OUP user menu

Prevalence of Chronic Pain in a Representative Sample in the United States

Jochen Hardt PhD, Clemma Jacobsen MS, Jack Goldberg PhD, Ralf Nickel MD, Dedra Buchwald MD
DOI: http://dx.doi.org/10.1111/j.1526-4637.2008.00425.x 803-812 First published online: 1 October 2008


Objective. Chronic pain is a common reason for seeking medical care. We estimated the prevalence of chronic regional and widespread pain in the United States population overall, and by age, sex, and race/ethnicity.

Setting. We examined the data from 10,291 respondents who participated in the 1999–2002 NHANES (National Health and Nutrition Examination Survey) and completed a pain questionnaire. Items allowed classification of chronic (≥3 months) pain as regional or widespread. We used regression models to test the association of sex and race/ethnicity with each pain outcome, adjusting for age.

Results. Chronic pain prevalence estimates were 10.1% for back pain, 7.1% for pain in the legs/feet, 4.1% for pain in the arms/hands, and 3.5% for headache. Chronic regional and widespread pain were reported by 11.0% and 3.6% of respondents, respectively. Women had higher odds than men for headache, abdominal pain, and chronic widespread pain. Mexican-Americans had lower odds compared with non-Hispanic whites and blacks for chronic back pain, legs/feet pain, arms/hands pain, and regional and widespread pain.

Conclusion. The population prevalence of chronic pain in the United States was lower than previously reported, with smaller sex-related differences and some variation by race/ethnicity.

  • Chronic Pain
  • Headache
  • Back Pain
  • Regional Pain
  • Widespread Pain


Chronic localized and widespread pain affect about 50% and about 10% of Americans, respectively [1,2]. It is, therefore, not surprising that chronic pain is the most common symptom for which patients seek medical care [3]. Chronic pain is associated with substantial psychological distress, functional impairment, and disability [4–6]. For example, chronic back pain is one of the major causes of absence from work in people under 45 years of age [7], and over a quarter of patients with chronic widespread pain receive some form of disability or injury compensation [8]. In the United States, 13% of the working population experienced pain in the last 2 weeks that reduced their ability to work [9], and chronic pain results in an estimated annual $61bn of lost productivity [9]. Although there are some successful treatments for chronic pain caused by cancer, treatment for other chronic pain is often difficult and unsuccessful [10–12]. This is especially true for chronic widespread pain [13].

Reviews of the literature have documented striking variations in the prevalence of chronic pain all over the world (e.g., [6,14]), including a World Health Organization multicenter study that reported baseline estimates of nonminor pain ranging from 5% to 41% for women and from 4% to 29% for men [15,16]. In addition to variations between different countries, prevalence estimates are heavily influenced by the methods used to assess pain and the conditions under which data are collected. For example, telephone surveys tend to result in lower rates of pain than mailed questionnaires [17–20]. Self-reported pain in the United States also differs by race and ethnicity, with most studies showing higher rates of pain for African Americans compared with non-Hispanic white people [21–23], but lower rates in African Americans and Hispanics have also been reported [2]. The underlying reasons for this variation are not well understood [24,25]. A study examining four methods of experimentally induced pain did not detect any difference in the pain threshold between African Americans, Hispanics, and non-Hispanic white people, although pain tolerance was somewhat higher in the latter group [26]. For example, non-Hispanic white people left their hands in very cold water (5°C) for an average of 47 seconds, compared with 42 seconds for African Americans and Hispanics. Most likely, a variety of cultural, socioeconomic, and biological factors, as well as interactions with the health care system, converge to affect the pain experience among Americans [24,25].

Previous population-based estimates of pain in the United States have relied on data from the National Health and Nutrition Survey (NHANES), a series of surveys initiated in 1971 to assemble national data on health and nutrition outcomes. Based on the NHANES data collected from 1971 to 1975, the overall population prevalence of musculoskeletal or joint pain lasting ≥1 month during the previous year was estimated at 14% [27]. When the same cohort was reinterviewed 8–10 years later, the prevalence of musculoskeletal or joint pain was 20%, with higher rates in women than men, and in older compared with younger respondents [28]. No differences by race/ethnicity were noted in either analysis. We used NHANES data from 1999 to 2002 to estimate the population prevalence of nonminor chronic localized and widespread pain in the United States, and to examine whether the pattern of chronic pain differs by age, sex, or race/ethnicity.


