CHAPTER 2 Epidemiology
INTRODUCTION
Epidemiology is the branch of medicine that deals with the study of the causes, distribution, and control of disease in populations.1 Epidemiology of spine pain provides insight into the scope of the problem and allows us to evaluate the impact of various treatment methods and preventative strategies. Without reliable epidemiologic data it is impossible to evaluate treatment or prevention with any accuracy. In reviewing the literature on the epidemiology of spine pain, it quickly becomes evident that there are significant gaps in our knowledge which require sound evidence-based medicine for resolution. Until we have reproducible data with set criteria for spine pain, in general and specific populations, we will be unable to accurately define its natural history or the benefit of selected treatments. Historically, the medical profession has held a variety of opinions on the cause of spine pain with associated treatments. This led to the teaching of treatments without any clear scientific evidence and has propagated potentially ineffectual approaches to ill-defined causes of spine problems. The long history of opinion-based clinical medicine and medical education is coming to a close. In this era of evidence-based medicine it is essential to determine the epidemiology of spine problems so we can proceed to focusing on effective treatment and prevention.
The terms often used in studying the effect of spine pain on populations are incidence and prevalence. Prevalence is the percentage of a population that is affected with a particular disease or symptoms at a given time or during a specific set time interval. There are many factors contributing to prevalence including, but not limited to, the number of new cases, the duration of symptoms, and individuals with spine pain moving in or out of the study population.2 The determination of prevalence only requires sampling at one time point. A cross-section of the population of interest should be sampled to ensure that the data will be generalizable to the population as a whole. The study population must reflect the population to which the information will be applied or the information will have little or no utility. Point prevalence is thought to be fairly accurate when obtained in surveys, whereas prevalence over long periods of time or an individual’s lifetime is often less accurate. Memory fades with time, particularly if pain has resolved.
Prevalence and incidence of spine pain allow us to better define the scope of the problem. They also allow for the formulation of theories of etiology by analysis of associated factors. They do not determine causation.3 The estimates of costs to society further define the problem and include economic, medical, and disability-related costs.
CHALLENGES
Consistency among studies with clearly defined criteria for an episode of spine pain would allow for comparison and pooling of data. Fleeting, transient, mild neck pain should not be evaluated in the same category as severe, intense, disabling, chronic neck pain. The study should include the question used and how it was administered. The length of the particular questions used and the method of administration can alter the responses obtained. The prevalence period must be defined. Only the same prevalence periods should be compared. Point prevalence represents the most reliable information to obtain in survey studies since memory is not required. The longer the recall period the more this is apt to be affected by memory.4 This can cause errors in both directions. Memory may fade with time or events may be remembered as occurring more recently than they actually occurred.5,6 Point prevalence avoids this issue.
Generalizability of the information from a study population is frequently the goal. To allow for the extrapolation of findings from the sampled population to the larger population requires that the sample population be representative of the group as a whole. It is essential that the study population be a random sample of the target population. It is important to define the population prior to sampling so the outcome is relevant.
Scope of the problem
Spine pain is nearly ubiquitous in industrial societies. It is among the most common medical problems in developed countries. It is present in rural workers and in sedentary through heavy-duty occupations. In the majority of cases causation remains muddled. The often repeated causal factors including obesity, heavy work, leg length discrepancy, and others have not been proven. The data for low back pain vary but the lifetime prevalence in industrial nations is high, 50–85% or greater.7,8 The annual incidence is approximately 5% with some reports up to 15%.8 Back pain accounted for 15 million physician visits in 1990 in the US.9 It is a major factor in lost work days and the first or second most common cause of disability.10 In people under 45 year of age it is the most common cause of disability in the US.10
History
Back pain has been present since the earliest of recorded time. In the Edwin Smith papyrus circa 1500 BCE there is a description of back pain, including the examination and diagnosis. Neanderthal skeletons and Egyptian mummies revealed degenerative spine changes. Hypocrites (460–370 BC) noted that back pain with sciatic pain lasted about 40 days and affected men 40–60 years old.11 Historically, chronic back pain was not thought to be secondary to injury until the mid nineteenth century. This was the time of the industrial revolution and the building of the railways. It was called railway spine and thought to be related to work, or even travel on the railroad, even if there was no identifiable injury.12 This led to the acceptance of spine pain as an occupational injury.
STATISTICS
Europe
In Britain, a study in the general population with 4515 respondents from three general practices determined the prevalence of neck and back pain.13 An episode of spine pain was defined as lasting at least 1 week in duration. The 1-month prevalence of all spinal pain was 29%. The prevalence for back pain was 24.5% for women and 21.3% for men. For neck pain the prevalence was 16.5% for women and 10.7% for men. Of the total spine pain, 40% was disabling.
A British study with 12 907 respondents to a survey found a 1-year prevalence of 34% and a weekly prevalence of 20% for neck pain.14 Of the total respondents, 11% reported neck pain within the past year that interfered with their normal activities. An episode was defined as pain lasting 1 day or longer.
