Sagittal standing posture in the general adult population: subective outcomes and determinants

Fábio Miguel Azevedo Araújo Sagittal standing posture in the general adult population: subjective outcomes and determinants. Porto | 2012 Fábio Miguel Azevedo Araújo Sagittal standing posture in the general adult population: subjective outcomes and determinants. Porto | 2012 Dissertação de candidatura ao grau de Mestre em Epidemiologia, apresentada à Faculdade de Medicina da Universidade do Porto, realizada sob a orientação científica da Professora Doutora Raquel Lucas Calado Ferreira, do Departamento de Epidemiologia Clínica, Medicina Preditiva e Saúde Pública da Faculdade de Medicina da Universidade do Porto. II Ao abrigo do Art.º 8º do Decreto-Lei n.º 288/70 esta dissertação teve como base dois manuscritos, nos quais colaborei ativamente na definição das hipóteses, recolha, análise e interpretação dos dados. Fui responsável pela redação da versão inicial dos dois manuscritos: I. Relevance of sagittal standing posture on quality of life measures among adults from the general population: a sex-specific association. II. Individual and contextual characteristics as determinants of sagittal standing posture: a population-based study among adults. III Table of contents Resumo 1 Abstract 5 Introduction 7 Sagittal spino-pelvic standing posture 7 Sagittal standing postural patterns 12 Clinical impact of non-neutral sagittal standing posture 15 Determinants of non-neutral sagittal standing posture 18 Sociodemographics 19 Anthropometrics 21 Behavioral characteristics 22 Methodological challenges concerning sagittal standing posture 23 Aims 25 Chapter I Relevance of sagittal standing posture on quality of life measures among adults from the general population: a sex-specific association 27 Chapter II Individual and contextual characteristics as determinants of sagittal standing posture: a population-based study among adults 45 Conclusions 65 References 67 IV 1 Resumo Introdução: As condições músculo-esqueléticas da coluna vertebral são uma das principais causas de dor e perda de qualidade de vida na população adulta geral. O impacto clínico do alinhamento espino-pélvico sagital sobre sintomas músculoesqueléticos e qualidade de vida tem vindo a ser demonstrado em amostras de pacientes com as mais diversas condições espinais, enquanto os determinantes da postura sagital têm sido principalmente avaliados em amostras muito selecionadas de indivíduos assintomáticos. No entanto, a relevância clínica da postura sagital em pé e os seus determinantes são ainda desconhecidos entre os adultos da população geral, especialmente em relação a padrões posturais globais não-neutros do plano sagital. Objetivos: Estudar resultados subjetivos em relação à postura sagital na posição de pé e identificar os seus determinantes em adultos da população geral, através dos seguintes objetivos específicos: 1) Analisar a relação da postura sagittal em pé com a severidade de dor da coluna vertebral e com a qualidade de vida relacionada com a saúde em homens e mulheres; 2) Estimar a associação de características sociodemográficas, antropométricas e comportamentais com parâmetros e padrões da postura sagital em pé. Métodos: Como parte do estudo EPIPorto, 489 adultos foram avaliados durante 20052008. A aquisição de dados radiográficos consistiu em radiografias de 36-polegadas obtidas com o participante numa posição de pé confortável, de onde foram registrados parâmetros espino-pélvicos individuais e os participantes foram depois classificados em um de quatro tipos de padrões posturais sagitais. A prevalência e a severidade de dor na coluna foram avaliadas através de questões auto-reportadas e a qualidade de vida relacionada com a saúde usando duas componentes principais do Short Form 36. Informação relativamente à idade, sexo, educação, ocupação, índice de massa corporal, circunferência da cinta, rácio das circunferências cinta-anca, atividade física total, atividade física de lazer, tempo despendido na posição sentada, hábitos tabágicos e consumo de tabaco, foram também obtidos. Resultados: Nos homens, apenas o rácio inclinação pélvica-incidência pélvica se encontrou estatisticamente associado com a severidade de dor da coluna vertebral e o eixo vertical sagital apresentou uma ligeira associação inversa e gradativa com a componente física da qualidade de vida. As mulheres no 1º e especialmente no 3º terço da incidência pélvica e declive sacral apresentaram maior prevalência e severidade de dor na coluna. Nas mulheres, um maior eixo vertical sagital esteve diretamente associado com prevalência e severidade de dor na coluna. Uma maior inclinação pélvica e rácio inclinação pélvica-incidência pélvica estiveram também 2 associados com a dor mais severa da coluna vertebral. Um gradiente inverso foi encontrado entre o eixo vertical sagital, inclinação pélvica ou rácio inclinação pélvicaincidência pélvica e a dimensão física de qualidade de vida relacionada com a saúde, que por sua vez, também apresentou uma relação gradativa e direta com o rácio lordose lombar-incidência pélvica. O parâmetro mais fortemente associado com a qualidade de vida foi o eixo vertical sagital (diferença média na pontuação do Short Form 36: 8,8 entre o 1º e o 3º terços; p<0,001). Idade avançada, menor educação formal, ocupação de trabalho manual, obesidade geral e central estiveram associados com um maior eixo vertical sagital e rácio inclinação pélvica-incidência pélvica. Comparativamente às respetivas categorias de referência e ao padrão postural neutro (tipo 3), e após ajuste para a idade, sexo, educação, índice de massa corporal, atividade física total e hábitos tabágicos, os adultos com sobrepeso tiveram maiores possibilidades de apresentarem um padrão postural tipo 2 (odds ratio [OR]=1,92) e tipo 4 (OR=2,13). Ser obeso esteve positivamente relacionado com padrão postural tipo 1 (OR=6,10). Obesidade central medida através da circunferência da cinta esteve diretamente associada com padrão postural tipo 1 (OR=3,45), enquanto o rácio das circunferências cinta-anca esteve inversamente associado com o padrão tipo 4 (OR=0,52). Houve também uma fraca associação direta entre sexo feminino e o padrão postural tipo 1. Relativamente a fatores comportamentais, os sujeitos no 2º e 3º terços de atividade física total exibiram todos os padrões posturais não neutros menos frequentemente, e os atuais fumadores estiveram mais propensos a apresentar um padrão postural tipo 4. Conclusão: A postura sagittal em pé não esteve consistentemente associada com dor ou qualidade de vida entre os homens adultos. Nas mulheres, incidência pélvica e declive sacral fora dos limites neutros, equilíbrio sagital aumentado, inclinação pélvica aumentada e maior retroversão pélvica podem estar envolvidos de forma causal no desenvolvimento de dor severa da coluna vertebral e consequentemente qualidade de vida diminuída. Um maior índice de massa corporal e obesidade central foram importantes potenciais determinantes de postura não neutra entre adultos da população geral. Espera-se que estratégias focadas na redução do índice de massa corporal promovam a prevenção de todas as posturas patológicas não neutras na posição de pé a um nível populacional. Tais estratégias poderão ter maior impacto nas mulheres, entre as quais a postura sagital provavelmente é um importante determinante de resultados subjetivos relevantes, tal como a dor e qualidade de vida. 3 Palavras-Chave: Postura, Coluna, Pélvis, Qualidade de Vida, Índice de Massa Corporal, Avaliação Populacional. 