Study Identifies Factors Preceding Headache Remission in Adolescents
Being a boy, having lower depression and anxiety scores, lower headache intensity, lower affective pain perception, fewer pain locations, fewer sleep issues, and lower pain-related disability predicted headache remission.
Julia Wager, PhD
Published in Headache, a new study examined factors associated with frequent headache remission in schoolchildren aged 10-18 years. All told, results on a step-wise multiple regression model showed that sex, headache intensity, depression, and anxiety were influential in headache remission and explained most of the variance between groups in the study.1
From a large school sample (n = 2280), a subgroup of 281 schoolchildren were included in the analysis. Of these, 156 participants had remitted frequent headache (57.1% female; mean age, 13.1 years) and 125 had non-remitted frequent headache (78.4% female; mean age, 13.5 years). There were 42 participants who fulfilled the criteria for frequent headache remission but were excluded because of other frequent pain conditions at T4-T5.
Led by senior author Julia Wager, PhD, a psychologist and head of research department at the German Pediatric Pain Center, univariate binary logistic regressions showed that headache remission was less likely with higher anxiety scores (OR, 0.86; 95% CI, 0.80-0.93), more difficulties falling asleep (OR, 0.96; 95% CI, 0.94-0.99) or sleeping throughout the night (OR, 0.96; 95% CI, 0.94-0.99). Furthermore, headache remission was less likely with higher pain-related disability (OR, 0.96; 95% CI, 0.94-0.99(, a greater number of pain locations (OR, 0.70; 95% CI, 0.49-0.98), and higher affective pain perception (OR, 0.92; 95% CI, 0.87-0.97).
To the authors’ knowledge, this was the first study to analyze a broad spectrum of biopsychosocial factors toward headache remission in this age group. Despite girls being more prevalent, the odds of girls experiencing headache remission were 0.366 times lower compared with boys (95% CI, 0.215-0.62).
In the multiple logistic regression model, the variables predictive of headache remission in stepwise regressions in at least two-thirds of the 25 datasets were sex (OR, 0.43; 95% CI, 0.248-0.76; P = .003), headache intensity (OR, 0.85; 95% CI, 0.73-0.99; P = .035), anxiety score (OR, 0.92; 95% CI, 0.85-1.01; P = .071), and depression score (OR, 0.94; 95% CI, 0.89-1.00; P = .041). This model explained 17% of the variance in group membership. Despite explaining only a small proportion of variance, this proportion exceeded previous research by Carasco and Kroner-Herwig, which examined only psychological characteristics (9%).
"In our study, we were able to explain a greater amount of variance by including pain characteristics and additional psychological factors in our model compared to previous work," Wager et al wrote. "Our results underscored the challenge of predicting frequent headache remission in the general population, despite examining a wide range of biopsychosocial factors."
In terms of between-group differences, schoolchildren with remitted headache had lower depression (F[1557.01] = 45.77, p < 0.001) and anxiety scores (F[1557.01] = 21.72, p < 0.001), as well as higher school satisfaction (F[1209.46] = 7.15, p = 0.008) than those with non-remitted headache. In addition, this group demonstrated fewer difficulties falling asleep (F[1856.52] = 41.21; P <.001) or sleeping through the night (F[1731.12] = 26.42; P <.001).
While higher anxiety scores seem to be a comorbidity of ongoing frequent headache, the investigators wrote that the results suggest a potential interplay between depression scores and frequent headache remission. In the non-remitted headache group, depression scores were similar at T1 and T5 (MT1 = 11.3; SDT1 = 5.5; MT5: 11.6, SDT5: 6.2; p = 0.502) whereas depression scores declined significantly during this time for those who remitted (MT1 = 8.7, SDT1 = 4.5; MT5 = 6.7, SDT5 = 4.7; p < 0.001).
In addition to the relatively short duration of the study, the data was limited by the artificial dichotomization of frequent headache as present versus absent, which reduced statistical power. In contrast, the study was strengthened by the fact that it is the first to communicate how physiological, psychological, and social factors interact with pediatric headache during remission. While previous work has mainly focused on pain-associated factors through cross-sectional designs, this study contributes to bridging this knowledge gap regarding the sequence of pain and its associated factors.
"Future works could establish a meaningful, weighted continuous pain outcome by integrating variables such as pain duration, intensity, and frequency," Wager et al wrote. "Furthermore, the absence of information on medical or psychotherapeutic treatment impedes our ability to conclude whether remission was spontaneous or purposeful."
