Original Article

Longitudinal Modeling of Non-Pharmacological Factors Related to Frequency, Severity and Duration in Both Migraine and Tension-Type Headaches

Introduction

The efficacy of interventions in a headache is often measured based on the patient’s estimates of three important parameters including frequency, severity, and attack duration ¹ . Frequency, duration, and severity of headaches, as well as associated symptoms, differ among patients and from one attack to another ² . Severe headaches are worrying, and frequent. On the other hand, long headache attacks cause problems, even if they are not severe in terms of intensity. The above-mentioned three parameters play an important role in diagnosing headache type, evaluation of the effectiveness of medication, comparison of different treatments and determination of the factors affecting the incidence of recurrent headaches ^3–6 . Thus, the factors that affect the attacks are those that can reduce or increase these parameters. There are a great body of studies in the literature that show socio-economic factors such as low income ^5,7 , poor education ^5,8–10 , and employment status ^9–11 are associated with the sense of pain. However, such risk factors are not significant in all studies ¹² . In addition, some other factors such as headache-associated symptoms may affect the frequency, severity and length of headaches the ignorance of which in headache research may lead to incorrect conclusions. Therefore, the studies related to headache should be performed by considering all these factors.

In the present study, we model frequency, severity and duration of attacks to discover the socio-demographic factors and headache-associated symptoms. Through such an approach to response modeling, the effects of other variables are adjusted, but not ignored. Moreover, higher diagnostic power is achieved in discovering the effective factors in the responses. Although previous studies have taken some of these factors into account in relation to frequency, severity, or duration of headaches, all of these associations have not been simultaneously investigated yet.

Methods

In this longitudinal study, 275 migraine or TTH patients were included through easy sampling method with regular patients during their usual clinical follow-ups in 2012 at four neurological clinics in Isfahan, Iran. The subjects were diagnosed based on the International Headache Society [IHS] criteria ¹³ . Informed consent was obtained from all participants before the investigation. All of them were treated with at least one prophylactic (including SSRIs, TCAs, Beta/Blockers and Anti-epileptics) and one analgesic drug (including NSAIDs, Acetaminophen, -Triptans, Caffeine, low-dose Codeine) drug. The headache characteristics were recorded on the first day [visit 1], third week [visit 2], and eighth week [visit 3] after the enrollment. The conditions of patients were stable in all three visits; hence, there was no need for any modification to the drug type or dosage. The stability was defined based on the neurologists’ opinions through on prepared forms and filled questioners in all visits. The patients who needed any changes in the type or dosage of drugs were excluded from the rest of the study.

The socio-demographic characteristics of all subjects including: age, sex, marital status (married/single), job status, working hours (per week), family history of headache (yes/no), length of headache history (years), place of living (rural/urban), and educational level (primary school, secondary school, diploma, and bachelor’s and above), as well as headache symptoms including photophobia and phonophobia (yes/no), lateralization (unilateral/bilateral) and pulsatile quality (yes/no), nausea (yes/no), vomiting (yes/no) and type of headache onset (suddenly/gradually) were asked in the first visit (baseline). Headache characteristics were frequency, severity, and duration of attacks. The averages of frequency and severity (in the last month by Persian MIDAS questionnaire) and attack duration (less than 12 h, between12 and 24 h, between 24 and 48 h and, over 48 h) were recorded in all of the three visits; the participants used at least one analgesic drug. Migraine Disability Assessment (MIDAS) questionnaire is a valid and reliable short questionnaire for the assessment of headache-related disability. It also includes two more questions about the frequency and severity of attacks in the last three months in the Persian version, the reliability and validity of confirmed by Zandifar et al. ¹⁴ . In the MIDAS, patients described the severity of headaches on a scale from zero to 10 with zero describing no existence of pain and 10 indicating the worst pain they had ever experienced.

Statistical analysis

Continuous variables were described by mean and standard error while categorical variables were described by frequency and percentage. Three considered dimensions of headache were frequency, severity, and attack duration. A set of demographic, social, and headache-associated symptoms were considered as the covariates. One of the responses was related to headache frequency in the last three months. For modeling this response, we used a multivariable negative binomial regression for panel data with random effects specification. Overdispersion in data was tested with a likelihood ratio test based on comparison of Poisson and Negative binomial distributions. This test assesses the equality of the mean and variance imposed by the Poisson distribution against the alternatives in which variance exceeds the mean ¹⁵ .

Other responses including headache severity and attack duration were ordinal variables. An ordered logit regression with random effect was used for modeling severity and because most headache attacks lasted under 12 h, random-effect Zero-Inflated Ordered Probit (ZIOP) model was used for modeling the duration of attacks in three-time intervals. Similarly, all of the above covariates were used in all of the models. Data analysis was performed using R Statistical Software, Version 3.4.3 “gamlss.mx” and “mixor” packages.

