SEARCH

SEARCH BY CITATION

Keywords:

  • headache;
  • migraine;
  • obesity;
  • body mass index

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Epidemiology of Obesity
  5. Epidemiology of Primary Headache
  6. The Epidemiological Association Between Obesity and Headache Disorders
  7. Obesity and Secondary Headaches
  8. Conclusion
  9. Key Points
  10. Statement of Authorship
  11. References
  12. Appendix 1

Individually, both obesity and headache are conditions associated with a substantial personal and societal impact. Recent data support that obesity is comorbid with headache in general and migraine specifically, as well as with certain secondary headache conditions such as idiopathic intracranial hypertension. In the current manuscript, we first briefly review the epidemiology of obesity and common primary and secondary headache disorders individually. This is followed by a systematic review of the general population data evaluating the association between obesity and headache in general, and then obesity and migraine and tension-type headache disorders. Finally, we briefly discuss the data on the association between obesity and a common secondary headache disorder that is associated with obesity, idiopathic intracranial hypertension. Taken together, these data suggest that it is important for clinicians and patients to be aware of the headache/migraine-obesity association, given that it is potentially modifiable. Hypotheses for mechanisms of the obesity-migraine association and treatment considerations for overweight and obese headache sufferers are discussed in the companion manuscript, as part II of this topic.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Epidemiology of Obesity
  5. Epidemiology of Primary Headache
  6. The Epidemiological Association Between Obesity and Headache Disorders
  7. Obesity and Secondary Headaches
  8. Conclusion
  9. Key Points
  10. Statement of Authorship
  11. References
  12. Appendix 1

Individually, both obesity and headache are conditions associated with a substantial personal and societal impact. Recent literature has consistently demonstrated an association between obesity and headache, and with migraine in particular. Specifically, research has attempted to identify the populations in which the headache-obesity association is the strongest. In addition, specific links between various headache/migraine characteristics (eg, headache frequency, severity) have also been investigated in association with obesity. While a significant amount of data has emerged regarding the headache/migraine-obesity association, the direction of this relationship is not yet clear. In addition, questions remain regarding the modifiable nature of the obesity–headache relationship and its implications in clinical practice.

Hypotheses for mechanisms of the obesity-migraine association and treatment considerations for overweight and obese headache sufferers are discussed in the companion manuscript, as part II of this topic. In the current manuscript, we first briefly review the epidemiology of obesity and common primary headache disorders individually. This is followed by a systematic review of the general population data evaluating the association between obesity and headache in general, and then obesity and migraine and tension-type headache (TTH) specifically. Finally, we briefly review the data on the association between obesity and a common secondary headache disorder that is associated with obesity, idiopathic intracranial hypertension (IIH).

Epidemiology of Obesity

  1. Top of page
  2. Abstract
  3. Introduction
  4. Epidemiology of Obesity
  5. Epidemiology of Primary Headache
  6. The Epidemiological Association Between Obesity and Headache Disorders
  7. Obesity and Secondary Headaches
  8. Conclusion
  9. Key Points
  10. Statement of Authorship
  11. References
  12. Appendix 1

According to the World Health Organization, obesity is classified as having a total body fat (TBF) percentage greater than 35% in women and greater than 25% in men.[1] However, because cost and ease of use, most epidemiological studies utilize anthropometric indices (such as the body mass index [BMI] or waist circumference [WC]) to estimate the threshold for total body obesity (TBO) and abdominal obesity, respectively. General obesity or TBO, based on the BMI, is estimated as a BMI ≥30 kg/m,[2] while abdominal obesity (abd-O), based on the WC, is estimated as a WC >88 cm in women or >102 cm in men[2, 3] (see Appendix 1 for body composition categories based on the BMI and WC).

The prevalence of obesity has increased globally over the past decades.[4, 5] In the United States, the prevalence of general obesity (BMI ≥30) increased from 33% (women) and 27% (men) in 1999-2000 to 35% (women) and 32% (men) by 2007-2008. Similarly, the prevalence of abd-O in the United States has increased over the past decade, with 62% of women and 43% of men fulfilling criteria for abdominal obesity in 2007-2008, as compared with 56% of women and 38% of men in 1999-2000.[4]

In addition to sex-specific differences, racial differences in adipose tissue distribution, as well as in obesity, have been identified.[6] In the United States, obesity prevalence (based on BMI) has been reported to be greatest in African Americans, followed by Caucasians, and lowest in Asian Americans.[4, 7] However, notably given the same BMI, Asians have a higher TBF percentage compared with their Caucasian counterparts;[8] and black women have a lower TBF percentage compared with white women.[9]

By itself, obesity is associated with substantial personal and financial burden.[10] Further contributing to this burden, obesity has been shown to be comorbid with multiple medical disorders (eg, hyperlipidemia, cardiovascular disease, depression).[11, 12] More recently, data also support an association between obesity and pain disorders,[13] including nonstructurally related pain disorders such as headache, and migraine in particular.[14]

Epidemiology of Primary Headache

  1. Top of page
  2. Abstract
  3. Introduction
  4. Epidemiology of Obesity
  5. Epidemiology of Primary Headache
  6. The Epidemiological Association Between Obesity and Headache Disorders
  7. Obesity and Secondary Headaches
  8. Conclusion
  9. Key Points
  10. Statement of Authorship
  11. References
  12. Appendix 1

Headache, in general, is incredibly common.[15] The global lifetime prevalence of headache (all types), is 66% (male 65%, female 69%);[15] while the 1-year period prevalence is approximately 47% (male 37%, female 52%).[15]

Of the primary headache disorders, TTH is the most common, with a lifetime prevalence of approximately 46% globally[15] and a 1-year period prevalence of 38%.[15] The 1-year incidence for TTH is between 14 and 44 per 1000 person-years.[16, 17] There is a female predilection for TTH, with a female to male ratio between 1.2:1 and 3:1.[18, 19]

Migraine, while less common than TTH in the general population, is the most common primary headache disorder presenting to a physician's office.[20] The lifetime prevalence of migraine is approximately 14% globally,[15] with a 1-year period prevalence of 12-15%.[15, 21] Migraine incidence has been estimated between 3 and 18 cases per 1000 person-years.[16, 17] As with TTH, migraine is more common in women (17.6%) than men (6.5%). Additionally, in both sexes, migraine is most prevalent in those of reproductive age (between 20 and 50 years of age).[15, 21]