10 Additionally to the evident important role of pelvic morphology in shaping overall sagittal standing alignment, pelvic incidence also has the functional role of determining individual capacity to compensate through postural adjustments when sagittal imbalance emerges (22-24). Pathologic or physiologic spinal changes could, in a meaningful way, unsettle the congruence of overall spino-pelvic sagittal arrangement with the consequence of displacing sagittal balance in a forward direction. Increasing pelvic tilt would be one of the most important compensatory mechanisms that could be activated in order to restore sagittal standing balance behind femoral heads. As pelvic tilt increase, sacral slope would decrease in the same proportion, and this process could be described as a backward rotation of the pelvis around the hips, i.e., pelvic retroversion (22-26). By definition, pelvic incidence constrains the ability to allow for compensatory pelvic retroversion in case of sagittal imbalance, which is easily performed for subjects with higher pelvic incidence. Considering that pelvic incidence results from the sum of pelvic tilt and sacral slope and that, in standing position, the minimal value of sacral slope is zero degrees, the ability to increase pelvic tilt will be as high as the overall pelvic incidence, since there is an increasing range through which adaptation can occur. Therefore, the theoretical maximum value of pelvic tilt is equal to the individual pelvic incidence value (22-24). From a clinical point of view, there has been in the past a special interest in the definition of “normative” neutral range values regarding regional spino-pelvic postural parameters of the sagittal plane, with two main purposes: to allow earlier identification of non-neutral pathologic standing posture during clinical evaluation, and also to benchmark procedures of corrective surgery having those values as reference. Kuntz et al (27) in a systematic review have calculated pooled estimates of mean and variance for parameters of sagittal spino-pelvic alignment and balance in asymptomatic adults, assuming that parameters were normally distributed in the population (table 1). One of the main conclusions of these authors is that there is a wide variation regarding neutral alignment of sagittal regional curves from the occiput to the pelvis for 95% of the asymptomatic adult population (mean ± 2 standard deviations). The greatest variation occurred in the cervical spine (C2-C7) showing values from 11º of khyphosis to -45º of lordosis. Thoracic kyphosis (T1-T12) and lumbosacral lordosis (T12-S1) range from 25º to 65º and -40º to -84º, respectively. Pelvic incidence showed the highest variation among pelvic parameters, ranging from 34º to 74º. Furthermore, they also underline that minimum and maximum values observed in the studies included in the pooled analysis, were often far outside the two standard deviations from the 11 estimated mean, reinforcing the wide variation supposed to exist in “normative” neutral sagittal alignment parameters among asymptomatic adults. Table 1. Pooled estimates of the mean and variance of the neutral sagittal spino-pelvic parameters in adults. Parameter Neutral values Mean* 2 standard deviations Occiput-C2 (º) -14 14 Cervical lordosis (º) C1-C2 -29 14 C2-C7 -17 28 Cervicothoracic junction angle (º) T1-T5 14 16 Thoracic kyphosis (º) T1-T12 45 20 T4-T12 41 22 Thoracolumbar junction angle (º) T10-L2 6 16 T10-T12 9 14 Lumbosacral lordosis (º) T12-S1 -62 22 L1-L5 -44 22 L4-L5 -17 10 L5-S1 -24 12 Spinal balance C2-S1 Sagittal vertical axis (mm) 13 60 C7-S1 Sagittal vertical axis (mm) 0 48 T1-Hip axis sagittal tilt (º) -1 6 T9-Hip axis sagittal tilt (º) -11 6 Thoracic apex T7 T3-T11 Lumbar apex L4 L2-L5 Pelvic alignment (º) Pelvic incidence 54 20 Pelvic tilt 13 12 Sacral slope 41 16 *In spinal alignment parameters positive values represent a kyphotic curve and negative values represent a lordotic curve. Adapted from Kuntz et al (27). 12 Regarding sagittal balance parameters, they remained distributed across relatively narrow ranges, as happen for example with sagittal vertical axis (C7-S1), which have estimated mean ± 2 standard deviations of 0 ± 48 mm. The utility of these “normative” sagittal alignment ranges turns out to be very limited, due to the difficulty in defining what is considered “normal” in the sagittal standing posture for a specific adult. It seems that more important than knowing “normative” values of individual sagittal parameters, is to understand the close relationship between spino-pelvic regions along the sagittal plane. Sagittal standing postural patterns Several authors (8, 18, 28, 29) have advocated that analyzing sagittal postural patterns, instead of the conventional analysis focused on isolated sagittal alignment parameters, should provide a more complete understanding of the complex overall sagittal standing alignment. First, interaction among separate segments of sagittal alignment should exist and needs to be considered in the analysis of standing posture (28, 29). Second, the same angular change in a similar segment of different subjects may have a different effect on overall sagittal alignment due to the compensatory relationship between separate spino-pelvic segments (29). Third, and finally, the great variability in neutral “normative” ranges of regional spino-pelvic parameters limits the usefulness of isolated parameters when studying sagittal standing posture (8, 18). Roussouly et al (8), have analyzed the standing radiographs of a sample of 160 asymptomatic adults, having 27 years as mean age (range: 18-48 years) and constituted by 74 men and 86 women. Based on the theoretical framework of correlations between individual regional sagittal alignment parameters and also on the geometrical analysis of thoracic and lumbar spinal curves, they have proposed a classification of four types of overall sagittal standing posture in “healthy” adults (figure 2): - Type 1: The sacral slope is smaller than or equal to 35º, which is associated with a low pelvic incidence. The apex of the lumbar lordosis is located in the center of L5 vertebral body. The lower arc of lumbar lordosis is minimal, decreasing toward zero as the sacral slope approaches the horizontal. The inflexion point is low and posterior, creating a short lumbar lordosis dorsally inclined. The thoracic kyphosis is long with an extension to the thoracolumbar area. - Type 2: The sacral slope is smaller than or equal to 35º and pelvic incidence is low. The apex of the lumbar lordosis is located at base of the L4 vertebral body. The lower arc of lumbar lordosis is relatively flat. The inflection point is higher and more 13 anterior, decreasing the dorsal inclination of lumbar lordosis but increasing the number of vertebral bodies included in this curve. The entire spine is relatively hypolordotic and hypokyphotic. - Type 3: The sacral slope is between 36º and 44º, showing a high pelvic incidence. The apex of lumbar