Introduction

Low back pain (LBP), a subcategory of back pain (BP), refers to the lumbar segment of the spinal column and is one of the most common musculoskeletal complaints of human beings. It is also a common reason for medical visits in the United States. Approximately 80% of the population has at least 1 episode of acute LBP at some time in a lifetime that sometimes leads to medical consultation. Many patients, however, do not seem to seek medical care and the majority of those attending primary care do not receive a specific diagnosis, which has given rise to the term, nonspecific LBP .

LBP is classified as chronic when symptoms last for more than 3 months. The prevalence of chronic LBP in the population is lower than that of acute LBP and often depends on age; gender; mechanical factors, such as heavy or unusual load activities; pathologic conditions, such as musculoligamentous sprains and strains; herniated disks and spondylolistesis; and even psychosocial factors. Primary care physicians, orthopedic surgeons, and chiropractors take care of most patients with acute and chronic LBP. Regardless of treatment, greater than 90% have complete relief of their BP within 6 weeks.

The interest of rheumatologists in LBP is primarily from the identification of patients with rheumatic diseases, in particular those related to inflammatory arthritides affecting the spine and sacroiliac joints, such as ankylosing spondylitis (AS). AS is now part of a larger spectrum of spondyloarthritides (SpAs), and the whole subset is named, axial SpAs (axSpAs) . The term covers AS and the nonradiographic form that is defined by the absence of structural changes. This article concentrates largely on AS, because most of the published studies have used AS criteria. Knowledge of the signs and symptoms that distinguish the inflammatory involvement of the spine from that of noninflammatory—mainly mechanical—conditions is fundamental for the identification of patients with AS and other SpAs in clinics. BP is also seen in rheumatic diseases, such as osteoarthritis, osteoporosis, and rheumatoid arthritis.

Regarding localization, the involvement of the spine in patients with AS and other SpAs is usually not limited to the lumbar region; rather, it extends from the sacroiliac joints up to the thoracic, lumbar, and cervical segments. Instead of LBP—the term most commonly used for the group of mechanical, noninflammatory conditions—the involvement of the spine in patients with AS is usually referred as inflammatory back pain (IBP) .

Concept development

Conceptually, IBP consists of a group of symptoms either present or absent, representing the most important characteristic clinical features of spinal and sacroiliac involvement in patients with AS and axSpAs. IBP is neither a complaint nor a symptom but a group of positive and negative variables. The most relevant characteristics of IBP were described approximately in the 1950s by several investigators. Hart and colleagues and Wilkinson and Bywaters recognized the role of physical rest and physical activity on BP in patients with AS. For example, in their study of 202 patients with AS, Wilkinson and Bywaters described, “most patients with a spinal onset noticed definite aggravation of the pain after resting, pain and stiffness being most marked on rising from bed and again after reclining in an armchair in the evening. Many found relief after activity, and a few found it necessary to get out of bed during the night to ‘limber up’ before completing their night’s rest.” They also noticed, “many, including some with quiescent disease, found that their pain was aggravated by heavy exertion or by jolting the spine.” Hart and colleagues had previously described something similar: “A frequent feature of the pain and stiffness was the aggravation caused by immobility. Waking in the morning stiff and in pain, the patient gradually became more supple during the day, feeling at his best from the afternoon until bedtime.”

Such clinical features were acknowledged and, therefore, included in the 1961 Rome criteria for the diagnosis of AS. Two of 5 clinical criteria referred to “Low back pain and stiffness for more than 3 months which is not relieved by rest” and “Pain and stiffness in the thoracic region.” The 1966 New York revision of Rome’s criteria which resulted in the New York criteria for AS changed those two parameters to one: “History or the presence of pain at the dorso-lumbar junction or in the lumbar spine.” This simplification process eliminated the 2 most important features of IBP and reduced the specificity of the New York criteria. van der Linden and colleagues analyzed the diagnostic value of the Rome criteria and the New York criteria in patients with AS, first-degree relatives (with and without sacroiliitis and with and without HLA-B27), and healthy controls (with and without HLA-B27) and proposed modifying the latter by reincorporating “Low back pain and stiffness for more than 3 months which improves with exercise, but is not relieved by rest.”