Migraine Incidence and Prevalence Rates Across Time

While it has been recognized that obesity incidence and prevalence rates have increased in the past several decades (particularly between 1999 and 2008),[4, 22] it is controversial as to if migraine incidence and prevalence rates have likewise increased in past decades. Several studies have reported that the incidence and/or prevalence of migraine in adolescents and adults have increased, particularly between the late 1980s to late 1990s and the mid-to-late 1990s to the early 2000s (Table 1).23-28 One study evaluating changes in migraine prevalence over time reported no increase in migraine prevalence between 1989 and 2001.[17] However, study methodologies in the earlier studies, at least in part, limit our ability to draw firm conclusions. For example, some of these studies measured medically ascertained migraine (eg, clinician-diagnosed)[23, 24] and therefore are also likely measuring secular changes in medical consultation for migraine. Additionally, other studies used self-reported or non-International Classification of Headache Disorders (ICHD) migraine diagnoses (Table 1).[25, 29, 30] Further, others have discussed an apparent increase29-31 or a lack of change[21, 32] in migraine prevalence over time without formal statistical evaluations being conducted. Regardless of whether the incidence and/or prevalence of migraine have increased, the existence or absence of such changes may be irrelevant to the validity of the migraine-obesity association, as such comparisons likely represent an example of an ecological fallacy[33] (see also http://www.stanford.edu/class/ed260/freedman549.pdf for further description and examples of ecological fallacies).

Table 1. Migraine Studies Evaluating Changes in Migraine Prevalence Across Time
Author Pub YearStudy DesignHA DxAge Range1st Time Period (n)2nd Time Period (n)Migraine Prevalence (%)
Total CohortWomenMen
1st TP2nd TP1st TP2nd TP1st TP2nd TP
  1. a

    P < .001 as compared with 1st TP.

  2. b

    Not significant.

  3. c

    P value not available.

  4. The data from this study were collected by the Center for Disease Control (CDC) that reported that from 1980 through 1989, the prevalence of migraine in US increased from 25.8 per 1000 persons to 41.0 per 1000 persons. In addition, the CDC reported that prevalence among those under 45 years of age increased by 77% in women and 64% in men in this time period. P value not available.

  5. CS-GP = cross-sectional general population; Dx = diagnosis; HA = headache; ICHD = International Classification of Headache Disorders; Long-GP = longitudinal general population; PP = period prevalence; SR = self report; TP = time-period; Yr = year.

Le et al[25] 2012Long-GPSR20-41199420021925a24311316
(n = 22,053)(n = 14,810)
Linde et al[26] 2010CS-GPICHD-Mod>191995-972006-081213a157.4
(n = 51,383)(n = 39,690)
Lyngberg et al[17] 2005CS-GPICHD25-361989200111.315.5b15.623.57.15.4
(n = 25)(n = 32)
Wang et al[27] 2005Long-GPICHD13-151999200157a6956
(n = 7942)(n = 7658)
Laurell et al[28] 2004CS-GPICHD and ICHD-Mod7-151955199747a1210
(n = 8993)(n = 1371)
Cheung[31] 2000CS-GPICHD>151992-9319981.54.7c
(n = 7356)(n = 1436)
Sillanpaa and Anttila[30] 1996CS-GPNon-ICHD<161974199226b2527
(n = 1927)(n = 1436)
Centers for Disease Control and Prevention[29] 1991CS-GPSR>171979-198119892.64.1
(n = 321,000)(n = 125,000)

The Epidemiological Association Between Obesity and Headache Disorders

  1. Top of page
  2. Abstract
  3. Introduction
  4. Epidemiology of Obesity
  5. Epidemiology of Primary Headache
  6. The Epidemiological Association Between Obesity and Headache Disorders
  7. Obesity and Secondary Headaches
  8. Conclusion
  9. Key Points
  10. Statement of Authorship
  11. References
  12. Appendix 1

Headache disorders, and migraine in particular, have long been recognized to be comorbid with multiple psychiatric (eg, depression, post-traumatic stress disorder) and medical conditions (eg, stroke, epilpesy).[34, 35] In the following sections, we review the existing literature examining the association between obesity and headache in general. We then review the literature evaluating the relationship between obesity and specific primary headache disorders, migraine, and TTH, as well as the relationship between obesity and an obesity-related secondary headache disorder, IIH.

Search Method for Obesity and Common Primary Headache Disorders

A systematic search of PubMed database was conducted on August 1, 2013 using the keywords “headache,” “migraine,” “tension-type headache,” “cluster headache,” “paroxysmal hemicrania,” “trigeminal autonomic cephalalgia” AND “obese/obesity,” “body mass index/BMI,” “overweight,” or “body fat.” In addition, reference lists of relevant articles were reviewed for possible inclusion. All English language, cross-sectional and longitiudinal, general population, and epidemiological studies of adult or adolescent (age >12) populations published between January 2000 and July 2013 were included. Over 500 studies were identified through the earlier search terms, of which 16 (including 2 on headache and migraine, 3 on TTH and migraine, and 11 on migraine only) fulfilled the earlier inclusion criteria and are reviewed later. No general population studies were identified evaluating the association between trigeminal autonomic cephalalgias and obesity, and thus, only headache in general, migraine, and TTH general population studies are reviewed later.

Obesity and Headache in General

Two studies have demonstrated a positive association between obesity and headache in general. In 2003, Scher et al conducted the first longitudinal, general population study evaluating the relationship between obesity and headache (see Table 2).[36] A total of 1192 adults, of predominantly reproductive age (18–65 years of age) with episodic headache (EH) (2-104 headache [HA] days/year) and chronic daily headache (CDH) (≥180 HA days/year) were evaluated at baseline and 11 months later. At baseline, obesity (self-reported: sr-BMI ≥30) was 34% more common in CDH participants (odds ratio [OR] 1.34, 95% confidence interval [CI] 1.0-1.8) than those with EH. Additionally, at the 11-month follow-up evaluation, EH participants with obesity were over 5 times more likely to have progressed to CDH than non-obese (sr-BMI <25) EH participants, (OR 5.28, 95% CI 1.3-21.1).[36]

Table 2. Headache and Obesity General Population Studies of Predominantly Reproductive-Age Participants
Author (Year)Study Analysis (Database)Population inclusionSexMean Age (Range)RaceExposure CategoriesOutcomeHA & BMI Dx (sr vs m)Statistical AdjustmentsFindings
Size
  1. No additional information regarding location, frequency, severity, or accompanying symptoms of HA were given.