lordosis is in the center of the L4 vertebral body. The lower arc of lumbar lordosis becomes more prominent. The inflection point is at the thoracolumbar junction (T12-L1), and lumbar lordosis inclination is near the vertical. An average of four vertebral bodies constitutes the arc of lumbar lordosis. The spine is well balanced. - Type 4: The sacral slope is greater than or equal to 45º, which is associated with a high pelvic incidence. The apex of the lumbar lordosis is located at the base of the L3 vertebral body or higher. The lower arc of lumbar lordosis is prominent, and its inclination is in line with the vertical or ventrally tilted. The number of vertebrae in lordotic orientation is greater than five, and a state of segmental hyperextension exists. Figure 2. Representation of the sagittal characteristics in each of the four sagittal postural patterns types proposed by Roussouly et al (PI: Pelvic incidence) (23). In order to assign to each subject one of the four types of sagittal postural patterns defined by Roussouly et al (8), the use of cut off values of sacral slope should allow the correct identification of types 3 and 4, since they show unique ranges of sacral slope values within all postural patterns identified. Regarding the distinction 14 between type 1 and type 2 postural patterns, their identification becomes more difficult than using only the distribution of a unique individual sagittal alignment parameter. In a recent work of the same author (30), it was suggested that additionally to sacral slope range, the use of the cut off “three vertebrae in lumbar lordosis” should easily allow the distinction between type 1 (three or less vertebrae) and type 2 (more than three vertebrae) postural patterns, without meaningfully compromising the spino-pelvic complex that those sagittal patterns should represent. Chanplakorn et al (31) analyzed the standing radiographs of the lumbo-pelvic region of 100 Thai asymptomatic adult (70 men and 30 women), with a mean age of 33.3 years (range: 21-50 years). Based on correlations between sagittal lumbo-pelvic parameters, they proposed three sagittal postural patterns that should be identified by the pelvic radius S1 angle, which is inversely related with lumbar lordosis degree. The type 1 postural pattern was named as high pelvic radius S1 angle (> 45º), type 2 as average pelvic radius S1 angle (35º-45º), and type 3 as low pelvic radius S1 angle (< 35º) (18, 31). Types 1, 2 and 3 developed by Chanplakorn et al have similar sagittal lumbo-pelvic alignment characteristics to those of types 2, 3 and 4 (respectively), identified by Roussouly et al (8). However, the former authors (31) were not able to identify type 1 postural pattern of Roussouly classification, because the number of vertebrae in lumbar lordosis were not assessed. Similarly, Lee et al (32) studied radiographs of 86 Korean asymptomatic adults with mean age of 28.2 years (range: 19-39 years), of which 54 were men and 32 were women. Subjects were grouped in three types of sagittal postural patterns also based on the correlations between individual sagittal alignment parameters, using cut off values of the horizontal lumbar level (i.e., the lumbar level or disc space nearest to the horizontal axis). Type 1 postural pattern was defined when the horizontal lumbar level was L3 or above, type 2 when horizontal lumbar level was at L3-L4 or L4, and type 3 was defined when the horizontal lumbar level was below L4. Also the present postural patterns could be matched to Roussouly’s classification: types 1, 2 and 3 of the present classification are similar regarding sagittal alignment to types 4, 3 and 1 described by Roussouly et al (8), respectively. Postural pattern type 2 (Roussouly classification) had no matching, and the authors (32) argued that their type 2 postural pattern appeared to include types 2 and 3 of Roussouly classification. Wang et al (29) evaluated the global curves of standing spinal sagittal alignment of 450 Chinese osteoporotic women of mean age 75.3 years (range: 60-95 years), using for that the Spinal-Mouse® noninvasive system. Classification of sagittal postural patterns was made based on visual evaluation of spinal curves, palpation of the spine, results of Spinal-Mouse® system and following the categorization proposed by Satoh 15 et al (33) that describe sagittal changes of thoracic and lumbar curves in osteoporotic postural deformities. Five types of sagittal postural patterns were described (29): normal (type 1) – without apparent change in spinal curve; round back (type 2) – increased thoracic kyphosis and normal lumbar lordosis; hollow round back (type 3) – increased thoracic kyphosis and lumbar lordosis; whole kyphosis (type 4) – with extensive kyphosis from the thoracic and lumbar region; and the modified round back (type 5) – increased thoracic kyphosis and decreased lumbar lordosis. However, these sagittal postural patterns were defined based on a previous description of spinal osteoporotic deformities without any reference to pelvic parameters, which are the main focus on asymptomatic adults’ classifications (8, 31, 32). In addition, the classification among Chinese osteoporotic women results from a combination of quantitative and qualitative postural criterions, while postural patterns in asymptomatic adults are generally defined using cut off values of isolated radiographic alignment parameters. Therefore, the direct comparison between Wang et al (29) postural patterns and those identified in asymptomatic adults cannot be performed, since there are considerable differences between classifications. Regarding the prevalence estimates of sagittal postural patterns limited evidence is available, referring only to asymptomatic adults or adult low back pain patients (30) and also Chinese osteoporotic women (29). Particularly, using the Roussouly classification in 709 asymptomatic adults (354 men and 355 women) of mean age 36.8 years, prevalence estimates were: type 1 (4.5%), type 2 (23.3%), type 3 (47.7%), and type 4 (24.5%) (30). In 198 low back pain patients (111 men and 87 women) with mean age 39.4 years, 5.1% presented a type 1 postural pattern, 37.4% presented a type 2, 38.9% presented a type 3, and 18.7% presented a type 4 postural pattern (30). Clinical impact of non-neutral sagittal standing posture The clinical relevance of a well-balanced spine in the standing sagittal plane was demonstrated in several studies among adult patient populations, that have highlighted the observed association between an anterior displacement of sagittal balance and poorer health-related quality of life scores (13-15, 34-36). Accordingly, the role of maintaining neutral pelvic (13, 14) and regional spinal (13-15, 37, 38) sagittal alignment in order to preserve health-related quality of life has also been emphasized, even in patients mostly showing spinal deformity in the coronal plane, as scoliosis (1315, 34, 35, 37, 38). Lafage et al (13) have investigated the relationship between spino-pelvic coronal and sagittal alignment and balance parameters with measures of health-related 16 quality of life in 125 adult patients suffering from spinal deformity, both in the coronal or sagittal plane. Over 100 spino-pelvic radiographic parameters were assessed, and