Results

Overall, 275 patients aged 13 to 59 participated in the present study. They consisted of 210 (76.4%) migraine patients and 65 (23.6%) TTH patients, among whom 26% were men and 74% women. In all three visits, patients were asked about the frequency, severity and duration of headache attacks during the last month. Due to the possible differences of these factors among the migraine and TTH patients, the dependent variables were defined based on headache type, and as reported in Table 1.

Description and comparison of socio-demographic factors and headache-associated symptoms in migraine and TTH patients are presented in Table 2. The migraine patients in our study had a longer headache history than TTH patients (P<0.001). On the other hand, the unemployed patients (P=0.021) and patients with family history of headache (P=0.049) were more likely to experience TTH. As expected, according to the IHS classification of headaches ¹³ , unilaterality, pulsatility, nausea, vomiting, photophobia, and phonophobia were mostly associated with migraine headaches (P<0.001). There was no significant difference in the types of onset (gradually or suddenly) between migraine and TTH attacks.

The relationship among the triad of frequency, severity and duration of headache attacks as well as the socio-demographic factors is reported in Table 3. People with TTH had more frequent (P<0.001) and less severe (P=0.004) headaches than migraine patients and less suffered from attacks over 12 hour (P=0.048). Older people had less frequent attacks (P<0.001), but age was not related to the severity and duration of attacks. No difference was observed in the frequency and severity of headaches in males and females, but female patients experienced attacks over 12 h (P=0.035) more. No significant relation was found between marital status and frequency as well as severity of attacks, but married people had shorter attack durations (P=0.005). The education of people was compared with the last level of education (bachelor’s degree and above), and just primary school educated had less under 12 h headache than people with the bachelors’ degree and above (P=0.040). However, there was no significant difference in frequency and severity. Nonworking people experienced more frequent (P=0.007) and severe (P=0.019) attacks than their working counterparts, but no significant difference was seen in the duration of attacks. The frequency of headaches increased with increase in working hours (P<0.001), but no relationship was found between severity and duration of headaches. Rural people had more frequent headaches than those who were living in urban areas (P<0.001), but the severity and duration of attacks were not different.

In Table 4, the results for headache-related symptoms are presented. With increase in the length of headache history, frequency and severity increased (P=0.003 and P=0.025, respectively), but headaches over 12 h duration experienced a significant decrease (P=0.030). The family history of headache was not related to frequency, severity, and duration of attacks. The attacks with nausea, vomiting, photophobia, and phonophobia were more severe (P<0.001, P=0.022, P=0.004, P=0.009, respectively), the unilateral and gradual-onset attacks were more frequent than others (P<0.001 and P=0.005) and with vomiting headaches of under 12 h increased (P=0.013). There was no significant difference in frequency, severity, and duration of pulsating and non-pulsating headaches. The variance of the random effect from the negative binomial model was estimated to be 6.51% (P<0.001), corresponding to an intra-class correlation (ICC) of 57.6%, given by the Pearson’s correlation statistic. Moreover, the random effect variance of the Ordered Logit and Zero-inflated Ordered Probit were estimated to be 3.52 (P<0.001) and 2.21 (P<0.001), corresponding to a ICC of 62.2% and 54.4% respectively, given by Kendall’s tau statistic.

Discussion

Comparison of different headache treatments or evaluations of the effects of medications is usually carried out based on the changes in the triad of frequency, severity, and duration of headache attacks. Disregarding the variables affecting the above-mentioned parameters may lead to bias and the results may become distorted. In the present study, we found some socio-demographic factors and headache-associated symptoms related to one or more parameters.

We found that TTH and migraine-type headaches were different in frequency, severity, and duration. TTH attacks were more frequent, less severe, and longer than migraine attacks. Pryse-Phillips et al. ¹⁶ achieved similar results for frequency and duration; Eskin et al. ¹⁷ for severity and duration; and Celentano et al. ¹⁸ for severity. On the other hand, Hennry et al. ¹⁹ in frequency, Celentano et al. ¹⁸ in duration achieved adverse findings.

The gender difference effect on various aspects of headache has been commonly mentioned in the literature. Most of the studies have demonstrated that migraine or TTH types are more frequent, severe, and/or longer in women ^20–22 . However, no difference was found in frequency, severity and duration of TTH attacks between male and female subjects ²³ . Migraine was longer in men and there was no difference in headache severity ²⁴ . We saw more duration of headaches in women, which was in line with the other results who found longer headache duration in women with non-significant changes in attack frequency and pain severity ^25,26 . Because of severity difference between genders, the male and female headaches were examined separately ²⁷ . Being married was associated with more severe headaches in women. However, in the present study, we did not observe any relationship between marital status and both headache severity and frequency. In our survey, married people only experienced less headache durations. On the other hand, while relationship was found between low education and headache severity, we only observed that low level of education was related to the duration of attacks ²⁷ .

