Pressure pain sensitivity: A new stress measure in children and adolescents with type 1 diabetes?

Abstract

Type 1 diabetes (T1D) is associated with general- and diabetes-specific stress which has multiple adverse effects. Hence measuring stress is of great importance. An algometer measuring pressure pain sensitivity (PPS) has been shown to correlate to certain stress measures in adults. However, it has never been investigated in children and adolescents. The aim of our study was to examine associations between PPS and glycated hemoglobin (HbA1c), salivary cortisol and two questionnaires as well as to identify whether the algometer can be used as a clinical tool among children and adolescents with T1D. Eighty-three participants aged 6-18 years and diagnosed with T1D were included in this study with data from two study visits. Salivary cortisol, PPS and questionnaires were collected, measured, and answered on site. HbA1c was collected from medical files. We found correlations between PPS and HbA1c (rho = 0.35, P = 0.046), cortisol (rho = -0.25, P = 0.02) and Perceived Stress Scale (rho = -0.44, P = 0.02) in different subgroups based on age. Males scored higher in PPS than females (P < 0.001). We found PPS to be correlated to HbA1c but otherwise inconsistent in results. High PPS values indicated either measurement difficulties or hypersensibility towards pain.

Key Words: Stress, Children and adolescents, Type 1 diabetes, Autonomic dysfunction

Core Tip: The aim of present study was to examine whether pressure pain sensitivity (PPS) in children and adolescents associates with other stress measures and determine if it can be used as a clinical tool in this population. Our study revealed some unexpected discrepancies examining PPS in a pediatric population with type 1 diabetes, highlighting the need for more research to validate if PPS is a clinically useful measure in children.

TO THE EDITOR

Type 1 diabetes (T1D) is associated to general- and diabetes-specific stress, which is linked to high glycated hemoglobin (HbA1c)[1], increased morbidity and mortality, and decreased quality of life[2]. In adults, Ballegaard et al[3,4] found pressure pain sensitivity (PPS) measured with an algometer to be correlated with established stress measures, thereby introducing an objective, non-invasive method of measuring stress. The aim of this study was to examine how PPS in children and adolescents associates with other stress measures and whether it can be used as a clinical tool in this population.

Data were collected as part of a prospective study of dermatological complications to diabetes devices. The present study included individuals with T1D between the ages of 6-18 years, and two study visits for each participant were selected based on available data. Exclusion criterium was participant or caregiver not being able to speak or read Danish.

PPS was measured using the algometer with two to three consecutive measurements on the index finger and tibia for method introduction before placement on the sternum. Increasing pressure was put on the skin for three seconds and participants were asked to say “stop” just before or as the pressure turned into pain or discomfort. The measurement was also stopped when a noxious withdrawal reflex (NWR) was observed, or if there was an activation of an alarm at maximum pressure (= 30 on PPS scale). PPS-score ranges from 30-100 with values ≥ 60 being the cut-off point for high level of stress in adults, based on receiver operating characteristic curves[4,5]. Salivary cortisol was analyzed with radioimmunoassay and HbA1c collected from medical files. Perceived Stress Scale (PSS) and World Health Organization-5 Well-Being Index (WHO-5), two questionnaires regarding stress and well-being, were completed during visits. PSS was answered by participants from age 10 years and WHO-5 from age 6 years. Examinations were conducted between April 1, 2020 and April 9, 2021. The study was approved by the Danish Data Protection Agency (P-2020-2) and the Regional Committee in Health Research Ethics (H-18059790) and followed Danish legislation regarding consent. Statistical analyses were made using the statistical software package R, version 4.2.2. Spearman’s rank correlation was used for correlation analyses. Sex differences were analyzed using Mann-Whitney U test. The study population was analyzed as a whole and divided in two age groups (6-12 years, 13-18 years). A P value of < 0.05 was considered statistically significant.