two of those representing sagittal balance (T1 spino-pelvic inclination and sagittal vertical axis) being the most significantly correlated with scores of health-related quality of life: a positive sagittal balance was directly associated with higher levels of disabling pain, decreased social function, worse overall quality of life, and mainly, with worse activity and physical function, as well as with “standing disability” (13). Previous evidence (15, 34) in adult patients, with or without prior spinal surgery, have also identified sagittal vertical axis as a clinically meaningful radiographic parameter. Positive sagittal balance was described as the most reliable predictor of clinical symptoms even when controlling for age effect (34), being related with higher pain (15, 34), smaller self image (34), smaller social function (34), and diminished activity and physical function (15, 34). Similar results were also found in 59 adult patients operated for scoliosis correction with a minimum of two postoperative years (14), but the effect of sagittal vertical axis on function domain loss statistical significance in multivariate analysis, being explained by age and pelvic tilt angle effects. Regarding standing alignment, pelvic tilt was identified as the most important individual radiographic alignment parameter with a negative effect on health-related quality of life measures as pelvic tilt increase, in essence regarding activity, physical function and “walking disability” (13). In this way, effective ambulation seems to be compromised due to the activation of spino-pelvic sagittal compensatory mechanisms represented by a higher pelvic tilt, that should negatively affect lower limb alignment through limited hip extension and consequently worsened walking performance (13). In orthopaedic operated patients, pelvic tilt was positively associated with poorer scores regarding pain, self image, social function, mental health, satisfaction, and specially with poorer physical function scores, where the effect on physical functioning was independent of age and sagittal vertical axis (14). The combined clinical relevance of sagittal vertical axis and pelvic tilt was also assessed using four distinct groups (13): (1) patients with sagittal vertical axis and pelvic tilt within “normal” limits; (2) patients that may have developed pelvic retroversion in order to efficiently compensate for sagittal positive imbalance; (3) patients that, having a sagittal positive imbalance, do not develop pelvic retroversion; and (4) patients showing pelvic retroversion that is unable to efficiently compensate sagittal positive imbalance. There is a progression of worse health-related quality of life measures from group 1 to group 4, where individuals group 4 have the highest average levels of physical limitations and worse overall health-related quality of life (13). 17 Although less consistently identified as a clinically relevant pelvic parameter that negatively affect health-related quality of life, it was suggested that in orthopaedic operated patients an increased pelvic incidence is associated with worse social function, satisfaction and mainly physical function scores (14). In relation to spinal alignment parameters, lumbar lordosis has been frequently suggested as a determinant of health-related quality of life measures (13-15, 37, 38), where decreased lordosis in the lumbar region is associated with increased pain (13, 38), decreased physical function (13-15), “standing and walking disability” (13), worse general health (37) and poorer overall scores of health-related quality of life (13, 15). The relation between loss of lumbar lordosis and health-related quality of life seems to be, at least, partially independent of positive sagittal balance (15). Once the need for studying sagittal standing posture as a global construct is a relatively recent concept, evidence supporting the clinical relevance of sagittal postural patterns is still very scarce. When Roussouly et al (8) developed their postural classification among asymptomatic adults, they also stated that, in clinical practice, patients with disc herniation mostly showed a type 1 or type 2 postural patterns, and patients with spinal stenosis more frequently exhibited a type 4, while rarely observed patients with “significant” complaints presented a type 3 postural pattern. Thereafter, based on biomechanical analysis of sagittal spino-pelvic organization, it was suggested that compressive forces (resulting from the sum of gravity and muscle action), would act differently according to lumbar sagittal orientation that is unique for each postural pattern (23). Thus, it could be expected that type-specific mechanical stresses would predispose those with type 1 and 2 postural patterns to develop disc pathology and those with type 4 to develop posterior facets arthritis, Baastrup disease or vertebral listhesis (23). Despite the fact that the previous hypothesis has not yet been confirmed, it is supported by studies focusing on non-neutral individual sagittal alignment parameters (9-12). Lower pelvic incidence in younger patients was suggested to promote disc herniation (10). Patients with disc herniation also present decreased sacral slope (1012), lumbar lordosis (10-12) and thoracic kyphosis (10); even after being matched with asymptomatic adult controls regarding pelvic incidence values (10). All these sagittal alignment characteristics correspond to those described in type 1 and type 2 postural patterns (8). On the other hand, a recent new classification of L5-S1 spondylolisthesis (9) proposed that all types of high-grade spondylolisthesis show high pelvic incidence, and those without pelvic compensatory retroversion also have increased sacral slope and lumbar lordosis, which are the sagittal alignment characteristics described in type 4 postural pattern (8). 18 Additionally, adult low back pain patients have smaller pelvic incidence, sacral slope and flattened sagittal spinal curves (30), in accordance with other studies identifying a more vertical sacrum and smaller lumbar lordosis as sagittal characteristics of adult low back pain patients (39, 40). Finally, using the Roussouly classification it was confirmed that adult low back pain patients were more likely than asymptomatic adult controls to show a type 2 postural pattern, and less likely to show a type 3 postural pattern (30). Concordantly, in adolescence, where sagittal postural patterns probably reflect a constitutional “true” sagittal morphotype instead of resulting from secondary alignment adaptations, all non-neutral postural patterns (sway, flat and hyperlordotic) were positively associated with different types of back pain presence, independently of gender, weight and height (28). Clinical relevance of sagittal postural patterns is also supported by the results found among Chinese osteoporotic women (29), where comparatively to the neutral postural pattern (type 1), the round back (type 2), whole kyphosis (type 4) and modified round back (type 5) were characterized by restricted spinal mobility (flexion, extension and flexion-to-extension range of movement) and worst sagittal standing balance, which in turn are known risk factors for worse quality of life in elderly subjects (41-43), and