Survey and Sample

The NHANES is a series of large population-based health surveys that have been administered to representative samples of Americans since 1971 and made available for public use. All surveys have oversampled racial/ethnic minorities to ensure adequate representation in the datasets, but the NHANES does not oversample women. The first three surveys are referred to as NHANES I (1971–1975), NHANES II (1976–1980), and NHANES III (1988–1994). Since 1999, NHANES has been administered continuously, with data released to the public every 2 years. From 1999 to 2002, a random subsample of 10,291 respondents ≥20 years old completed an additional miscellaneous pain questionnaire in English through personal interviews. With respect to sex, age, socioeconomic status, and rural vs urban residence, the NHANES is a representative sample of the general United States population, with an oversampling of ethnic minorities. It should be noted, however, that the NHANES is a household survey, and institutionalized subjects are excluded. In addition, homeless people and households without telephones are less likely to be included. The miscellaneous pain questionnaire, the dataset, and further details about the 1999–2002 NHANES cohorts are available at http://www.cdc.gov/nchs/nhanes.htm.


The 1999–2002 NHANES data include self-report information on age in years, sex, and race/ethnicity (non-Hispanic white people, African American, Mexican-American, and Other). The Other category includes non-Mexican Hispanics, people reporting >1 race/ethnicity, and diverse small subgroups with insufficient numbers for reliable analysis. We have included the Other category primarily to provide complete data [29].

Assessment of Pain

The miscellaneous pain questionnaire collected information on the location and duration of self-reported pain. Respondents were first asked if they had a problem with pain within the last month (yes/no). Those endorsing pain were asked about the duration of the pain (<1 month, ≥1 month but <3 months, ≥3 months but <1 year, ≥1 year), and which of 32 body regions were affected. They were instructed to only report pain that lasted ≥24 hours and was not “fleeting or minor.” For this analysis, we defined chronic pain according to the American College of Rheumatology criteria as pain of ≥3 months duration [4]. Hence, all outcomes in this analysis refer to current, chronic, nonminor pain.

We analyzed the following seven specific pain locations from the miscellaneous pain questionnaire: [1] back pain (lower back, upper back, spine, shoulder, and neck) [2], pain in the legs/feet [3], pain in the arms/hands [4], headache [5], abdominal pain [6], pain in the face/teeth, and [7] chest pain. These seven NHANES pain location outcomes were coded as binary variables (present/not present). Additionally, we created a three-level variable based on American College of Rheumatology criteria (no chronic pain, localized but not widespread pain, widespread pain) [4]. Localized pain was defined as pain in ≥1 of the 32 body regions that did not meet the American College of Rheumatology criteria for widespread pain. Pain was classified as widespread if it was experienced above and below the waist, on both sides of the body, and at ≥1 axial location (spine, chest, upper or lower back). The full wording of the questionnaire is available on the Internet ([30], and a short version is provided as an appendix).

Statistical Analyses

We calculated unweighted means and percents to describe the demographic and pain outcomes for the NHANES sample, and weighted means and percents to make inferences to the total United States population. We also described all pain outcomes stratified by sex and race/ethnicity, and calculated prevalence estimates for the seven NHANES localized pain locations and three American College of Rheumatology pain types stratified by 10-year age categories and sex. All inferential prevalence rates and means are presented as point estimates with 95% confidence intervals.

We used regression models to test the association of sex and race/ethnicity with each pain outcome, adjusting for age. Because previous research has suggested a nonlinear relationship between age and pain [31], we included a quadratic term for age in all models. Men were the reference category for sex comparisons, and the non-Hispanic white group was the reference category for race/ethnicity comparisons. We used binary logistic regression to evaluate the association of sex and race/ethnicity for the seven binary NHANES pain location outcomes, and multinomial logistic regression for the three-category American College of Rheumatology pain type outcome. We considered an α rate of 0.05 as the threshold for statistical significance.

We weighted the data using the 4-year (1999–2002) weight strata and masked variance units (pseudo-primary sampling unit [PSU]) variables available in the NHANES public release data. All analyses were conducted with the SVY commands in Stata version 8.1 [32], which use the Taylor series method to calculate properly weighted variance estimates for the general population.