In one of the few prospective studies the lifetime and annual prevalence of low back pain in the UK was 59% and 42%, respectively.15 This was a mailed survey with 1455 respondents. An incidence rate of 4% was found. Age was associated with increased prevalence. An episode of back pain was defined as lasting longer than 1 day and not associated with menstrual cycle, pregnancy, or febrile illness.
Guez, in a Swedish study of 4392 adults, found an 18% prevalence of chronic neck pain with continuous pain lasting longer than 6 months.16 Of the subjects with neck pain, 30% had a history of trauma. No data were reported on the interval between trauma and neck pain. The definition of neck injury was injury that was severe enough to lead to a physician visit. In another Swedish population study with 6000 respondents, 48% of men and 38% of women reported neck pain on a self-administered questionnaire.17 The prevalence as a whole was 43% with women having a significantly higher prevalence than men. Chronic neck pain defined as lasting greater than 6 months was reported in 22% of women and 16% of men. A history of head or neck trauma was present in 25% of the subjects who developed chronic neck pain. Linton, in a Swedish study, surveyed 3000 persons and found a 2-year prevalence of 73% for low back pain.18 Of these, 17% utilized sick time and another 14% had been off work but did not use sick time.
In the Mini-Finland Health Survey 8000 people were interviewed and examined.19 Lifetime prevalence of neck pain was 71%. Chronic neck pain was diagnosed in 9.5% of the men and 13.5% of the women. An association was found between neck pain and history of injury and mental and physical stress at work. In a survey study of 10 000 Norwegians, the 1-year prevalence rate of neck pain was 34.4%.20 Neck pain lasting for more than 6 months had a prevalence of 13.8%.
In a telephone survey of 1964 participants in Catalonia, Spain, the 6-month prevalence of low back pain was 50.9%.21 Back pain was more common in women, manual workers and less-educated respondents. Back pain limited the daily activities in 36.7% and was responsible for time off work in 17% and disability pension in 6.5%.
In a Belgian study of 618 blue collar workers in the steel industry, lifetime prevalence was 66%, 1-year prevalence was 53%, and 1-week prevalence was 25%.22 An episode was any ‘problem in the low back.’ Most of these episodes were mild and categorized as fatigue or common low back pain. Only 17% sought medical advice and only 11% were limited in their occupational or domestic activities.
In the Netherlands in a survey with 3664 respondents, low back pain had a prevalence of 26.9% and neck pain 20.6%. Low back pain was the most common musculoskeletal pain and neck pain was the third most common.23
North America
National Health and Nutrition Examination Survey-CDC (NHANES 1999–2000) found the prevalence of low back pain (LBP) within the past 3 months to be 37.44% with a sample size of 4880.24 Neck pain over the previous 3 months lasting at least 1 day revealed a prevalence of 18.46%. Deyo analyzed the NHANES II data (1976–1980) with a survey population of 27 801 and found a lifetime prevalence in the US of LBP of 13.8% and prevalence in the previous year of 10.3%.25 In the Deyo study an episode of LBP was defined as lasting at least 2 weeks.
Canada has been the site for many epidemiologic studies for both lumbar and cervical ailments. Cassidy et al., with 1131 respondents to a mailed survey in Saskatchewan, found 28.4% point prevalence and 84.1% lifetime prevalence of back pain.26 The 6-month prevalence was graded into five intensities and disability categories. This was an attempt to stratify the prevalence so that transient nondisabling pain could be differentiated from disabling back pain. Low intensity/low disability back pain accounted for 48.9% of the population that had back pain in the previous 6 months. High intensity/high disability back pain was reported by 10.7% of this population. The remaining 12.3% of the subjects in the 6-month prevalence group reported high intensity/low disability back pain. Women were twice as likely as men to report severe disabling back pain; low intensity was equal between genders. The authors conclude that general prevalence is not terribly useful information since the majority of responders who had episodes of back pain had low intensity/nondisabling episodes of back pain.
Cote et al. looked at the prevalence of neck pain in the same random survey of the Saskatchewan population.27 The lifetime prevalence of neck pain was 66.7% and point prevalence was 22.2%. Neck pain was defined as any pain between the occiput and third thoracic vertebrae as detailed on a mannequin diagram. Subjects were stratified by intensity of pain and disability in a fashion similar to the study on back pain. Women experienced more neck pain than men in all severity groups. Women had a 58.8% 6-month prevalence and men had a 47.2% 6-month prevalence. The 6-month prevalence of low intensity/low disability neck pain was 39.75% and 10.1% for high intensity/low disability neck pain. A total of 4.6% of the surveyed population reported highly disabling neck pain for the previous 6-month period. Interestingly, low intensity/low disability neck pain was found to decrease with age. High disability neck pain was more prevalent in women than in men.
Kopec et al., in a longitudinal study of households in 10 provinces in Canada, were able to determine the incidence of back pain.10