could increase risk of falls (44, 45). Particularly, worse scores in different domains of health-related quality of life (pain, activity of daily living, posture and figure and global score) were found in type 2 and type 3 (hollow round back) postural patterns, with type 4 additionally showing worse scores in the general health domain, comparatively to the neutral postural pattern (41). Determinants of non-neutral sagittal standing posture The development of a non-neutral sagittal posture is a complex phenomenon that results from an accumulation of diverse interrelated exposures throughout the life course. Nevertheless, in adults, the number of potential determinants of sagittal posture that were clarified is very limited, where only some evidence exists regarding the influence of age (5, 16, 18, 25, 29, 46-50), sex (5, 7, 16, 17, 29, 30, 32, 46, 49, 50) and body mass index (5, 7, 18, 50) on sagittal spino-pelvic standing posture, and yet, results between studies are controversial. Furthermore, previous adult studies have been performed in selected samples based on participants clinical features (asymptomatic or low back pain patients), which results in limited generalizability of the already scarce evidence that these studies have produced regarding the potential determinants of sagittal posture. The study of other potential determinants is clearly lacking, with a special interest in exposures that were shown to be associated with back pain presence, but that their relation with sagittal standing posture is unknown, as 19 socioeconomic status (51), central obesity (52, 53), physical activity levels (54, 55), time spent in sitting position (56, 57) or smoking habits (58). In addition, it is likely that all the previous internal and external factors relate with each other around a specific socioeconomic status (51, 59-61), what turns even more difficult to identify the characteristics with a causal role in the development of a non-neutral sagittal standing posture. In order to summarize and describe evidence of potential determinants of nonneutral sagittal standing posture and respective underlying pathophysiologic mechanisms, the considered relevant characteristics were categorized in three different groups: sociodemographics (age, sex, education and occupation), anthropometrics (body mass index and central obesity), and behavioral characteristics (physical activity levels, time spent in sitting position and smoking habits). Sociodemographics An ageing effect promoting positive sagittal standing balance among adults has been consistently reported (25, 46-48). However, the effect of ageing on individual sagittal spino-pelvic alignment has been controversial. Some authors have defended that with ageing, pelvic incidence (5) and pelvic tilt (16, 25) should increase and sacral slope decrease (16), while others have suggested no change in pelvic incidence (16, 25, 48), pelvic tilt (5, 48, 49) or sacral slope (5, 25, 46, 48-50). Similarly, different results support that as age increases, lumbar lordosis may decrease (5, 46, 49), increase (50) or remain unchanged (25, 46, 48), and also, thoracic kyphosis may increase (5, 25, 49, 50) or not change (46, 48). Asymptomatic adults have a more positive sagittal balance than asymptomatic adolescents (47), and age has been positively correlated with sagittal balance in asymptomatic adults aged between 21 and more than 60 years (25), 40 and 82 years (46), and 70 and 85 years (48). Positive sagittal balance should result from an agerelated decrease in lumbar lordosis (46, 48) and increased thoracic kyphosis (25), that will in turn induce increased pelvic retroversion in order to keep the gravity line relatively stable within basis of support (25, 26). In 709 asymptomatic adults (age range: 18-81 years), age was significantly correlated with pelvic tilt (r = 0.16), sacral slope (r = -0.10), pelvic tilt-pelvic incidence ratio (r = 0.19), sacral slope-pelvic incidence ratio (r = -0.19) and pelvic tilt-sacral slope ratio (r = 0.21), but the small magnitude of correlations supported authors hypothesis that physiological age-related changes in the pelvic region might not be of clinical relevance in asymptomatic adults (16). The same conclusion is sustained by the results observed in 300 asymptomatic adults (age range: 20-70 years), where only weak 26 27 Chapter I Relevance of sagittal standing posture on quality of life measures among adults from the general population: a sex-specific association. 28 29 Abstract Objectives: to analyze the relation of sagittal standing posture with back pain severity and health-related quality of life in men and women from the general population. Methods: As part of the EPIPorto population-based study among adults, 178 men and 311 women were assessed during 2005-2008. Age, education and body mass index were recorded. Radiographic data collection consisted of 36-inch sagittal radiographs obtained in free-standing posture. Sex-specific thirds of individual spinopelvic parameters were computed and one of four sagittal types of postural patterns attributed to each participant. Back pain prevalence and severity were assessed trough self-reported questions and health-related quality of life using two main components of the Short Form 36 (SF-36). Results: In men, only pelvic tilt-pelvic incidence ratio was statistically associated with back pain severity and sagittal vertical axis sagittal vertical axis had a slight inverse dose-dependent association with the physical component of SF-36. Women in the 1st and especially in the 3rd third of pelvic incidence and sacral slope presented higher prevalence and severe back pain. In women, higher sagittal vertical axis was directly associated with back pain prevalence and severity. Higher pelvic tilt and pelvic tilt-pelvic incidence ratio were also associated with more severe back pain. An inverse dose-response relation was found between sagittal vertical axis, pelvic tilt or pelvic tilt-pelvic incidence ratio and physical component of SF-36, which also had a direct dose response relation with lumbar lordosis-pelvic incidence ratio. The parameter more strongly associated with quality of life was sagittal vertical axis (mean difference in SF-36 score: 8.8 between the 1st and 3rd thirds; p<0.001). Conclusion: Sagittal standing posture was not consistently associated with pain and quality of life among men. Pelvic incidence and sacral slope outside neutral ranges, increased sagittal balance, pelvic tilt and pelvic retroversion may be involved in causing severe back pain and consequently decreased quality of life among women. Keywords: Posture, Spine, Pelvis, Back Pain, Quality of Life, Population Survey. Introduction Musculoskeletal spinal conditions are among the leading causes of pain and decreased quality of life in the general adult population (1-3). Due to predicted aging trends, it is expected that the resultant burden on individuals, societies and health systems will increase dramatically in the upcoming decades (2, 3). 