Twenty cases were excluded due to incomplete data, leaving 10,271 subjects for analysis. Table 1 shows the unweighted demographic and chronic pain characteristics of the respondents who completed the miscellaneous pain questionnaire, and the weighted inferential statistics for the U.S. general population. The unweighted age and sex distributions in the sample were similar to the weighted population estimates. NHANES respondents were, on average, 50 years old, compared with a population mean of 45 years of age. Approximately 50% of respondents and of the overall population were female. Due to the oversampling of racial/ethnic minorities, the sample contained higher percentages of black people (19% vs 11%), Mexican-Americans (23% vs 7%), and a lower percentage of white people (50% vs 72%) than the general population. The highest population prevalence estimates for the seven NHANES pain location outcomes occurred for back (10%) and legs/feet (7%), followed by arms/hands and headache (both 4%). Abdominal, face/teeth, and chest pain had the lowest prevalence(all 1%). The population prevalence estimates for chronic widespread and chronic localized pain were 4% and 11%, respectively.

View this table:
Table 1

Descriptive statistics for demographic and chronic pain outcomes

CharacteristicNHANES 1999–2002 N = 10,271United States
Unweighted ValueWeighted Estimate(95% CI)
  Age, years50.246.1(45.4–46.8)
  Female, %53.352.2(51.3–53.1)
Race/ethnicity, %
  Non-Hispanic white49.771.5(67.7–75.0)
  Non-Hispanic black19.010.9 (8.6–13.6)
  Mexican-American23.3 7.1 (5.5–9.0)
  Other* 8.110.5 (7.4–14.8)
Pain location, %
  Back 9.010.1 (9.0–11.2)
  Legs/feet 6.8 7.1 (6.4–7.8)
  Arms/hands 3.5 4.1 (3.4–4.8)
  Headache 2.9 3.5 (3.0–4.1)
  Abdominal 1.2 1.1 (0.9–1.4)
  Face/teeth 0.7 0.8 (0.6–1.2)
  Chest 0.6 0.7 (0.5–0.9)
ACR pain type, %
  Widespread 3.2 3.6 (3.1–4.2)
  Regional10.011.0 (9.9–12.3)
  • The NHANES column displays unweighted sample means and percentages; the United States columns display weighted inferential statistics for the general population.

  • * Includes non-Mexican Hispanics and people reporting >1 race or ethnicity;

  • Any chronic pain that does not qualify as widespread; CI = confidence interval.

  • ACR = American College of Rheumatology.

Table 2 reports the population prevalence estimates for the NHANES pain location and American College of Rheumatology pain type, stratified by race/ethnicity within sex. In general, Mexican-Americans had a lower prevalence than other racial/ethnic groups. The NHANES pain location outcomes for both men and women were lower among Mexican-Americans than among non-Hispanic white people and non-Hispanic black people for back pain, pain in the legs/feet, and pain in the arms/hands. For the American College of Rheumatology pain types, Mexican-American women and men had a lower prevalence of both chronic localized and chronic widespread pain than non-Hispanic white and black people. Point estimates of localized and widespread chronic pain were lower in non-Hispanic black people than in non-Hispanic white people, except for chronic widespread pain in men.