This study comprised 83 participants, 51% male, mean (± SD) age was 12.6 (± 2.9) years and median (Q1-Q3) T1D duration was 0.8 years (0.01; 3.4). Forty-one percent were diagnosed with T1D within three months prior to their first study visit. Tables 1 and 2 show correlation analyses at the first visit (Table 1) and in between visits (Table 2) for the whole population as well as when stratified into age groups. Unexpectedly, negative correlations were found between PPS and cortisol in the total population (rho = -0.25, P = 0.02), in the 13 to 18-year-olds (rho = -0.35, P = 0.045), and between PPS and PSS in the 6 to 12-year-olds (rho = -0.44, P = 0.02). A positive correlation between PPS and HbA1c was present in the 13 to 18-year-olds (rho = 0.35, P = 0.046) and this finding persisted when comparing differences in PPS and HbA1c between the two visits (rho = 0.45, P = 0.048). Males scored higher than females in PPS in the total population (median difference = 17.5, P < 0.001), this being driven by the younger age group (P < 0.001) since no sex difference was present among 13 to 18-year-olds. No sex differences were found among the other variables. The PPS measurements were strongly and internally correlated when measured on the index finger, the tibia, and the sternum (all r > 0.5, all P < 0.001).

Table 1 Pressure pain sensitivity and correlations to stress measures during the first study visit.

	Age: 6-12 yr; (M = 24, F = 25)		Age: 13-18 yr; (M = 18, F = 16)		Total population; (M = 42, F = 41)
	Median (Q1; Q3), N (missing)	Rho (P value)	Median (Q1; Q3), N (missing)	Rho (P value)	Median (Q1; Q3), N (missing)	Rho (P value)
Pressure pain sensitivity1	85 (68; 99), 0		76 (66; 91), 0		81 (66; 96), 0
Perceived stress scale2	12 (8; 16), 23	-0.44 (0.02)3	16 (10; 18), 9	0.16 (0.45)	15 (9; 18), 32	-0.18 (0.20)
WHO-5	72 (64; 84), 7	0.29 (0.06)	62 (46; 76), 4	-0.21 (0.26)	68 (59; 80), 11	0.14 (0.23)
Salivary cortisol (nmol/L)	6.8 (5.8; 8.5), 0	-0.13 (0.38)	9.4 (5.6; 15.3), 0	-0.35 (0.045)3	7.4 (5.7; 10.4), 0	-0.25 (0.02)3
HbA1c (mmol/mol)	63 (53; 8), 0	-0.27 (0.06)	67 (59; 72), 1	0.35 (0.046)3	66 (53; 80), 1	-0.07 (0.51)

¹Pressure pain sensitivity score is an average of two to three measurements at the sternum. Third measurement was performed if a difference > 10 was present between the two first measurements. Distribution of cause of termination: First measurement: 18 with noxious withdrawal reflex, 65 said “stop”; second measurement: 15 with NWR, 66 said “stop”, 1 alarm; third measurement: 4 with NWR, 2 said “stop”.

²Answered by children above age 10 yr.

³Indicates a significant correlation (Spearman; P < 0.05).

M: Male; F: Female; Q1; Q3: First and third quartile; HbA1c: Glycated hemoglobin; WHO-5: World Health Organization-5 Well-Being Index.

Table 2 Within-variable differences between the two study visits and the correlations between changes in Pressure pain sensitivity and changes in Perceived stress scale, World Health Organization-5 Well-Being Index, salivary cortisol and glycated hemoglobin respectively.

	Age: 6-12 yr; (M = 24, F = 25)		Age: 13-18 yr; (M = 18, F = 16)		Total population; (M = 42, F = 41)
	Median (Q1; Q3), N (missing)	Rho (P value)	Median (Q1; Q3), N (missing)	Rho (P value)	Median (Q1; Q3), N (missing)	Rho (P value)
Pressure pain sensitivity1	4.5 (0; 13), 0		3 (-2.9; 8.4), 0		3.5 (-0.8; 11.2), 0
Perceived stress scale2	-2 (-7; 0), 24	-0.08 (0.70)	1.5 (-2.3; 4.3), 10	0.06 (0.77)	-1 (-5; 3), 34	-0.04 (0.80)
WHO-5	4 (-8; 12), 8	-0.08 (0.63)	0 (-12; 8), 5	0.34 (0.068)	0 (-8; 11), 13	0.10 (0.43)
Salivary cortisol (nmol/L)	-0.07 (-0.38; 1.3)	-0.14 (0.33)	-0.2 (-3.1; 5.4), 0	0.25 (0.16)	-0.07 (-3.5; 0.75), 0	0.02 (0.87)
HbA1c (mmol/mol)3	0 (-5.5; 2.5), 4	0.05 (0.82)	0 (-5; 2.3), 2	0.45 (0.048)4	0 (-5; 2.5), 6	0.19 (0.22)