  2. a

    Reference BMI.

  3. Abd-O = abdominal obesity (defined as WC ≥102 cm in men; WC ≥88 cm in women); BMI = body mass index; CI = confidence interval; CDH = chronic daily headache; CS = cross-sectional; EH = episodic headache; EM = episodic migraine; GP = general population study; HA = headache; HFEM = high-frequency episodic migraine; ICHD = International Classification of Headache Disorders; LFEM = low-frequency episodic migraine; Long = longitudinal; m = measured; M = men; Mex Am = Mexican American; Mig = migraine; NHANES = National Health and Nutrition Examination Survey; NR = not reported; OR = odds ratio; Prob-EM = probable episodic migraine; Sev-ETTH = severe episodic tension type HA; sr = self-reported; TBO = total body obesity; VLFEM = very low headache frequency EM; W = women; WC = waist circumference.

Brown et al[44] (2000)CS-GP (ALSWH)

1. No HA/Mig

2. HA/Mig

(Controls: Participants with no HA/Mig)

Note: Freq not reported

W20 (18-23)

NR

Birth Origin:

88.6% Australian

3.6% Asian

7.8% Other

BMI <17 (n = 377)

BMI 17-<18.5 (n = 1104)

BMI 18.5-<20 (n = 2240)

BMI 20-<25a (n = 6536)

BMI 25-<30 (n = 1817)

BMI>30 (n = 738)

OR HA or Mig

HA: Non-ICHD

BMI: sr

1. Age

2. Education

3. Smoking

4. Exercise

5. Area of residence

The OR of HA or Mig were increased in women with sr-BMI 25-<30 (OR 1.12, CI 1.00-1.25) and sr-BMI >30 (OR 1.47, CI 1.25-1.73) compared with women with normal weight (sr-BMI 20 to <25).

No conclusion can be drawn regarding obesity and HA freq

n = 12,855 (HA/Mig 7229)
Scher et al[36] (2003)Long-GP

1. EH: 2-104 HA/year

2. CDH: 180 + HA/year

(Controls: EH)

M & W Combined40 (18-65)

White 73%

Non-white 25%

Unknown 2%

BMI <25a (n = 599)

BMI 25-30 (n = 390)

BMI ≥30 (n = 316)

1) OR of prevalent CDH

2) OR CDH incidence in EH participants

HA: Non-ICHD

BMI: sr

1. Age

2. Sex

3. Race

4. Education

5. Marital status

6. BMI

1. BMI

2. Baseline headache frequency

The odds of new onset CDH was 5-fold greater in those EH participants with sr-BMI ≥30 (OR 5.28, CI 1.3-21.1) compared with EH participants with normal weight (sr-BMI<25).

The OR of CDH were increased in headache participants (ie, EH + CDH) with sr-BMI 25-30 (OR 1.26, CI 1.0-1.7) and sr-BMI ≥30 (OR 1.34, CI 1.0-1.8) as compared with headache participants with normal weight.

n = 1932 (EH 798)
Bigal et al[45] (2006)CS-GP (Master Health Study)

1. 14 or < non-migraine HA days/month

2. No HA

3. CDH

4. EM

(Controls: CDH, non-migraine HA and No HA)

M & W Combined & Stratified39 (18-89)

White 72%

Black 25%

BMI <18.5 (n = 941)

BMI 18.5-24.9a (n = 15,501)

BMI 25-29.9 (n = 9,258)

BMI 30-34.9 (n = 3,133)

BMI ≥35 (n = 1,382)

OR of EM in all BMI categories evaluate as compared to those with non-migraine HA, CDH, no HA

HA: ICHD-1

BMI: sr

1. Income

2. Marital status

3. Use of acute or preventative medications

4. Depression

Among those with EM, the odds of HFEM (ie, 10-14 headache days/month) were increased in those with sr-BMI 25-29.9 (OR 1.3, CI 1.1-1.9), sr-BMI 30-34.9 (OR 2.9, CI 1.9-4.4), and sr-BMI ≥35 groups (OR 5.7, CI 3.6-8.8) compared with those with normal weight (sr-BMI 18.5-24.9).

The OR of LFEM were not increased in those with sr-BMI 30-34.9 vs those with sr-BMI 18.5-24.9, including those with <3 HA days/month (OR 0.8; CI 0.6-0.99) and those with 3-9 HA days/month (OR 0.8; CI 0.7-1.1).

n = 30,215 (EM: 3791)
Ford et al[38] (2008)CS-GP (NHANES)

1. Sev HA or Mig

2. No Sev HA or Mig

(Controls: No Sev HA or Mig)

Note: Freq not reported

M & W Combined & Stratified46 (20-85)

White 50%

Black 18%

Mex Am 24%

BMI <18.5 (n = 121)

BMI 18.5-24.9a (n = 2313)

BMI 25-29.9 (n = 2784)

BMI ≥30 (n = 2383)

OR Mig

HA: Non-ICHD

BMI: m

1. Age

2. Sex

3. Race

4. Education

5. Marital status

6. Physical activity

7. Smoking

8. Alcohol

9. SBP

10. CRP

11. Cholesterol

12. Diabetes

The OR of migraine or severe HA was increased in those with m-BMI ≥30 (OR 1.37, CI 1.09-1.72) compared with those with normal weight (m-BMI 18.5-25).