View this table:
Table 2

Population prevalence estimates in the United States by sex and race/ethnicity

%(95% CI)%(95% CI)
Pain location
  Non-Hispanic white11.4 (9.9–13.2)10.1 (8.8–11.7)
  Non-Hispanic black 9.7 (7.8–12.2) 8.2 (6.3–10.7)
  Mexican-American 6.6 (5.1–8.5) 4.1 (3.1–5.4)
  Other* 9.8 (6.5–14.4) 8.6 (5.3–13.8)
  All respondents10.7 (9.3–12.3) 9.3 (8.3–10.5)
  Non-Hispanic white 8.5 (7.4–9.8) 6.7 (5.6–8.0)
  Non-Hispanic black 7.4 (5.8–9.5) 5.7 (3.8–8.5)
  Mexican-American 4.5 (3.5–5.9) 3.0 (2.3–4.0)
  Other 5.5 (3.2–9.2) 6.2 (3.6–10.3)
  All respondents 7.8 (6.8–8.9) 6.3 (5.4–7.2)
  Non-Hispanic white 5.0 (3.8–6.5) 4.0 (3.1–5.2)
  Non-Hispanic black 3.7 (2.3–5.9) 3.1 (2.1–4.6)
  Mexican-American 1.8 (1.2–2.6) 1.6 (1.0–2.7)
  Other 3.4 (1.4–8.2) 3.0 (1.4–6.1)
  All respondents 4.4 (3.5–5.6) 3.7 (3.0–4.5)
  Non-Hispanic white 4.2 (3.2–5.5) 2.1 (1.7–2.8)
  Non-Hispanic black 5.8 (4.2–7.8) 2.1 (1.3–3.2)
  Mexican-American 4.6 (3.0–6.9) 1.1 (0.5–2.3)
  Other 7.2 (4.4–11.7) 2.4 (1.2–4.7)
  All respondents 4.7 (3.9–5.8) 2.1 (1.7–2.5)
  Non-Hispanic white 1.3 (0.9–1.8) 0.7 (0.3–1.3)
  Non-Hispanic black 2.0 (1.3–3.0) 0.7 (0.3–1.5)
  Mexican-American 1.1 (0.7–1.7) 0.8 (0.4–1.6)
  Other 1.7 (0.7–4.1) 1.5 (0.6–4.0)
  All respondents 1.4 (1.1–1.9) 0.8 (0.5–1.3)
  Non-Hispanic white 0.9 (0.5–1.4) 0.8 (0.4–1.3)
  Non-Hispanic black 1.3 (0.9–2.0) 0.2 (0.0–0.9)
  Mexican-American 0.8 (0.3–2.1) 0.3 (0.0–1.1)
  Other 1.2 (0.3–4.4) 0.6 (0.2–2.5)
  All respondents 0.9 (0.7–1.4) 0.7 (0.4–1.1)
  Non-Hispanic white 1.0 (0.6–1.4) 0.6 (0.4–1.0)
  Non-Hispanic black 1.0 (0.5–2.0) 0.2 (0.0–0.8)
  Mexican-American 0.7 (0.3–1.7) 0.6 (0.3–1.2)
  Other 0.4 (0.0–2.5)<0.1 (0.0–0.8)
  All respondents 0.9 (0.6–1.2) 0.5 (0.3–0.8)
ACR pain type
  Non-Hispanic white12.2(10.4–14.2)11.5(10.0–13.2)
  Non-Hispanic black10.7 (8.6–13.2) 8.4 (6.5–10.9)
  Mexican-American 8.4 (6.5–10.8) 4.5 (3.6–5.6)
  Other11.3 (8.0–15.6) 7.9 (4.9–12.5)
  All respondents11.7(10.2–13.3)10.3 (9.0–11.8)
  Non-Hispanic white 4.8 (3.7–6.1) 2.9 (2.1–3.8)
  Non-Hispanic black 3.9 (2.6–5.7) 3.2 (2.1–4.9)
  Mexican-American 2.0 (1.3–3.1) 1.7 (1.0–2.8)
  Other 3.0 (1.4–6.2) 3.9 (2.0–7.5)
  All respondents 4.3 (3.5–5.3) 2.9 (2.3–3.7)
  • * Includes non-Mexican Hispanics and people reporting >1 nationality.

  • Any chronic pain that does not qualify as widespread.

  • CI = confidence interval; ACR = American College of Rheumatology.

Table 3 shows the results of the logistic regression models for the pain outcomes, simultaneously adjusting for sex, race/ethnicity, and age. Women had higher odds than men for all pain sites, but these were statistically significant only for headache, abdominal pain, and chronic widespread pain. Mexican-Americans had lower odds for all pain outcomes compared with the other races. The differences were statistically significant compared with non-Hispanic white and black people for chronic back pain, pain in the legs/feet, and pain in the arms/hands, as well as for chronic localized and widespread pain.

View this table:
Table 3

Regression models evaluating the association of chronic pain with sex and race/ethnicity, adjusting for age