30 Extreme non-neutral sagittal spino-pelvic alignment was suggested to have a determinant role in the pathologic development of spinal orthopaedic conditions, such as spondylolisthesis (4) and lumbar degenerative diseases (5-7), thereby contributing importantly for the global burden of musculoskeletal spinal conditions at a population level. Several clinical studies have consistently demonstrated associations of sagittal standing posture with pain, physical disability and reduced quality of life in patients with (8-11) or without (8, 9, 12-15) previous spinal surgery. Specifically, an anterior displacement of sagittal balance and an increased pelvic tilt have emerged as the radiographic parameters most predictive of health-related quality of life measures among highly selected samples of patients (10, 12). However, the relevance of sagittal standing posture on health-related quality of life has not been assessed in the general adult population, (8-15). Although evaluations of the influence of sagittal posture on meaningful clinical variables have been performed in adult samples less severely impaired (16-21), previous studies have relevant limitations, namely the lack of an analysis of pelvic parameters (16-21), the use of a surface technique to evaluate standing posture (18, 20, 21), or the inclusion of samples clearly unrepresentative of the general population regarding age (17-21) or clinical features (16-21). Attempts over the last years to define a neutral range of sagittal spino-pelvic postural parameters were made among asymptomatic adults without relevant orthopaedic conditions (22). However, the wide variability observed within those normative samples precludes the use of reference alignment values to identify sagittal misalignment situations of clinical relevance. Therefore, it is believed that sagittal standing overall postural patterns should allow an improved comprehension of the effect of posture on quality of life measures comparatively to isolated parameters of standing alignment (23-25). However, the importance of sagittal postural patterns as determinants of quality of life in adults has not been evaluated. By using individual sagittal spino-pelvic parameters and standing postural patterns, our aim was to analyze the relation of sagittal standing posture with back pain severity and health-related quality of life in men and women from the general population. Participants and Methods This cross-sectional study was performed as part of a longitudinal evaluation of non-institutionalized adult inhabitants of the city of Porto (EPIPorto study). The recruitment of the initial sample has been previously described (26). Assembling of the cohort was conducted between 1999 and 2003 and comprised the assessment of 2485 31 individuals selected by random digit dialing and using households as the sampling frame (participation proportion of 70.0%). In each household identified, a resident aged 18 years or more was randomly selected for evaluation without replacement if a refusal occurred. The local ethics committee of Hospital S. João approved the study protocol. All participants gave written informed consent to participate in the study. Between 2005 and 2008, 1682 individuals (67.7%) were re-evaluated and the first 518 consecutively assessed trough sagittal radiographs of the spine and pelvic region were eligible for the present study. Subjects were excluded if they had instrumentation of the spine (n=2), hip arthroplasty (n=10), and inflammatory spinal arthropathies (n=17). The final sample included 489 subjects, namely 178 men (36.4%) and 311 women (63.6%). All participants were invited to visit the University of Porto Medical School. Age, complete years of formal education and the measurement of weight and height were obtained for all participants. Body mass index was computed as the weight in kilograms divided by the square of the height in meters. Thirty-six-inch standing sagittal radiographs were obtained for each participant during free-standing posture in accordance with institutional standardized protocol. All examiners were unaware of the hypothesis under study. Radiographic films were digitized using a Vidar scanner (Vidar Systems Corp., Herndon, VA, USA) with 75 dpi resolution and 12 gray levels, and a spine dedicated software with a precision of 0.1º in angles and 0.1 mm in distances (Sectra Imtec AB, Linköping, Sweden) was used to obtain spino-pelvic measures: pelvic incidence, pelvic tilt, sacral slope, lumbar lordosis, number of vertebrae in lumbar lordosis and sagittal vertical axis (described in Table 1). A trained physical therapist who remained blind regarding the outcomes of the study (i.e., back pain and healthrelated quality of life) recorded all radiographic parameters. Lumbar region was defined from the point where the orientation of sagittal spinal curves changed from kyphosis to lordosis (inflexion point) (27). Two ratios between parameters were computed: pelvic tilt-pelvic incidence ratio – to indicate the extent of pelvic retroversion for a given pelvic incidence; and lumbar lordosis-pelvic incidence ratio – to indicate the conformity of lumbar lordosis value for a specific pelvic incidence. Sex-specific thirds were created for all individual radiographic parameters. Roussouly et al (27) have proposed four types of lumbar lordosis in adults, each of them representing a different overall spino-pelvic postural pattern in the sagittal plane. Based on sacral slope (27) and on the number of vertebrae in lumbar lordosis (28), sagittal postural pattern was determined for each participant. Types 1 and 2 have a sacral slope smaller than 35.5º, while type 1 has up to three vertebrae in lumbar lordosis and type 2 has at least four. The remaining two postural patterns have unique 32 ranges of sacral slope, between 35.5º and 44.4º (type 3), and higher or equal to 44.5º (type 4). Prevalence of back pain not directly associated to trauma was assessed trough two different questions: “Have you ever had back pain?” and “During the past month have you had back pain?”. Participants also answered back pain severity questions: “In the last year have you sought a doctor because of your back pain, having been prescribed exams or treatment?”, “During the last year did you have more than 3 episodes of pain?” and “During the last 6 months did the pain ever last longer than one week?”. A combined back pain severity variable was created: for each affirmative answer in back pain severity questions a point was given and an additional point was attributed if the intensity of back pain month prevalence was higher than 50 mm in a visual analogue scale. Health-related quality of life was assessed using the Medical Outcomes Study Short Form 36 (SF-36) (29). SF-36 has been previously translated (30) and the adapted Portuguese version has been validated (31). Two summary health components were defined in the EPIPorto Portuguese adult sample, one regarding physical and the other mental health, a procedure shown to be valid and reproducible (32). Thirty men and 66 women were not included in this analysis because they did not present complete information on SF-36. Excluded men were less educated (p=0.004) and had slightly higher lumbar lordosis (p=0.060) than those included, while women not included were older (p=0.008) and less educated (p=0.001). Age in men and body mass index, radiographic parameters and types of postural patterns in both sexes, were similar between included and non-included men and women. Statistical