future research is needed to identify the specific factors underlying these concerns. This will also allow for greater clarity on how to best support these parents.

Despite the strength of this study, it featured some limitations. First, our study included a relatively small sample of participants, and it is possible that as a result it was underpowered to detect small differences between groups or the relationships between FoP and other variables. As social media was used for recruitment, the parents included in our study may be a more engaged or concerned cohort, which may not be representative. The generalisability of our findings may also be limited by the predominantly female, white and university educated sample of parents, meaning it is unclear whether these findings will generalise to fathers, more diverse or less educated populations. Furthermore, parents’ self-report of their child’s diagnosis was used to determine eligibility for this study. While concordance between self-report diagnosis and health records has been found to be good in individuals with diabetes,⁴⁴ there is little data on the accuracy of parent-report for JIA or diabetes. As such, future studies should verify participants’ diagnoses to increase confidence in their eligibility. The use of self-report also applies to other medical and demographic data, such as time missed from school due to illness, which could be subject to recall bias. Moreover, as discussed prior, this study was limited by the reliance on proxy reports for youth outcomes. Although this allowed us to include parents of children too young to self-report, the validity of proxy report measures has been criticised.⁴⁵ As such, we recommend that future research prioritises collecting self-report data. Finally, although the aim of this study was not to validate the parent version of the WARPS, we nonetheless conducted preliminary analyses to ensure it was appropriate to use in the present study. The Cronbach’s alpha (α=0.95) was excellent and the factor structure identified across a range of other illnesses was confirmed in this sample. However, further psychometric validation of this measure is required. Lastly, as mentioned previously, the cross-sectional design of this study means the directionality of findings is unclear. Future research should endeavour to collect follow-up data to clarify this.

Conclusions

This study was the first to investigate parental FoP in JIA and T1DM. Parental FoP was consistently associated with poorer parent and parent-reported youth psychosocial outcomes across both illness groups. This included more anxiety and depressive symptoms in parents and poorer parent-reported youth pain, anxiety, depressive symptoms and HRQoL. Additionally, parent FoP also explained a small but significant amount of additional variance (1.6%) in parent-reported youth HRQoL, after controlling for known covariates. Our secondary exploratory analyses also suggested that elevated FoP may be a common concern among parents of youth with JIA and T1DM. Despite the apparent relevance of FoP to parents of this study and its association with poorer outcomes, very little research has focused on this in the cancer and chronic illness literature and no interventions exist to treat FoP among parents. Hence, these results suggest a clear need to establish effective supports and resources for parents who are distressed about their child’s possible future disease progression.

Data availability statement

Data are available upon reasonable request.