Celentano et al. ¹⁸ saw increasing headache duration in older ages and Dodic et al. ²⁰ showed longer duration of migraine attacks in patients who were 40 yr and above. We saw inverse association between age and attack frequency. Moreover, age was not related to headache severity and duration.

In the present study, headaches were more frequent in rural residents as confirmed in another study ²⁸ who emphasized the higher spread of headaches in small towns. Moreover, we found out that higher headache severity and frequency were associated with unemployment and, parallel, frequency increased with working hours. These were in line with other studies ^22,27 who found relationship between unemployment and severe headaches in men. Moreover, women who worked half-time experienced more severe headaches than others who worked full-time.

Headache-associated symptoms were also related to frequency, severity, and/or duration of attacks. Some of previous studies observed more frequent ²⁰ , more severe and longer ¹⁸ headaches were related to nausea but we observed such a relationship only for severity of attacks. Headaches with vomiting were more severe and longer ¹⁸ . However, in the present study, headaches with vomiting were more severe and had shorter duration. This inconsistency may be explained by the traditional supposition that headache improves by vomiting. Several studies have observed relationship between unilateral headaches and frequency as well as duration ¹⁸ ; however, we found such a relation for more frequent attacks. The association of photophobia and phonophobia was indicated with migraine attacks frequency, but we only saw this relationship for the more severe attacks ²⁰ .

The family history of headaches, length of headache history, type of headache onset and pulsatility are the factors without dependent variables in the literature. Among these factors, suffering from longer headache history, was found positively associated with frequency, severity, and duration of attacks, and gradual-onset attacks were related to frequent headaches in the present study. However, here the economic, psychological, and environmental issues have been ignored, suggested to be considered in future studies.

Some results of the present study were similar to prior headache studies, and others were in contrast to them. This may stem from some main reasons; first, while the previous studies used simple statistical tests, we used expert analysis of the types of responses. Second, we modeled covariates effects in the presence of other covariates to adjust them, while other studies checked each factor alone without attention to the effects of other variable. Third, cultural differences between Iran and western countries might be the cause for some differences. For instance, the association of marriage with more severe headaches in some other studies most likely reflects cultural differences.

Conclusion

There are some socio-demographic and headache symptoms, that affect frequency, severity and duration of headache attacks. Hence, we recommend physicians to consider the factors that have a significant relationship with frequency, severity and duration of headache attacks in the visit of migraine and TTH patients. For future studies, we modeled the mentioned three responses separately, joint modeling of doubles or the triple of the responses is suggested concerning the correlation of responses.

Conflict of interest

The author reports no relevant conflict of interest.

Funding

This work was supported by the Isfahan University of Medical science, Isfahan, Iran (grant number 396444).

Highlights

• Ignoring the factors related to headache frequency, severity and duration in headache-related studies can lead researchers to produce incorrect results.

• TTH attacks are more frequent, less severe and longer than migraine attacks.

• Some socio-demographic factors are related to attack frequency, severity and duration.

• Some headache symptoms are also related to headache frequency, severity and duration.