No conclusion can be drawn regarding obesity and HA freq.

n = 7,601 (Mig: 1649)
Peterlin et al[39] (2010)CS-GP (NHANES)

1. Severe HA or Mig

2. No severe HA or Mig

(Controls: No Sev HA or Mig )

Note: Freq not reported

M & W Stratified38 (20-55)

Men:

White 70%

Black 11%

Women:

White 82%

Black 8%

1) BMI ≥30 (n = 4,585) BMI <30a (n = 11,046)

2) Abd ob (n = 6,631) No Abd-Oa (n = 9,000)

Abd OB defined as WC ≥88 cm in women and WC ≥102 cm in men

OR Mig

HA: Non-ICHD

BMI: m

1. Age

2. Education

3. Race

4. Smoking

5. Alcohol

6. Income

7. Diabetes

8. Abd-O or TBO

1. The OR of migraine or severe HA were increased in women with m-BMI ≥30 (OR 1.39, CI 1.25-1.56) and men with m-BMI ≥30 (OR 1.38, CI 1.20-1.59) compared with those with BMI <30.

2. The OR of migraine or severe HA were increased in women with Abd-O compared with women without Abd-O (OR 1.26, CI 1.1-1.45), and in men with Abd-O compared with men without Abd-O (OR 1.3, CI 1.13-1.49).

No conclusion can be drawn regarding obesity and HA freq.

n = 15,631 (Mig: 3915)
Robberstad et al[40] (2010)CS-GP (Head-Hunt Youth)

1) Mig

2) TTH

3) Non-classifiable HA

4) No HA

(Controls: No migraine or non-migraine HA)

Note: Freq not reported

M & W Combined & Stratified<18 (13-18)

White 98%

Non-white <3%

Overweight (based on pediatric cut BMI offs per sex and age) (n = 891)

(Number of participants of normal weight and obese were not given)

OR Mig

HA: Non-ICHD

BMI: m

1. Age

2. Sex

3. Smoking

4. Physical activity

The OR of migraine were increased in overweight or obese adolescents and young adults (per pediatric cut-off points) compared with those not overweight or obese (OR 1.6, CI 1.4-2.2).

No conclusion can be drawn regarding obesity and HA freq.

n = 5847 (Mig: 392)
Vo et al[41] (2011)CS-GP (OMEGA)

1) Physician DxMig

2) No Lifetime Physician MigDx

(Controls: Participants with no lifetime physician MigDx)

Note: Freq not reported

W<40 (18-40 s)

White 86%

Other 14%

BMI < 18.5 (n = 160)

BMI 18.5-24.9a (n = 2623)

BMI 25-29.9 (n = 612)

BMI ≥ 30 (n = 338)

OR Mig

HA: Non-ICHD

BMI: SR

1. age

2. Race

3. Education

4. Marital status

5. Smoking

6. Exercise status

7. History of chronic hypertension

8. Diabetes

The OR of migraine were increased in women with sr-BMI 30-34.9 (OR 1.48, CI 1.12-1.96), sr-BMI 35-39.9 (OR 2.07, CI 1.27-3.39), and sr-BMI ≥40 (OR 2.75, CI 1.60-4.70) compared with women with normal weight (BMI 18.5-24.9).

No conclusion can be drawn regarding obesity and HA freq.

n = 3733 (Mig: 672)
Yu et al[42] (2012)CS-GP

Those with:

1) EM (<15 HA days/month)

2) No EM (excluded those with headache ≥15 days/month)

(Controls: No EM)

M & W Combined43 (18-65)

Han Chinese 94%

Other 6%

BMI < 18.5 (n = 262)

BMI 18.5-<23a (n = 2,469)

BMI 23-< 25 (n = 1,064)

BMI 25-<30 (n = 1,044)

BMI ≥ 30 (n = 190)

OR EM

HA: ICHD2

BMI: M

1. Age

2. Sex

3. Education

4. Marital status

5. Occupation

6. Urban or rural habitation

The OR of EM was increased in Asians with m-BMI ≥30 (OR 2.10, CI 1.39-3.12) compared to Asians with normal weight (m-BMI 18.5-23).

Headache frequency among different BMI categories showed no significant difference.

n = 5029 (EM: 467)
Peterlin et al[43] (2013)CS-GP (NCS-R)

1) EM

2) No HA

(controls: No HA)

M & W Combined & Stratified<46.6 (18-50+)

White 85.5%

Black 14.5%

BMI < 18.5 (n = 129)

BMI18.5-24.9a (n = 1424)

BMI 25-29.9 (n = 1306)

BMI ≥ 30 (n = 1004)

OR EM

HA: ICHD2

BMI: SR

1. Age

2. Race

3. Sex

4. Smoking

5. Depression

6. Poverty index

7. Diabetes

The OR of EM was increased in those with sr-BMI ≥30 (OR 1.81, CI:1.27-2.57) compared with those with normal weight (sr-BMI 18.5-24.9). The OR of low-frequency EM (LFEM [≤108 HA days/year]) and very low HA frequency EM (VLFEM [≤60 HA days/year]) were increased in those with obesity (LFEM: OR 1.83, 95% CI 1.26-2.65; VLFEM: OR 1.89, 95% CI 1.29-2.78) as compared with those with normal weight. Additionally, this study substantiated that the risk of EM in those with obesity was stronger in those younger than 50 years of age (OR 1.86, 95% CI 1.20-2.89; P = .006) than in those older than 50 (OR 1.15, 95% CI 0.61-2.18).

Headache frequency among different BMI categories showed no significant difference.

n = 3862 (EM: 188)

Subsequently, Keith et al conducted a cross sectional analysis of over 200,000 participants (16–90 years of age) drawn from 11 large general population databases that also demonstrated an increased risk of headache in those with obesity, as well as an increasing risk of headache with increasing obesity status. Headache in general was evaluated based on questionnaires from each of the 11 utilized databases (and will be reviewed here), while self-reported migraine was reported in a subgroup of older women (≥45 years of age), from only the Women's Health Initiative database (and will be discussed in a later section). Participants with obesity (measured/sr: m/sr-BMI = 30) had an approximately 35% increased risk of headache as compared with those women with a normal BMI (BMI = 20). Those with morbid obesity (m/sr-BMI = 40) had an approximately 80% increased risk of headache.[37] Taken together these studies demonstrate that: (1) the risk of headache is increased in those with obesity and that this risk increases with increasing obesity status; (2) the association is stronger in those with chronic as compared with EH.