IBP was also important for the recognition of the SpAs as a group. The involvement of the axial skeleton in Amor and colleagues’ article refers to “lumbar or dorsal pain during the night, or morning stiffness of lumbar or dorsal spine”—which does not mention other features of IBP—and “buttock pain if affecting alternately the right or the left buttock.” The European Spondyloarthropathy Study Group (ESSG) criteria refer to “inflammatory spinal pain” affecting the back, dorsal, or cervical region, as defined by the presence of 4 out of 5 criteria proposed by Calin and colleagues in 1977—these were the first data based study on IBP.

Today, IBP is 1 of the clinical parameters for the classification of axSpA and peripheral SpA proposed by the Assessment of SpondyloArthritis International Society (ASAS). The definition of IBP for axSpA corresponds to the ASAS definition of IBP whereas that for peripheral SpA refers to IBP in the past “according to the rheumatologist’s judgment.” axSpA also includes “buttock pain alternating between right and left gluteal areas” in reference to axial involvement.

Concept development

Conceptually, IBP consists of a group of symptoms either present or absent, representing the most important characteristic clinical features of spinal and sacroiliac involvement in patients with AS and axSpAs. IBP is neither a complaint nor a symptom but a group of positive and negative variables. The most relevant characteristics of IBP were described approximately in the 1950s by several investigators. Hart and colleagues and Wilkinson and Bywaters recognized the role of physical rest and physical activity on BP in patients with AS. For example, in their study of 202 patients with AS, Wilkinson and Bywaters described, “most patients with a spinal onset noticed definite aggravation of the pain after resting, pain and stiffness being most marked on rising from bed and again after reclining in an armchair in the evening. Many found relief after activity, and a few found it necessary to get out of bed during the night to ‘limber up’ before completing their night’s rest.” They also noticed, “many, including some with quiescent disease, found that their pain was aggravated by heavy exertion or by jolting the spine.” Hart and colleagues had previously described something similar: “A frequent feature of the pain and stiffness was the aggravation caused by immobility. Waking in the morning stiff and in pain, the patient gradually became more supple during the day, feeling at his best from the afternoon until bedtime.”

Such clinical features were acknowledged and, therefore, included in the 1961 Rome criteria for the diagnosis of AS. Two of 5 clinical criteria referred to “Low back pain and stiffness for more than 3 months which is not relieved by rest” and “Pain and stiffness in the thoracic region.” The 1966 New York revision of Rome’s criteria which resulted in the New York criteria for AS changed those two parameters to one: “History or the presence of pain at the dorso-lumbar junction or in the lumbar spine.” This simplification process eliminated the 2 most important features of IBP and reduced the specificity of the New York criteria. van der Linden and colleagues analyzed the diagnostic value of the Rome criteria and the New York criteria in patients with AS, first-degree relatives (with and without sacroiliitis and with and without HLA-B27), and healthy controls (with and without HLA-B27) and proposed modifying the latter by reincorporating “Low back pain and stiffness for more than 3 months which improves with exercise, but is not relieved by rest.”

IBP was also important for the recognition of the SpAs as a group. The involvement of the axial skeleton in Amor and colleagues’ article refers to “lumbar or dorsal pain during the night, or morning stiffness of lumbar or dorsal spine”—which does not mention other features of IBP—and “buttock pain if affecting alternately the right or the left buttock.” The European Spondyloarthropathy Study Group (ESSG) criteria refer to “inflammatory spinal pain” affecting the back, dorsal, or cervical region, as defined by the presence of 4 out of 5 criteria proposed by Calin and colleagues in 1977—these were the first data based study on IBP.

Today, IBP is 1 of the clinical parameters for the classification of axSpA and peripheral SpA proposed by the Assessment of SpondyloArthritis International Society (ASAS). The definition of IBP for axSpA corresponds to the ASAS definition of IBP whereas that for peripheral SpA refers to IBP in the past “according to the rheumatologist’s judgment.” axSpA also includes “buttock pain alternating between right and left gluteal areas” in reference to axial involvement.

The recognition of patients with IBP

One of the initial steps in the identification of patients with AS or SpAs who complain of BP is the identification of IBP features. For that reason, clinicians have developed some sets of criteria and tools for the classification and diagnosis of IBP. Their diagnostic value—sensitivity as well as specificity and likelihood ratio (LR)—has been determined in various clinical settings in patients with AS, SpAs, or other diseases and in healthy individuals.