analysis Descriptive data were depicted as count (proportion) in categorical variables and as mean (standard deviation) or median (25th-75th percentiles) for continuous variables, as appropriate. The χ2 test was used to compare proportions between groups. Mean differences were compared trough t-student or analysis of variance (ANOVA) for normally distributed variables, while variables non-normally distributed were compared by Mann-Whitney or Kruskall-Wallis tests. Analysis was conducted separately for men and women. Results 33 Descriptive data are shown in Table 2. Men had slightly lower formal education, lower body mass index, and lower back pain prevalence and severity. Additionally, men also presented better health-related quality of life both in physical and mental components. In men, of all individual sagittal radiographic parameters, only pelvic tilt-pelvic incidence ratio was statistically associated with back pain severity (Table 3). Women in the 1st and especially in the 3rd third of pelvic incidence presented higher prevalence of severe back pain, and although not statistically significant, also a higher prevalence of back pain in the previous month. Also, women in the highest and lowest thirds of sacral slope were more likely to have recent as well as severe back pain. Higher sagittal vertical axis was positively associated with all back pain variables, and the largest difference was found regarding previous month pain (a 18.6% increase from the 1st to the 2nd third). Additionally, higher pelvic tilt (not significant) and pelvic tilt-pelvic incidence ratio were associated with higher back pain severity. When analyzing the association of sagittal standing posture with back pain variables adjusted for age, education and body mass index, similar conclusions were obtained in both sexes. In men, only sagittal vertical axis had a slight inverse dose-dependent association with the physical component of SF-36 (Table 4). In women, an inverse dose-response relation was found between sagittal vertical axis, pelvic tilt or pelvic tiltpelvic incidence ratio and physical health-related quality of life scores. A direct doseresponse association was found between lumbar lordosis-pelvic incidence ratio and scores in the physical component of SF-36. The largest mean difference (8.8) was found in sagittal vertical axis, while from other previous parameters, the highest mean difference of physical component z-score ranged between 4.1 and 4.3. In order to assess the combined effect of sagittal balance and alignment parameters among women, thirds of sagittal vertical axis and the thirds of each assessed individual alignment parameter were combined. Figure 1 depicts the median z-score on the physical component of SF-36 in women classified in nine different groups. None of the examined sagittal alignment parameters was relevantly associated with scores in the physical component of SF-36 within thirds of sagittal balance, meaning that alignment was not associated with quality of life when balance was taken into account. Conversely, sagittal vertical axis was generally inversely related with physical healthrelated quality of life, independently of the alignment parameters measured. However, after adjustment for age, education and body mass index, none of radiographic parameters of women were associated with physical health-related quality of life. When sensitivity analysis on back pain variables was performed excluding participants that had missing information on SF-36, the same conclusions were 34 supported with the exception of sacral slope thirds that were not associated with back pain variables among women. Discussion In the general adult population, individual parameters of sagittal spino-pelvic alignment and balance were consistently associated with meaningful health outcomes among women, particularly regarding back pain severity and the physical component of health-related quality of life. In contrast, isolated parameters of sagittal posture did not show consistent associations with quality of life measures among men. Sagittal postural patterns were not associated with quality of life measures in men or women of the general adult population. Possibly the most important limitation of this study is its cross-sectional nature, in that the temporal sequence between sagittal standing posture and quality of life measures could not be evaluated, since postural adaptations to back pain presence or severity are likely to exist. Posture parameters could then be a consequence rather than a cause of pain or decreased quality of life. Another important issue regarding exposure assessment is the possibility of high inter-examiner variability regarding procedures of radiographic evaluation, but it is not expected that systematic variability has influenced our main conclusions, since the assignment of radiology technician was performed merely based in logistic reasons. While some postural parameters were related with quality of life measures in men, a relevant consistent effect of isolated sagittal standing posture did not seem to exist. These results can be explained either by a true gender heterogeneity in the effects quantified, which would argue for a biological dissociation of posture from pain and quality of life only among men, or by sex differences in the reporting of subjective outcome measures such as the ones we assessed. Given that men reported back pain less frequently as well as higher health-related quality of life scores, smaller effect sizes among men are another important issue that may have limited our statistical power to identify significant associations. This is in agreement with the lower frequency of musculoskeletal disorders and consequent impact on pain and quality of life that has often been reported in men comparatively to women (33-35). Higher previous month prevalence of back pain and back pain severity were found in women presenting pelvic incidence and sacral slope values outside the neutral range, especially among those exhibiting the highest values. However, those results do not seem to generally imply a worse health-related quality of life. Pelvic incidence is 35 well recognized as the foundation of overall standing sagittal alignment (27, 36-38), and pelvic incidence values outside the neutral range were already suggested to be related with specific shear mechanical forces that may promote back pain development (27, 39). Particularly, high pelvic incidence values induce vertebral listhesis and are positively related with the severity of slip grade (4, 40). Additionally, it is not surprising that sacral slope has a similar influence on back pain variables, because it is highly determined by pelvic incidence (27, 36-38). In women, higher sagittal balance (forward displacement of the spine over the hips) was the parameter most strongly associated with back pain presence and severity, resulting in a clear gradient of decreasing physical health-related quality of life with increasing sagittal balance. These results are supported by previous observations in clinical samples showing that sagittal balance is strongly associated with pain and physical disability (8-12, 15). Furthermore, it was suggested that the documented influence of sagittal balance on symptoms and quality