Obesity and Migraine

Reproductive-Age Migraine and Obesity Studies

The majority of general population studies of predominantly reproductive aged individuals (mean age <50) have demonstrated a significant association between obesity and migraine38-43 (see Table 2). The first cross-sectional population-based study demonstrating an association between obesity and headache of any kind was conducted in Australia by Brown et al in 2000.[44] This study included over 14,000 young women between the ages of 18 and 23. Controls included those reporting no headache or migraine. Obese (sr-BMI ≥30) women were 47% more likely to report migraine or headache than non-obese (sr-BMI 25–29.9) women (OR 1.47, 95% CI 1.25-1.73).[44]

Following the Brown et al and Scher et al studies, Bigal et al evaluated the prevalence of obesity in those with episodic migraine (EM) as compared with those with no headache, non-migraine headache, and possibly CDH, and found no association.[45] However, subgroup analyses evaluating just EM participants were also conducted. Although obese individuals were reported to be more likely to have high-frequency EM (HFEM; 10-14 HA days/month), as compared with those of normal weight (BMI 30-34.9: OR 2.9, 95% CI 1.9-4.4; BMI ≥ 35: OR 5.7, 95% CI 3.6-8.8), those with obesity were not reported to be more likely to have lower frequencies of EM (ie, ≤9 HA days/month). Thus, in this study, obesity was not associated with increasing frequency of EM, but HFEM specifically.[45] A second cross-sectional study by Bigal et al in 2007 also supported an association between HFEM and obesity.[46]

Subsequently, the migraine-obesity literature has consistently identified an association between migraine and obesity in general population studies evaluating those of reproductive age (the age when migraine is most prevalent)38-43 (see Table 2) and no association in those of perireproductive/postreproductive age37,39,43,47-49 (see Table 3). In 2008, Ford et al conducted a cross-sectional analysis from the National Health and Nutrition Examination Survey (NHANES). Notably, obesity status was estimated for the first time based on m-BMI. In this study, obese (m-BMI ≥ 30) participants had a 37% greater odds of migraine or severe headaches (OR 1.37, 95% CI 1.09-1.72) as compared with those of normal weight.[38] Following Ford et al's study, a 2010 cross-sectional study (also utilizing the NHANES) further defined the relationship between migraine and obesity.[39] In contrast with Ford et al's study, the odds of migraine in those with obesity were compared with those who were non-obese. Notably, this study was the first to suggest that the odds of migraine in those with obesity (by BMI) differed based on age, with a stronger association in younger individuals (<55 years) as compared with older individuals (55 years or older). Specifically, in participants 20-55 years of age, the prevalence and odds of migraine and severe headaches were increased in both women and men with obesity (m-BMI ≥ 30) as compared with non-obese individuals (women: 39.1% vs 30.2%, P ≤ .001, OR 1.39, 95% CI 1.24-1.55; men: 20.2% vs 16.1%, P ≤ .001, OR 1.38, 95% CI 1.20-1.59). However, in those older than 55 years of age, the risk of migraine in those with obesity was not increased. This study also extended the migraine–obesity relationship to include an association between migraine and abd-O in those ≤55 years old. Specifically, migraine was more prevalent in those with abd-O as compared with those without abd-O (women: 36.9% vs 28.8.2%, P ≤ .001; men: 20.1% vs 15.9%, P ≤ .001), and the risk of migraine was increased by 30-39% in those with abd-O as compared with those without abd-O (women: OR 1.39, 95% CI 1.25-1.56; men: OR 1.3, 95% CI 1.13-1.49)[39] (see Table 2).

Table 3. Migraine and Obesity General Population Studies of Perireproductive and Postreproductive-Age Participants
Author (Year)Study Analysis (Database)Population InclusionsSexMean Age (range)RaceExposure CategoriesOutcomeHA & BMI Dx (sr vs m)Statistical AdjustmentsFindings
Size
  1. a

    This study utilizes a Swedish database of women invited for mammography screening.

  2. b

    This study was a meta-analysis evaluating the association between headache in general and obesity; only the WHI database was used when evaluating the association between obesity and migraine specifically.

  3. c

    Reference group.

  4. BMI = body mass index; cs = cross-sectional; CI = confidence interval; CS = cross-sectional; EM = episodic migraine; GP = general population study; HA = headache; HR = hazard ratio; hx=history; ICHD = International Classification of Headache Disorders; long = longitudinal; Mig = migraine; (m)-BMI = measured-BMI; M = men; NHANES = National Health and Nutrition Examination Survey; NR = not reported; OR = odds ratio; prev = previous; sr-BMI = self-reported BMI; W = women, WC = waist circumference; WHI = Women's Health Initiative; yr = year.

Mattsson et al[47] (2007)CS-GPa

1. Active Mig (≥1 Mig Attack In Prev Yr)

2. Inactive Mig (No Mig Past Yr But Prev Mig Hx)

3. No Mig

W54 (40-74)NR

Active migraine (n = 130)

Inactive migraine (n = 95)

No migraine (n = 459)c

Mean BMI

Obesity Status

HA: ICHD

BMI: (m)

1. Age

2. Education

The mean (m)-BMI was not significantly different in older women with active migraine (26.7 ± 4.7), inactive migraine (27.2 ± 4), and no migraine (26.7 ± 4.5), P = .32.

Active and inactive migraine were not associated with obesity (LR chi-square = 0.07,2 d.f., P = .96).

n = 684 (EM: 130)
Keith et al[37] (2008)CS-GP (WHI)b

1) Severe/Freq HA or Mig

2) No Severe/ Freq HA or Mig

W>50 (50-79)White 83%

BMI ∼20c

BMI ∼30

BMI ∼40

OR Mig

HA: non-ICHD

BMI: (m)

1. Age

2. Race

3. Smoking

As compared with those with (m)-BMI = 20, the odds of migraine were not increased in older women with BMI = 30 (OR 1.00, 95% CI 0.96-1.04) or those with BMI = 40 (OR 0.99; 95% CI 0.92-1.07).
n = 220,370 (EM: NR)
Winter et al[49] (2009)CS-GP (WHS)

1) Active Mig (≥1 Mig Attacks In Prev Yr)

2) Inactive Mig (No Mig Past Yr But Prev Hx Mig)

3) No Migc

W54 (≥45)NR

BMI < 23 (n = 18,301)

BMI 23-24.9c (n = 12,475)

BMI 25-26.9 (n = 9744)