¹Pressure pain sensitivity score is an average of two to three measurements at the sternum. Third measurement was performed if a difference > 10 was present between the two first measurements. Distribution of cause of termination: First measurement: 18 with noxious withdrawal reflex, 65 said “stop”; second measurement: 15 with NWR, 66 said “stop”, 1 alarm; third measurement: 4 with NWR, 2 said “stop”.

²Answered by children above age 10 yr.

³Participants newly diagnosed with Type 1 Diabetes (duration < 3 months) were excluded from analyses (n = 49 participants remained, n = 27 aged 6-12 yr, n = 22 aged 13-18 yr).

⁴Indicates a significant correlation (Spearman; P < 0.05).

M: Male; F: Female; Q1; Q3: First and third quartile; HbA1c: Glycated hemoglobin; WHO-5: World Health Organization-5 Well-Being Index.

To summarize, we found a positive correlation between PPS and HbA1c in the old age group and unexpectedly negative correlations at the first visit between PPS and cortisol and PPS and PSS, the latter being present in the young age group. The negative and missing correlations contrast the findings of Ballegaard et al[3] who in adults found significant correlations between PPS and physiological markers of stress (heart rate, blood pressure, pressure rate product), regulation of glucose metabolism in adults with type 2 diabetes[5-7], survival in persons with ischemic heart disease[8] and questionnaires regarding mental and physical health[4]. Interestingly, we found that the younger males scored higher than the younger females but no differences among the older participants. Conversely Ballegaard et al[3] found the opposite correlation in adults which is more in line with the general assumption that females score higher in stress than males[9]. When using the algometer, a sex specific scale is activated. Hence our opposing findings may be explained by an irrelevance of different scales for sex in younger children. Despite the set-up with measuring on finger and tibia first and then sternum, a high percentage of participants said stop right away even at repeated visits. The a priori fear of being hurt and/or problems with understanding the instruction may be more prominent in younger children. A possible scenario could also be that participants wished to appear strong and therefore waited too long before saying stop. The population’s high PPS levels can be interpreted as either an expression of a psychological response, measurement difficulties or a general centrally induced hypersensibility caused by autonomic imbalance due to T1D[7]. PPS was significantly lower in participants with NWR compared to participants saying “stop” but too few measurements terminated because of NWR were available to enable subgroup analyses. The study was running during the COVID-19 pandemic which might have influenced stress and well-being. Cortisol levels were not adjusted to potentially influencing factors such as exercise and food intake. The study had minimal selection bias since participants enrolled in a study regarding skin problems. Furthermore, the variation in age allowed subgroup analyses.

In conclusion, there was a moderate to strong internal correlation between PPS measured on the three locations, however, the correlations of PPS to other indicators of stress such as cortisol and PSS was unexpectedly negative. PPS-values were generally high compared to adults reflecting either measurement difficulties or hypersensibility towards pain and the use of sex-specific scale was less relevant in the youngest age group. Our study revealed some unexpected discrepancies examining PPS in a pediatric population with T1D highlighting the need for more research to validate if PPS is a clinically useful measure in children.

ACKNOWLEDGEMENTS

We would like to thank all the children and parents who participated in the project and made this research possible. We would also like to thank the nurses and doctors at both hospitals for their assistance in recruiting patients for the study. We would also like to thank Olivia McCarthy for language editing.