Neurology -- Correspondence for Couch et al., 69 (11) 1169-1177

Correspondence published:

Head or neck injury increases the risk of chronic daily headache: A population-based study: Lars J. Stovner, Harald Schrader (27 December 2007)

Reply from the Authors: James R. Couch Jr, Richard B Lipton, Walter F Stewart, Ann I Scher (27 December 2007)

Head or neck injury increases the risk of chronic daily headache: A population-based study 27 December 2007

Lars J. Stovner,
Norwegian National Headache Centre, Dept. of Neuroscience, NTNU, Trondheim; Norway
St. Olavs Hospital, N-7006 Trondheim, Norway,
Harald Schrader
Send Correspondence to journal:
Re: Head or neck injury increases the risk of chronic daily headache: A population-based study

lars.stovner{at}ntnu.no Lars J. Stovner, et al.

In the case-control study of Couch et al. on headache prevalence related to self-reported previous head and neck injuries, some methodological problems are insufficiently addressed. [1]
In a case-control study, it is difficult to ascertain whether the headache started or worsened after the trauma. It is also generally acknowledged that self-reported exposure in case control studies may be subject to differential information bias. For example, an unfortunate outcome like chronic headache will serve as a stimulus for the patient to scrutinize the past and increase the likelihood to recall certain events. This is briefly mentioned in the Discussion, but the true importance of this problem is not acknowledged, since it is stated that “Recall bias may lead to over- or under-reporting of head injury in CDH cases and controls”. [1]
We believe that over-reporting is much more likely in Western societies where head or neck pain is widely believed to be caused by mild or moderate head or neck traumas. For example, it has been demonstrated that 32% of employees in a Canadian company but only 1% of employees in a Lithuanian company expected they could get long-lasting headache after an imaginary scenario with whiplash after rear-end car collision. [2]
The authors should have mentioned that cohort studies, which are methodologically superior in this respect, are more beneficial in examining the relationship between trauma and development of headache, and that there are at least four different cohort studies in accident victims (either whiplash or concussion), with adequate control groups, which have shown no association between these types of trauma and long-lasting head or neck pain. [3] These studies from Lithuania have the additional advantage that they are performed in a culture with little negative expectation of chronic complaints and minimal possibility for economic compensation.
Therefore, we are concerned that only one of these four studies is cited, [4] and then as a single study which is the exception to the rule (“..with occasional contrary evidence” [1]), and with no reference to its methodological advantages.
Misattribution of headaches to trauma may have negative consequences in the way these patients are handled and for the legal system. It may also direct research efforts on this patient group in the wrong direction. Therefore, careful and critical evaluation of the methodological fallacies is critical when studies on causality in “post-traumatic conditions” are investigated.
References
1. Couch JR, Lipton RB, Stewart WF, Scher AI. Head or neck injury increases the risk of chronic daily headache: a population-based study. Neurology 2007;69:1169-1177.
2. Ferrari R, Obelieniene D, Russell A, Darlington P, Gervais R, Green P. Laypersons' expectation of the sequelae of whiplash injury. A cross- cultural comparative study between Canada and Lithuania. Med Sci Monit 2002;8:CR728-CR734.
3. Mickeviciene D, Schrader H, Obelieniene D, et al. A controlled prospective inception cohort study on the post-concussion syndrome outside the medicolegal context. Eur. J. Neurol. 2004;11:411-419.
4. Schrader H, Stovner LJ, Obelieniene D, et al. Examination of the diagnostic validity of 'headache attributed to whiplash injury': a controlled, prospective study. Eur J Neurol 2006;13:1226-1232.
Disclosure: The authors report no conflicts of interest.

Reply from the Authors 27 December 2007

James R. Couch Jr,
University of Oklahoma Medical School
711 Stanton L. Young Blvd, Suite 215, Oklahoma City, OK 73104,
Richard B Lipton, Walter F Stewart, Ann I Scher
Send Correspondence to journal:
Re: Reply from the Authors

james-couch{at}ouhsc.edu James R. Couch Jr, et al.

We thank Drs. Stovner and Schrader for their comments. Epidemiologists debate the merits of case-control vs. cohort studies and each study design has its strengths and weaknesses.
While cohort studies are usually less prone to recall bias, case-control studies are often preferred when the outcome is rare, the latency between exposure and the outcome is long or uncertain, and if there is uncertainty about the exposure assessment (e.g., any injury vs. medically-ascertained injury, proximate vs. lifetime injuries).
Bias in outcome assessment due to over- or under-reporting of headache can occur in both study designs. The cohort studies by Schrader or Stovner have followed the natural history of headache after head injury with loss of consciousness or after rear-end collision. [1,3,5-7] They found a transient—-but not long-term-- difference in headache prevalence between cases and matched controls with follow-up of 1 to 3 years.
We do not believe that these studies are directly comparable. If CDH is a rare complication of head or neck injury (HANI), these studies are under-powered to detect it. Additionally, they cannot address the longer term consequences of HANI or the role of less severe HANI in the development of CDH in the general population as subjects were identified through hospital or police records.
The objective of our parent study was to evaluate, in a general population, a broad range of potential risk factors for CDH. Head or neck injury was one of the many exposures we assessed and there was no special emphasis on this entity. Study subjects were identified from a well-defined population cohort. We assessed lifetime injuries of any severity and we did not rely on the study subjects to associate their injuries with their headaches.
Our results suggest that lifetime HANI increased the risk of CDH with a dose-response relationship. Our finding that the risk of CDH associated with HANI was not limited to injuries proximate to the onset of CDH was an interesting and unexpected conclusion that would not have been evident in a prospective cohort study with short-term follow-up.
The comments about cultural differences in reporting are interesting but not substantiated by direct evidence, and we believe, not relevant due to our design. Furthermore, the majority (~70%) of our cases with CDH and HANI did not identify injury as the cause of their frequent headaches when asked at the conclusion of the interview.
References
5. Mickeviciene D, Schrader H, Nestvold K, et al. A controlled historical cohort study on the post-concussion syndrome. Eur J Neurol 2002;9:581-587.
6. Obelieniene D, Schrader H, Bovim G, Miseviciene I, Sand T. Pain after whiplash: a prospective controlled inception cohort study. J Neurol Neurosurg Psychiatry 1999;66:279-283.
7. Obelieniene D, Bovim G, Schrader H, et al. Headache after whiplash: a historical cohort study outside the medico-legal context. Cephalalgia 1998;18:559-564.
Disclosure: The article to which this correspondence refers was supported by GlaxoSmithKline, the Migraine Trust, and the American Headache Society. The funding sources had no control over study design, data collection and analysis, manuscript preparation, or decision to publish. Drs. Couch, Lipton, and Stewart have all received research grants from GlaxoSmithKline in excess of $10,000. Drs. Couch and Lipton have received honoraria from GlaxoSmithKline in excess of $10,000. Dr. Scher received a grant in excess of $10,000 from GlaxoSmithKline for support of this study.