BMI 27-29.9 (n = 10,105)

BMI 30-34.9 (n = 8338)

BMI ≥ 35 (n = 4504)

OR Mig

HA: non-ICHD

BMI: (sr)

1. Age

2. Smoking

3. Exercise

4. Alcohol

5. HTN

6. Menopausal status

7. Hormone use

8. Hypercholesterolemia

The OR of migraine in older women with (sr)-BMI ≥ 35 (OR 1.03, CI 0.95-1.12) was not increased as compared with those with BMI < 23(ref).
n = 63,467 (Mig: 9195)
Peterlin et al[39] (2010)CS-GP (NHANES)

1) Severe HA or Mig

2) No severe HA or Migc

M & W Stratified68 (≥55)M: White 81.8% Black 8.2% W: White 79.8% Black 9.1%

1)BMI ≥ 30 (n = 4585) BMI< 30c (n = 11,046)

2) Abd ob (n = 6631) No Abd-O≠ (n = 9000)

Abd OB defined as WC ≥ 88 cm in women and WC ≥ 102 cm in men

OR Mig

HA: non-ICHD

BMI: (m)

1. Age

2. Education

3. Race

4. Smoking

5. Alcohol

6. Income

7. Diabetes

8. abd-O or TBO

The OR of migraine were not increased in older women or older men with (m)-BMI ≥30, (Women: OR 0.85, CI 0.67-1.08; Men: OR 0.82, CI 0.58-1.16) as compared with those with BMI < 30.

The OR of migraine were decreased in older women with Abd-O (OR 0.73, CI 0.56-0.97) as compared with those without Abd-O; there was no difference in OR of migraine in men with Abd-O as compared with without Abd-O (OR 0.97; CI 0.70-1.34).

n = 15,631 (Mig: 3915)
Winter et al[48] (2012)Long-GP (WHS)

Baseline: BMI 18.5- < 25

After 12.9 years

1) BMI 18.5 to <25c

2) BMI ≥ 25

3) BMI ≥ 30

W54 (≥45)White 95%

No migraine (n = 15,679)c

Any migraine (n = 3483)

HR over-wt (BMI ≥ 25) or obese (BMI ≥30)

HA: Non-ICHD

BMI: (sr)

1. Age

2. Race

3. Smoking

4. Exercise

5. Alcohol

6. HTN

7.Menopausal status

8. Hormone use

9. Hypercholesterolemia

10. Depression

11. Calorie intake

The HR for older women with any history of migraine (HR 1.00, CI 0.83-1.19) to become obese ([sr]-BMI ≥ 30) was not increased compared with women without migraine.
n = 19,162 (EM: 3483)
Peterlin et al[43] (2013)CS-GP (NCS-R)

1) EM

2) No HAc

M & W Combined & Stratified>50 (50-98)White 85.5% Black 14.5%

BMI < 18.5 (n = 129)

BMI18.5-24.9c (n = 1424)

BMI 25-29.9 (n = 1306)

BMI ≥ 30 (n = 1004)

OR EM

HA: ICHD

BMI: (sr)

1. Age

2. Race

3. Sex

4. Smoking

5. Depression

6. Poverty index

7. Diabetes

The OR of episodic migraine was not increased in older men and women with (sr)-BMI ≥ 30 (OR 1.15, CI 0.61-2.18) as compared with those with normal weight ([sr]-BMI 18.5-24.9).
n = 3862 (EM: 188)

The migraine-obesity association was also demonstrated in a general population study of younger reproductive-aged individuals.[40] In a cross-sectional analysis of 5847 adolescents (13–18 years old), Robberstad et al reported that those with migraine or probable migraine were 60% more likely to be overweight or obese (m-BMI) than those without migraine (OR 1.6, 95% CI 1.4-2.2).[40] Likewise, Vo et al also confirmed the obesity–migraine relationship in adult women of reproductive age in a cross-sectional, general population study of over 3700 premenopausal women.[41] Similar to Keith et al's study evaluating obesity and headache in general,[37] this study also established that the risk of migraine increased with increasing obesity.[41] While the overall odds of migraine in women with obesity (sr-BMI 30–34.9) was increased by almost 1.5-fold as compared with women without obesity (OR 1.48, 95% CI 1.12-1.96), those women with class II obesity (sr-BMI 35-39.9) demonstrated over a 2-fold increased risk of migraine (OR 2.07, 95% CI 1.27-3.39), while women with morbid obesity (class III; sr-BMI ≥40) had almost a 3-fold increased risk of migraine (OR 2.75, 95% CI 1.60-4.70)[41] (see Table 2).

Most recently, Peterlin et al conducted a general population, cross-sectional analysis of over 3800 participants evaluating the EM–obesity relationship.[43] This study extended the EM–obesity relationship to include those of all frequencies, including those with LFEM. In general, obese individuals had an 81% increased risk of EM as compared with those of normal weight (OR 1.81, 95% CI 1.27-2.57; P = .001). In addition, subgroup analyses demonstrated that the odds of LFEM (≤108 headache days/year) and very low headache frequency EM (VLFEM; ≤60 HA days/year) were increased by 83-89% in those with obesity (LFEM: OR 1.83, 95% CI 1.26-2.65; VLFEM: OR 1.89, 95% CI 1.29-2.78) as compared with those with normal weight. However, there were no significant increases in the mean headache frequency (normal 43.8 ± 42.3, overweight 39.2 ± 35.6, obese 42.3 ± 45 headache days/year; P = .37) in participants with EM based on obesity status from normal to overweight to obese (P = .37), or between participants of normal weight and those with EM who were obese (P = .67). Finally, this study also substantiated the age variation in the migraine–obesity relationship. Specifically, age-stratified results showed that the risk of EM in those with obesity was increased by 86% in participants younger than 50 years of age,(OR 1.86, 95% CI 1.20-2.89; P = .006) but was not significantly increased in those older than 50 (OR 1.15, 95% CI 0.61-2.18)[43] (see Table 2).

Reproductive-Age Migraine and Obesity Studies in Asian Populations

The obesity-migraine association has also been demonstrated in at least 1 Asian general population study as well.[42] Yu et al conducted a cross-sectional study of over 5000 participants evaluating the migraine-obesity association. Asians with morbid obesity (m-BMI ≥30) had a more than a 2-fold increased odds of EM (OR 2.10, 95% CI 1.39-3.12) as compared with normal-weighted (m-BMI 18.5-23) Asian individuals.[42] However, an association between obesity and headache frequency was not found (Table 2).