Female OR4 (95% CI)Non-Hispanic Black OR (95% CI)Mexican-American OR (95% CI)Other OR (95% CI)
Pain location§
  Back1.2 (1.0–1.4)0.8 (0.7–1.0)0.5 (0.4–0.7)0.9 (0.6–1.3)
  Legs/feet1.3 (1.0–1.5)0.9 (0.7–1.2)0.6 (0.5–0.7)0.8 (0.5–1.3)
  Arms/hands1.2 (0.9–1.7)0.8 (0.5–1.2)0.4 (0.3–0.7)0.7 (0.3–1.7)
  Head2.4 (1.8–3.1)1.2 (0.9–1.7)0.9 (0.6–1.3)1.5 (0.9–2.6)
  Abdomen1.9 (1.1–3.3)1.5 (1.0–2.2)1.0 (0.7–1.5)1.6 (0.8–3.4)
  Face/teeth1.5 (0.9–2.4)1.0 (0.5–1.9)0.6 (0.3–1.2)1.1 (0.3–4.0)
  Chest1.8 (1.0–3.1)0.8 (0.4–1.5)0.9 (0.5–1.8)0.3 (0.1–1.6)
ACR pain type
  Regional1.2 (1.0–1.4)0.8 (0.7–1.0)0.6 (0.4–0.7)0.8 (0.6–1.1)
  Widespread1.5 (1.1–2.1)0.9 (0.7–1.3)0.5 (0.4–0.8)0.9 (0.4–1.9)
  • * Reference group is composed of males;

  • Reference group is non-Hispanic white people;

  • Includes non-Mexican Hispanics and people reporting >1 nationality;

  • § Individual binary logistic regression models evaluating odds of reporting pain at each location vs not reporting pain;

  • Multinomial logistic regression with three-category outcome—no chronic pain (reference), regional but not widespread chronic pain, and widespread chronic pain.

  • OR = odds ratio; CI = confidence interval; ACR = American College of Rheumatology.

A nonlinear age effect was evident in all pain outcomes except for those with very low prevalence estimates (abdomen, face/teeth, and chest). In general, pain rates increased up to 50–59 years of age, then declined for people 60 years of age and older. The exception to this pattern was headache, which showed the highest rates for 30- to 59-year-olds, with lower rates for respondents in their 20s and 60s, and the lowest rates reported by people ≥70 years of age. Figure 1 graphs the trends for chronic localized and widespread pain across 10-year age categories, stratified by sex.

Figure 1

Regional and widespread chronic pain prevalence by 10-year age categories and sex. Error bars indicate 95% confidence intervals. Note: The two plots have different Y-axes.


We found that the overall population estimates for nonminor current localized and widespread pain of ≥3 months duration were low compared with previous reports. This holds true in the international comparison [7,14–16], with United States surveys [2,33], and with reports from previous NHANES investigations [27,28,34–36]. In spite of the generally lower prevalence rates, the confidence intervals around our estimates do not explicitly rule out rates that correspond to the range of values observed in these other studies. The low rates detected in our analyses could be explained by the method of data collection used by the NHANES. For instance, in-person interviews as opposed to mailed questionnaires [17–20], or the wording of the questions, could result in lower pain rates. Furthermore, the rates reported here reflect estimates for current chronic pain. Other studies typically did not require pain to be current or even recent, and likely captured any lifetime occurrence of chronic pain, resulting in higher rates. Finally, the NHANES items focused on nonminor pain. Previous epidemiological investigations that inquired about pain intensity and disability found that only 8–30% of participants rated their pain as nonminor [18,37–40].

We found that the associations with age were nonlinear for almost all chronic pain conditions. The prevalence for both localized and widespread pain increased from younger to middle ages, then leveled (localized pain) or decreased (widespread pain) among older people. This pattern has been previously reported for back pain, shoulder/arm pain, headache, and chronic widespread pain [9,20,31,33]. In contrast, other studies have described an almost linear increase of pain prevalence with age [41,42]. Our analysis might not reflect older adults who suffer from nonminor pain and do not live in private homes, because the NHANES is a household survey and does not solicit information from institutionalized individuals or those in assisted living situations. Thus, our results are not necessarily inconsistent with the observation that pain increases with age.

Except for headaches, we did not observe meaningful differences in the rates of pain between women and men, even though some other differences were statistically significant. This pattern corresponds to most previous studies, which have observed a 2:1 female predominance for headache, but not for back and chest pain [17,43–45]. Women also more often experienced abdominal and facial pain, but these differences were less pronounced, inconsistent, and not always significant. Likewise, only one other study has detected sex differences for all pain regions [46]. However, our results did not reflect the striking female predominance reported in virtually all previous investigations of chronic widespread pain [1,31,33,47–52]. Although we found that women had significantly higher odds of chronic widespread pain than men, the sex-related difference was less than in other surveys. The reasons for this are unclear, but the personal interview and focus on nonminor pain may differentially influence the reporting of pain in men and women.