Calin and colleagues developed 17 questions addressing the characteristics of BP as well as 1 each on familial aggregation, age at onset, physician consultation, and past radiographic studies for screening patients with nonspecific causes of BP and to identify those with possible AS. These questions were tested in HLA-B27–positive AS as well as in HLA-B27–negative patients with normal sacroiliac joints from an orthopedic clinic and in healthy controls. A positive response to 4 of 5 specific questions allowed the differentiation of AS BP and nonspecific BP with 95% sensitivity and 85% specificity ( Box 1 ). Yet, in other studies, sensitivity was as low as 23% and 38% and specificity 75%. Despite these later findings, Calin and colleagues’ criteria had been widely used in clinical and epidemiologic studies.

Rudwaleit and colleagues developed a set of criteria for IBP, which included some changes to Calin and colleagues’ criteria, specifically patient’s maximum age, the addition of 2 parameters (morning stiffness and improvement with exercise), and the removal of 2 other parameters (duration of symptoms and insidious onset) ( Box 2 ). Another major difference is that the criteria set tested required that patients must have chronic BP and be less than 45 years old at onset. When intended for IBP classification, the presence of at least 2 parameters raises the sensitivity to 70.3%, the specificity to 81.2%, and the positive LR to 3.7. For IBP diagnosis, the presence of at least 3 parameters yields low sensitivity (33.6%) but high specificity (97.3%) and positive LR (12.4). If none of the parameters is present, sensitivity lowers to 10.9%, specificity to 57.1%, and positive LR to 0.25.

The diagnostic value of IBP as a unique manifestation of SpA (using 75% sensitivity, 76% specificity, 3.1 positive LR, and 0.33 negative LR) is markedly increased when HLA-B27 and MRI of the sacroiliac joints are positive (90% sensitivity and specificity, 9 positive LR, and 0.11 negative LR). The combination of IBP with other SpA features, specifically those listed in Amor and colleagues’ or ESSG criteria, yields variations in the diagnostic properties of the Berlin criteria.

The ASAS set of criteria for IBP resulted from an exercise with real cases in which 13 experts determined whether 20 patients with BP of diverse causes had IBP or not. Thus, 8 items related to IBP were assessed and considered positive or negative by each expert. According to the experts’ opinion, 61 of 109 (56%) judgments on the nature of BP corresponded to IBP. The concordance rate of global judgment on the presence or absence of IBP was 0.83. Except for BP duration greater than 3 months, the frequency of age at onset greater than 40 years, insidious onset, morning stiffness of the back, improvement with exercise, no improvement with rest, alternating buttock pain, and pain at night with improvement on getting out of bed were significantly higher in the group of patients judged with IBP versus those with non-IBP. The final set of IBP parameters in ASAS proposal included those that in the logistic regression analysis were independently contributory to IBP ( Box 3 ).

Weisman and colleagues developed an ascertainment tool for the identification of patients with AS based on the presence of IBP among patients with chronic BP. The first phase of the study included a literature review in search of potential items as well as their selection of by board members and finally the item generation by focus groups of patients with AS according to disease duration. The next phases included a feasibility study of the tool in AS and chronic BP patients, item reduction, and tool validation in 145 patients with AS and 308 patients with chronic BP. The final version of the tool included 12 items with which investigators expect to identify patients with AS in early stages of the disease ( Box 4 ). Two of these items refer to symptoms on non-IBP, mechanical complaints that might exclude the diagnosis of IBP.

In addition, Braun and colleagues identified which clinical parameters were predictive for a diagnosis of axSpA in patients with chronic BP presenting in primary care. For a diagnosis of axSpA, the items, age at onset less than or equal to 35 years, improvement by exercise, improvement with nonsteroidal anti-inflammatory drugs, waking up in the second half of the night, and alternating buttock pain, had 47.8% sensitivity, 86.1% specificity (area under the curve [AUC] 71.3%), positive LR of 3.4, and negative LR of 0.6. For the diagnosis of AS, greater than or equal to 3 criteria had a sensitivity of 57.4%, specificity of 85.6% (AUC 75.7%), positive LR of 4.0, and negative LR of 0.5. Finally, for a diagnosis of nonradiographic SpA, the presence of greater than or equal to 1 criterion had a sensitivity of 81.8%, specificity of 35.9% (AUC 64.9%), positive LR of 1.3, and negative LR of 0.5. Morning stiffness was irrelevant as a parameter indicating axSpA in primary care. These data indicate which parameters could be useful for identifying patients with axSpA, AS, and nonradiographic SpA in primary care settings.