Post-Reproductive-Age Migraine and Obesity Studies

In contrast with those studies evaluating predominantly reproductive-age individuals, all studies evaluating perimenopausal/postmenopausal populations have not found an association between migraine and TBO or abd-O37,39,43,47-49 (see Table 3).

Obesity and Chronic Migraine (CM)

In addition to EM, the current data also support an association between CM and obesity. In a cross-sectional analysis of more than 30,000 participants, Bigal and Lipton evaluated the odds of CM in those with obesity as compared with controls (those with no headaches and ≤108 non-migraine headaches per year). The risk of CM was increased by 50% in those with BMI between 30 and 34.9 (OR 1.5, 95% CI 1.2-1.8) and by 100% in those with BMI >35 (OR 2.0, 95% CI 1.4-2.4) compared with those of normal weight (BMI 18.5-24.9).[50]

Finally, a recent cross-sectional, general population study of almost 7000 participants (migraine diagnosis based on ICHD-2) suggested that CM participants were 72% more likely to be obese (sr-BMI >30) as compared with those without headache (OR 1.72, 95% CI 1.02-2.92).[51] However, this finding was no longer significant after adjustments for frequent acute pain medication use (OR 1.85, 95% CI 0.54-6.27).[51]

Obesity and Migraine Conclusion

The current epidemiological literature on obesity and headache does not yet allow us to identify the direction of the headache/migraine–obesity relationship. However, taken together, the earlier studies support that the risk of both episodic and CM is increased in those with obesity, and that this risk may be strongest in those of reproductive age. Further, although the existing data support that the risk of migraine increases with increasing obesity status, whether the migraine–obesity relationship is dose-dependent with increasing headache frequency has not been determined.

Obesity and TTH

Substantially less data exist examining the TTH and obesity association. Three general population studies have specifically evaluated this association.[40, 46, 50] While Robberstad et al reported a 40% increased risk of TTH (episodic and chronic together) in adolescents 13-18 years of age who were overweight or obese (OR 1.4, 95% CI 1.1-1.6),[40] Bigal et al reported no association between ETTH and obesity in the predominantly adult population.[46] When evaluating CTTH, a 40% increased risk of CTTH was found in those with a BMI ≥35 as compared with those with a BMI 18.5-24.9 (OR 1.4, 95% CI 1.1-1.9) in another cross-sectional general population study by Bigal et al.[50]

Obesity and Secondary Headaches

  1. Top of page
  2. Abstract
  3. Introduction
  4. Epidemiology of Obesity
  5. Epidemiology of Primary Headache
  6. The Epidemiological Association Between Obesity and Headache Disorders
  7. Obesity and Secondary Headaches
  8. Conclusion
  9. Key Points
  10. Statement of Authorship
  11. References
  12. Appendix 1

Secondary headaches are headaches that occur in close temporal relation to another disorder known to be able to cause headaches (eg, intracranial neoplasm, infection).[52] IIH is a secondary headache condition with close links to both obesity and the female gender – both characteristics also associated with migraine and deserves special attention by physicians treating headaches.

Search Method for Obesity and Secondary Headache Disorders

A systematic search of PubMed database was conducted on August 1, 2013 using keywords “secondary headache,” “idiopathic intracranial hypertension,” “pseudotumor cerebri” AND “obese/obesity,” “body mass index/BMI,” “overweight,” or “body fat” for general population epidemiological studies published between January 2000 and July 2013. In addition, reference lists of relevant articles were reviewed for possible inclusion. Because of a paucity of data, search dates were extended to include those studies published between July 1983 and July 2013, as well as to include both general population and clinic-based epidemiological studies of adult or adolescent (age >12) populations. Over 300 studies were identified, of which 12 fulfilled the earlier inclusion criteria and are reviewed later.

Obesity and IIH

IIH, also known as pseudotumor cerebri, is most commonly characterized by a progressive headache associated with increased cerebral spinal fluid (CSF) pressure (>200 mm H2O in the non-obese and >250 mm H2O in the obese), normal CSF chemistry, and the absence of other structural, vascular, metabolic, toxic, or hormonal causes of increased intracranial pressure. These patients often present with enlarged blind spot, papilledema, visual field deficits, and occasionally 6th nerve palsies.[52] Headache is reported by about 75-94% of IIH patients[53, 54] and is often the first symptom reported.[55, 56] With withdrawal of CSF and lowered pressure, headache usually improves.[52] There is a substantial financial burden associated with IIH, and this cost may be increasing with the increasing incidence and prevalence of obesity. According to 1 report, 38% of IIH patients have been hospitalized, with a total economic cost of IIH patients exceeding more than $400 million.[57]

IIH is much more common in women than men, reported as 84-90% women in case-control studies.[58, 59] The association between obesity and IIH may be more robust in women as well, as case-control studies have shown that while 25-65% of male IIH sufferers are overweight, around 80% of females with IIH are overweight.59-61 However, even in men with IIH, the prevalence of obesity is higher than that of healthy male controls.[60]

The average annual age-adjusted incidence rate for IIH had been estimated to be about 1-3 per 100,000 in the general female population;[56, 62, 63] and in women with obesity, the incidence has been estimated to be as high as 12-19 per 100,000.[58, 63] Further, a lower incidence has been reported in populations with a lower prevalence of obesity,[64] while a higher incidence has been reported in a population of obese patients seeking bariatric surgery,[65] further supporting the role of obesity in IIH. The prevalence of IIH is around 11 per 100,000 in the general female population and 86 per 100,000 in obese women.[63] Small case-control studies have suggested that both BMI and weight gain are risk factors for IIH.[54, 55, 66] Specifically, higher levels of percent weight gain was found to be associated with progressively greater risk of IIH (1 year 5-10% weight gain: OR 3.6, 95% CI 1.1-11.9, P = .04; 1 year 11-15% weight gain: OR 10.2, 95% CI 1.9-56.5, P = .008; 1 year >15% weight gain: OR 15.2, 95% CI 1.5-151.2, P = .02).[66] Finally, IIH patients with normal BMI may have better visual outcomes compared with those with obesity.[67] Taken together, the previous studies demonstrate that: (1) IIH is more common in women, especially obese women; (2) both higher BMI and higher annual percent weight gain are associated with greater risk of IIH; (3) IIH can occur in those of normal weight, and they may experience a less severe phenotype.