Back pain is the most frequently cited region of nonminor pain, affecting at least half of those who report any localized pain [6,38]. This pattern was evident in our analysis. Chronic pain in other locations typically occurs less frequently; however, prevalence estimates are strongly dependent on the criteria or definitions used [7,53,54], precluding direct comparisons of our rates to those reported by others. Finally, although not entirely comparable in terms of pain duration or severity, data from previous NHANES cycles are noteworthy. The population prevalence of low back pain of ≥1 month duration in the 1976–1980 NHANES was estimated at 10.3% [55], but in 1988–1994 the rates had increased to 20% [56]. The 1-month pain duration in these studies was less stringent than the 3 months required by the American College of Rheumatology criteria, and is probably responsible for the higher prevalence of pain in the earlier NHANES cohorts compared with the 1999–2002 data. Pain of 1-month duration may not be chronic, and our lower rates represent more accurate estimates of current, nonminor, chronic pain.

One aspect of this study that deserves special mention is our estimates for pain syndromes by race/ethnicity. Mexican-Americans had lower odds of reporting chronic pain in the limbs and back than either black or non-Hispanic white people. Black and non-Hispanic white people did not have significantly different odds of reporting any chronic pain outcome. In a comparable study, a cross-sectional telephone survey was conducted on a nationally representative probability sample of non-Hispanic white subjects, non-Hispanic black subjects, and Hispanic subjects of any race to explore relationships between chronic pain and race/ethnicity [36]. About 33% of each group reported “frequent or persistent pain” for ≥3 months during the past year, but the pain experience and treatment preferences differed by race/ethnicity. Others have observed differences in the rates of pain in various regions between white, black, and Mexican-American adults [2,2,36], and international investigations have reported that pain differs by race/ethnicity or cultural group. For example, one study reported that the prevalence of back pain was lower among Mexicans and New Zealanders than among Americans, Japanese, Colombians, and Italians [57]. In contrast to our findings, British researchers found that musculoskeletal pain was more common among ethnic minorities than among white people [25], and white people more often experienced face/jaw pain than black people, a finding we cannot address as NHANES did not include an analogous pain question. Lastly, another study in the United States found that rates of headache, back pain and pain in various other sites did not differ between white and black people [58]. Taken together, these studies underscore the need to incorporate methods into survey research to better understand the meaning of pain and the pain experience in well-described, diverse, population-based samples.

This study has several limitations. First, our results may not be directly comparable to previous efforts, as estimates vary as a function of the intensity, duration, episodic or recurrent nature of the pain, and the number of body sites involved. Regarding the latter, individuals with localized pain may have had pain in >1 area, but did not meet the American College of Rheumatology criteria for widespread pain. A related issue is that the data collected did not allow us to ascertain the intensity or frequency of the pain or its impact on function. Second, pain was assessed in the 1999–2002 NHANES by personal interview—a method not often used in epidemiological studies of chronic pain and thus may not be directly comparable to data obtained from other study designs. Third, because the NHANES did not perform physical examinations or medical chart reviews, subjective complaints could not be correlated with objective findings. Fourth, some individuals with recurrent chronic pain, such as migraine headaches, were likely not detected by the definition of chronic pain used in the NHANES. Fifth, some observations, such as the lower rates of chronic pain among Mexican-Americans, could not be attributed to a single component cause or easily explained by the data collected by the NHANES [59]. Sixth, smaller ethnic groups were collapsed by the NHANES into the category “Other,” and we could not analyze them separately. Nonetheless, our estimates of clearly defined current, nonminor, chronic pain were based on a large and representative American sample and are likely to be among the most stable estimates available for many of these conditions overall, as well as by race/ethnicity.

In conclusion, this is a large national epidemiological survey that encompasses diverse nonminor chronic localized and widespread pain conditions and includes adequate numbers of minority respondents. Our findings suggest that the prevalence of chronic pain at various locations is lower than previously reported. Over 85% of the population did not report any current, nonminor, chronic localized or widespread pain. Furthermore, in contrast to previous reports, we found strikingly fewer sex differences, and with the possible exception of Mexican-Americans, generally similar rates of chronic pain among racial/ethnic groups. Future research should use more stringent and carefully crafted pain assessments to facilitate comparisons across pain severity, duration, and body sites.


Appendix: Abbreved Miscellanous Pain Questionnaire

View this table:


View Abstract