References

1. Niere K, Jerak A. Measurement of headache frequency, intensity and duration: comparison of patient report by questionnaire and headache diary. Physiother Res Int 2004; 9(4):149-56.
2. Zandifar A, Banihashemi M, Haghdoost F, Masjedi SS, Manouchehri N, Asgari F. Reliability and validity of the persian HIT‐6 questionnaire in migraine and tension‐type headache. Pain Pract 2014; 14(7):625-31.
3. Schmitz N, Admiraal‐Behloul F, Arkink EB, Kruit MC, Schoonman GG, Ferrari MD. Attack frequency and disease duration as indicators for brain damage in migraine. Headache J Head Face Pain 2008; 48(7):1044-55.
4. Xue CCL, MApplSc LD, Polus B, English RA, Zheng Z, Costa C Da. Electroacupuncture for Tension‐type Headache on Distal Acupoints Only: A Randomized, Controlled, Crossover Trial. Headache J Head Face Pain 2004; 44(4):333-41.
5. Latza U, Kohlmann T, Deck R, Raspe H. Influence of occupational factors on the relation between socioeconomic status and self-reported back pain in a population-based sample of German adults with back pain. Spine 2000; 25(11):1390-7.
6. Ferrara LA, Pacioni D, Di Fronzo V, Russo BF, Speranza E, Carlino V. Low-lipid diet reduces frequency and severity of acute migraine attacks. Nutr Metab Cardiovasc Dis 2015; 25(4):370-5.
7. Lipton RB, Stewart WF, Diamond S, Diamond ML, Reed M. Prevalence and burden of migraine in the United States: data from the American Migraine Study II. Headache J Head Face Pain 2001; 41(7):646-57.
8. Hagen KB, Bjørndal A, Uhlig T, Kvien TK. A population study of factors associated with general practitioner consultation for non-inflammatory musculoskeletal pain. Ann Rheum Dis 2000; 59(10):788-93.
9. Blyth FM, March LM, Brnabic AJM, Jorm LR, Williamson M, Cousins MJ. Chronic pain in Australia: a prevalence study. Pain 2001; 89(2-3):127-34.
10. Smith BH, Elliott AM, Chambers WA, Smith WC, Hannaford PC, Penny K. The impact of chronic pain in the community. Fam Pract 2001; 18(3):292-9.
11. Bergman S, Herrström P, Högström K, Petersson IF, Svensson B, Jacobsson LT. Chronic musculoskeletal pain, prevalence rates, and sociodemographic associations in a Swedish population study. J Rheumatol 2001; 28(6):1369-77.
12. Picavet HSJ, Schouten J. Musculoskeletal pain in the Netherlands: prevalences, consequences and risk groups, the DMC3-study. Pain 2003; 102(1-2):167-78.
13. Olesen J, Steiner TJ. The International classification of headache disorders, 2nd ed (ICDH-II). BMJ Publishing Group Ltd; 2004.
14. Zandifar A, Asgari F, Haghdoost F, Masjedi SS, NavidManouchehri NavidManouchehri, Banihashemi M. Reliability and validity of the migraine disability assessment scale among migraine and tension type headache in Iranian patients. Biomed Res Int 2014(2014):1-7.
15. Samani EB, Ganjali M. Mixed correlated bivariate ordinal and negative binomial longitudinal responses with nonignorable missing values. Commun Stat - Theory Methods 2014; 43:2659-73.
16. Pryse-Phillips W, Findlay H, Tugwell P, Edmeads J, Murray TJ, Nelson RF. A Canadian population survey on the clinical, epidemiologic and societal impact of migraine and tension-type headache. Can J Neurol Sci 1992; 19(3):333-9.
17. Eskin M, Akyol A, Çelik EY, Gültekin BK. Social problem‐solving, perceived stress, depression and life‐satisfaction in patients suffering from tension type and migraine headaches. Scand J Psychol 2013; 54(4):337-43.
18. Celentano DD, Stewart WF, Linet MS. The relationship of headache symptoms with severity and duration of attacks. J Clin Epidemiol 1990; 43(9):983-94.
19. Henry P, Auray JP, Gaudin AF, Dartigues JF, Duru G, Lantéri–Minet M. Prevalence and clinical characteristics of migraine in France. Neurology 2002; 59(2):232-7.
20. Dodick DW, Lipton RB, Goadsby PJ, Tfelt‐Hansen P, Ferrari MD, Diener H. Predictors of migraine headache recurrence: a pooled analysis from the eletriptan database. Headache J Head Face Pain 2008; 48(2):184-93.
21. Bener A. Frequency of headache and migraine in Qatar. Neuroepidemiology 2006; 27(2):61-6.
22. Kelman L. Pain characteristics of the acute migraine attack. Headache J Head Face Pain 2006; 46(6):942-53.
23. Roh JK, Kim JS, Ahn YO. Epidemiologic and Clinical Characteristics of Migraine and Tension‐Type Headache in Korea. Headache J Head Face Pain 1998; 38(5):356-65.
24. Dent W, Stelzhammer B, Meindl M, Matuja WBP, Schmutzhard E, Winkler AS. Migraine attack frequency, duration, and pain intensity: disease burden derived from a community‐based survey in northern Tanzania. Headache J Head Face Pain 2011; 51(10):1483-92.
25. Ertas M, Baykan B, Orhan EK, Zarifoglu M, Karli N, Saip S. One-year prevalence and the impact of migraine and tension-type headache in Turkey: a nationwide home-based study in adults. J Headache Pain 2012; 13(2):147-57.
26. Steiner TJ, Scher AI, Stewart WF, Kolodner K, Liberman J, Lipton RB. The prevalence and disability burden of adult migraine in England and their relationships to age, gender and ethnicity. Cephalalgia 2003; 23(7):519-27.
27. Bingefors K, Isacson D. Epidemiology, co‐morbidity, and impact on health‐related quality of life of self‐reported headache and musculoskeletal pain—a gender perspective. Eur J pain 2004; 8(5):435-50.
28. Göbel H, Petersen-Braun M, Soyka D. The epidemiology of headache in Germany: a nationwide survey of a representative sample on the basis of the headache classification of the International Headache Society. Cephalalgia 1994; 14(2):97-106.