The precise cause of IIH is not known. It has been proposed that a deficiency in intracranial CSF absorption secondary to venous hypertension whether because of intracranial venous stenosis or because of raised intrathoracic pressure secondary to increased abdominal obesity may be the etiology.[68] However, abdominal adiposity is unlikely the sole cause of IIH, as studies have demonstrated significantly different CSF pressures when comparing patients with IIH with papilledema with patients with chronic TTH with similar BMIs.[69] In fact, instead of abdominal adiposity, 1 study found that lower body adiposity (defined as weight-to-hip ratio <0.76) was actually more than 6-fold more common in IIH than in control group of obese women.[70] Further, other comorbid conditions with obesity (ie, sleep apnea, systemic hypertension) have also been suggested as possible links between obesity and IIH.[55, 71, 72] Large prospective studies are warranted to better define the causal relationship between obesity and IIH.

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Epidemiology of Obesity
  5. Epidemiology of Primary Headache
  6. The Epidemiological Association Between Obesity and Headache Disorders
  7. Obesity and Secondary Headaches
  8. Conclusion
  9. Key Points
  10. Statement of Authorship
  11. References
  12. Appendix 1

Population studies have consistently identified an association between obesity and headache in general, as well as migraine specifically. The disease risk associated with obesity includes both episodic and CM. Specifically, obesity is estimated to increase the risk of migraine by 40-80%, and this risk increases with increasing obesity status from normal weight to overweight to obese. Further, the migraine-obesity disease risk is modified by age, being greatest in those of reproductive age – the age when migraine and headaches are most common and potentially most disabling. The association between obesity and TTH is less robust, and this association may be stronger in those with chronic TTH as compared with episodic TTH. Finally, IIH is a secondary headache condition that is significantly associated with obesity as well as the process of weight gain. Taken together, these data suggest that it is important for clinicians treating patients with headache disorders, and particularly migraine, who are obese or at risk of becoming obese, to be aware of the headache/migraine-obesity association, given that it is potentially modifiable. Hypotheses for mechanisms of the obesity-migraine association and treatment considerations for overweight and obese headache sufferers in light of these data are discussed in the companion manuscript, as part II of this topic.

Key Points

  1. Top of page
  2. Abstract
  3. Introduction
  4. Epidemiology of Obesity
  5. Epidemiology of Primary Headache
  6. The Epidemiological Association Between Obesity and Headache Disorders
  7. Obesity and Secondary Headaches
  8. Conclusion
  9. Key Points
  10. Statement of Authorship
  11. References
  12. Appendix 1
  • Obesity is comorbid1 with both episodic and chronic migraine.
  • The risk of migraine may be strongest in those of reproductive age.
  • The risk of migraine overall increases with increasing obesity status from normal to overweight to obese.
  • Whether the migraine-obesity relationship is dose-dependent with increasing headache frequency has not been determined.
  • Those with episodic headache who are obese have a greater risk of headache chronification than those with episodic headache who are not obese.
Footnotes
  1. 1

    By the seminal definition of Feinstein, the definition of comorbidity is: “any distinct additional entity that has existed or may occur during the clinical course of a patient who has the index disease under study.” {{431 Feinstein 1970;}}

Statement of Authorship

  1. Top of page
  2. Abstract
  3. Introduction
  4. Epidemiology of Obesity
  5. Epidemiology of Primary Headache
  6. The Epidemiological Association Between Obesity and Headache Disorders
  7. Obesity and Secondary Headaches
  8. Conclusion
  9. Key Points
  10. Statement of Authorship
  11. References
  12. Appendix 1

Category 1

  • (a)
    Conception and Design
    Nu Cindy Chai, B. Lee Peterlin
  • (b)
    Acquisition of Data
    Nu Cindy Chai, B. Lee Peterlin
  • (c)
    Analysis and Interpretation of Data
    Nu Cindy Chai, Abhay Moghekar, Dale S. Bond, Ann I. Scher, B. Lee Peterlin

Category 2

  • (a)
    Drafting the Manuscript
    Nu Cindy Chai
  • (b)
    Revising It for Intellectual Content
    Abhay Moghekar, Dale S. Bond, Ann I. Scher, B. Lee Peterlin

Category 3

  • (a)
    Final Approval of the Completed Manuscript
    Nu Cindy Chai, Abhay Moghekar, Dale S. Bond, Ann I. Scher, B. Lee Peterlin

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Epidemiology of Obesity
  5. Epidemiology of Primary Headache
  6. The Epidemiological Association Between Obesity and Headache Disorders
  7. Obesity and Secondary Headaches
  8. Conclusion
  9. Key Points
  10. Statement of Authorship
  11. References
  12. Appendix 1

Appendix 1

  1. Top of page
  2. Abstract
  3. Introduction
  4. Epidemiology of Obesity
  5. Epidemiology of Primary Headache
  6. The Epidemiological Association Between Obesity and Headache Disorders
  7. Obesity and Secondary Headaches
  8. Conclusion
  9. Key Points
  10. Statement of Authorship
  11. References
  12. Appendix 1

Table A1

Appendix 1: WHO Criteria for Total Body Obesity & Abdominal Obesity
  1. BMI = body mass index; WC = waist circumference.

WHO Criteria for Total Body Obesity
Women & Men
BMI < 18.5Underweight 
BMI 18.5-24.9Normal weight 
BMI 25-29.9Grade I obesityOverweight
BMI 30-39.9Grade II ObesitySevere overweight
BMI ≥ 40Grade III ObesityMorbid Obesity
WHO Criteria for Abdominal Obesity
Men
WC < 94Normal weight 
WC 94-102Action Level 1Overweight
WC > 102 cmAction Level 2Abdominal Obesity
Women
WC < 80Normal weight 
WC 80-88 cmAction Level 1Overweight
WC > 88 cmAction Level 2Abdominal Obesity