Inflammatory versus noninflammatory mechanical BP

The differentiation of IBP from noninflammatory mechanical BP is often difficult, particularly in the community, primary care clinics, and certain age groups, specifically children and adolescents and older people. As discussed previously, the classification and diagnosis of BP according to IBP criteria depend on the number of items met by an individual with BP. None of the IBP criteria existing thus far have proposed a list of clinical situations that may exclude the classification and diagnosis of IBP. Weisman and colleagues approached that problem by including 2 answers related to noninflammatory BP. Alternatively, some studies, including one recently performed in primary care, found low negative LR, suggesting good performance of the criteria.

Overlap of inflammatory and noninflammatory mechanical BP symptoms occurs and the lack of exclusion criteria might favor the number of false-positive IBP individuals. A survey in an industrial complex found BP in 1880 (65%) employees, identifying 491 (26.1%) fulfilling greater than or equal to 4 of 5 of Calin and colleagues’ criteria but only 12 cases with AS.

Walker and Williamson conducted a survey among both orthopedically and neurosurgically trained spine surgeons, rheumatologists, medical practitioners with a special interest in musculoskeletal medicine, chiropractors, and manipulative physiotherapists to determine the signs and symptoms that could differentiate inflammatory and mechanical LBP. Participants had to rate 27 signs and symptoms (26 generated by the investigators) on an 11-point semantic differential scale from strongly disagree (0) to strongly agree regarding their association with mechanical or inflammatory LBP. The item, morning pain on waking, had the highest level of agreement as an indicator of ILBP, whereas constant pain, pain that wakes, and stiffness after resting showed moderate agreement. Regarding mechanical LBP, pain when lifting showed the highest level of agreement whereas intermittent pain during the day, pain that develops later in the day, pain on standing for a while, pain bending forward a little, pain on trunk flexion or extension, pain doing a sit-up, pain when driving long distances, pain getting out of a chair, and pain on repetitive bending, running, coughing or sneezing were considered moderate indicators of mechanical LBP. Based on such questions, Riksman and colleagues developed the Mechanical and Inflammatory Low Back Pain Scale to distinguish the 2 conditions. Tested in LBP patients attending chiropractic clinics, the scale distinguished 6 patients with inflammatory LBP, 5 with mechanical LBP, and 39 (78%) with mixed symptoms. A community-based study of the prevalence of AS in Norway showed that most patients with AS gave positive responses to 3 questions related to IBP (2 on BP chronicity and 1 stiffness), yet several non-AS individuals responded in the same way. Based on such findings, Gran identified the clinical features—mostly related to IBP—allowing differentiation of patients with AS from non-AS individuals, but he did not calculated their sensitivity and specificity.

Overlaps of IBP and non-IBP symptoms may also occur in specific age-group populations. The identification of IBP in adolescents with SpAs complaining of BP is often difficult. Most patients have both inflammatory and mechanical signs and symptoms at some time during the course of the disease. Many of these patients report pain relief with both spinal exercises and rest; alternatively, they often deny having pain at night but only some spinal stiffness when waking up. Alternatively, 2 recent studies in the US have reported a high prevalence of IBP and SpAs, according to different sets of criteria, in individuals between 50 and 69 years. The prevalence rates of IBP in 1915 individuals in that age group (representing 37.5% of the total sample) were 4.1% and 5.0%, according to Calin and colleagues’ and the ESSG criteria, respectively. The age at IBP onset in 258 (26.6%) individuals of 980 with IBP occurred at greater than 45 years of age. The prevalence of SpAs in the same age group was 1.5% and 1.4% according to Amor and colleagues and the ESSG criteria. None of these studies analyzed the characteristics of IBP according to age. More importantly, none of them searched for alternative conditions, specifically noninflammatory mechanical conditions affecting the spine or sacroiliac joints in those age groups. It is in these situations in which clear methods to distinguish IBP from non-IBP are needed.