of life would also explain the observed association of increased pelvic tilt (10, 12) and decreased lumbar lordosis (9, 10, 12) with worse quality of life measures. As sagittal balance progressively increases, compensatory pelvic retroversion takes place in order to restore a mechanically efficient posture, and consequently, lumbar lordosis angle will decrease. Our results support this compensatory mechanism among adult women of the general population, where an association with back pain severity was observed for pelvic tilt and pelvic retroversion, possibly accounting for the inverse dose-response relation of those parameters with the physical component of quality of life. Additionally, smaller lumbar lordosis was related with worse physical health-related quality of life, especially when the degree of lumbar lordosis was measured taking into account individual pelvic incidence. Furthermore, of all parameters, sagittal balance seemed to be the one with the most important potential effect regarding the impairment of physical health-related quality of life among women, likely reflecting the major determinant role of an increased sagittal vertical axis in changing alignment parameters. However, the relation of sagittal standing posture with the physical component of SF-36 was partly explained by women’s characteristics such as age, education and body mass index, since those are major determinants of sagittal balance (41, 42). The lack of relation between sagittal standing posture and the mental health-related quality of life, sustains that the influence of posture in the physical component of quality of life is likely mediated by the presence and severity of back pain. Although it has already been shown that the type 2 postural pattern is more frequent in chronic low back patients comparatively to control subjects (28), our results do not support the usefulness of sagittal postural patterns to predict quality of life 42 Table 4 – Mean (standard deviation) of health-related quality of life according to thirds of individual sagittal radiographic parameters and types of postural patterns, separately by sex. Men Women SF-36 health-related quality of life SF-36 health-related quality of life Parameter/Thirds* Physical component zscore P Mental component zscore P Physical component zscore P Mental component zscore P Pelvic incidence 1st 51.8 (9.3) 0.601 52.2 (8.4) 0.885 49.8 (10.0) 0.303 47.3 (10.1) 0.136 2nd 52.7 (8.6) 52.7 (8.5) 47.5 (10.2) 50.3 (9.3) 3rd 53.7 (9.2) 51.8 (9.3) 47.8 (9.9) 48.2 (10.7) Pelvic tilt 1st 52.0 (9.9) 0.794 51.5 (9.2) 0.504 50.8 (9.9) 0.019 48.8 (9.5) 0.861 2nd 53.2 (7.8) 53.4 (8.0) 47.5 (10.1) 49.0 (10.1) 3rd 52.9 (9.4) 51.8 (8.8) 46.7 (9.7) 48.1 (10.7) Sacral Slope 1st 51.6 (9.1) 0.465 51.7 (9.7) 0.764 48.3 (10.2) 0.740 48.3 (9.8) 0.687 2nd 52.7 (9.4) 52.1 (8.7) 47.8 (10.0) 48.2 (9.7) 3rd 53.9 (8.5) 53.0 (7.5) 49.0 (10.1) 49.5 (10.8) Lumbar lordosis 1st 50.9 (9.1) 0.174 51.9 (9.5) 0.397 47.8 (10.4) 0.113 48.2 (10.0) 0.328 2nd 54.2 (8.8) 53.5 (6.7) 47.2 (10.0) 49.9 (8.8) 3rd 53.0 (9.0) 51.2 (9.7) 50.3 (9.6) 47.7 (11.4) Sagittal vertical axis 1st 54.6 (8.3) 0.059 52.4 (8.6) 0.033 53.0 (9.9) <0.001 48.0 (10.4) 0.787 2nd 53.2 (8.4) 50.0 (9.1) 47.9 (9.3) 48.8 (10.0) 3rd 50.3 (9.9) 54.5 (7.7) 44.2 (9.0) 49.1 (9.8) Lumbar lordosis-pelvic incidence ratio 1st 52.9 (9.1) 0.973 51.6 (8.3) 0.143 46.0 (10.0) 0.019 49.0 (10.3) 0.251 2nd 52.7 (8.8) 54.2 (7.1) 48.8 (9.8) 47.1 (10.4) 3rd 52.5 (9.3) 50.9 (10.1) 50.3 (10.0) 49.7 (9.5) Pelvic tilt-pelvic incidence ratio 1st 52.4 (9.6) 0.378 51.9 (8.5) 0.189 50.4 (10.2) 0.026 47.6 (10.0) 0.331 2nd 54.1 (8.1) 53.9 (7.8) 48.4 (10.1) 48.3 (10.6) 3rd 51.6 (9.3) 50.8 (9.5) 46.2 (9.5) 49.9 (9.6) Types of postural patterns† 1 42.6 (33.6-54.5) 0.698 50.8 (47.5-59.5) 0.843 48.6 (34.9-52.1) 0.550 48.9 (39.2-55.7) 0.677 2 54.1 (47.0-58.9) 54.1 (47.5-58.9) 49.0 (40.7-56.6) 50.1 (41.1-55.4) 3 55.3 (48.7-60.5) 53.8 (47.3-60.1) 49.0 (41.0-54.5) 50.0 (42.2-55.4) 4 53.4 (46.9-61.5) 51.1 (45.9-57.2) 51.0 (40.6-58.0) 52.9 (39.1-59.5) SF-36, Short Form 36. *Cut offs used to define thirds of individual parameters: Pelvic incidence (º), 49.3 and 59.1 in men or 48.9 and 59.1 in women; Pelvic tilt (º), 12.8 and 18.4 in men or 12.7 and 19.5 in women; Sacral slope (º), 34.3 and 42.8 in men or 35.1 and 41.6 in women; Lumbar lordosis (º), 56.5 and 66.4 in men or 57.3 and 65.9 in women; Sagittal vertical axis (mm), -31.2 and -1.9 in men or -34.2 and -4.5 in women; Lumbar lordosis-pelvic incidence ratio, 1.05 and 1.22 in men or 1.05 and 1.23 in women; Pelvic tiltpelvic incidence ratio, 0.26 and 0.33 in men or 0.25 and 0.34 in women. †Values are reported as median (25th-75th percentiles). Bold type indicates statistical significance. 43 Figure 1.Combined effect of sagittal vertical axis thirds and each alignment parameter thirds on physical health-related quality of life (Short Form 36), in women. 20 30 40 50 60 1st 2nd 3rd 1st 2nd 3rd 1st 2nd 3rd 1st 2nd 3rd 1st 2nd 3rd 1st 2nd 3rd Sagittal vertical axis thirds Physical component z-score (median) 1st third of alignment 2nd third of alignment 3rd third of alignment Pelvic incidence p=0.454 p=0.256 p=0.987 p=0.430 p=0.761 p=0.512 p=0.933 p=0.135 p=0.660 p=0.926 p=0.773 p=0.311 p=0.628 p=0.561 p=0.486 p=0.186 p=0.292 p=0.835 Pelvic tilt Sacral slope Lumbar lordosis Lumbar lordosis-pelvic incidence ratio Pelvic tilt-pelvic incidence ratio 20 30 40 50 60 1st 2nd 3rd 1st 2nd 3rd 1st 2nd 3rd 1st 2nd 3rd 1st 2nd 3rd 1st 2nd 3rd Pelvic incidence thirds Pelvic tilt thirds Sacral slope thirds Lumbar lordosis thirds Lumbar lordosis-pelvic incidence ratio thirds Pelvic tilt-pelvic incidence ratio thirds Physical component z-score (median) 1st third of sagittal vertical axis 2nd third of sagittal vertical axis 3rd third of sagittal vertical axis p<0.001 p=0.025 p=0.011 p=0.004 p=0.010 p=0.039 p=0.031 p=0.001 p=0.001 p=0.006 p=0.005 p=0.024 p=0.012 p=0.053 p=0.002 p=0.002 p=0.001 p=0.240 44 45 Chapter II Individual and contextual characteristics as determinants of sagittal standing posture: a population-based study among adults. 46 47 Abstract Objectives: To estimate the association between sociodemographic, anthropometric and behavioral characteristics with sagittal standing posture among adults from the general population. Methods: As part of the EpiPorto study, 489 adults were assessed during 2005-2008. Radiographic data collection consisted of 36-inch sagittal radiographs obtained in free-standing posture, from where individual spinopelvic parameters were recorded and participants were classified into one of four types of sagittal postural patterns. Data regarding age, sex, education, occupation, body mass index (BMI), waist circumference, waist-hip ratio, total physical activity, leisuretime physical activity, time spent in sitting position, smoking status and tobacco consumption, were collected. Results: Older age, lower formal education, blue collar occupation, overall and central obesity were associated with increased sagittal vertical axis and pelvic tilt-pelvic incidence ratio. Comparatively to the respective reference categories and to the neutral postural pattern (type 3), independently of age, sex, education, BMI, total physical activity and smoking status, overweight adults had higher odds of type 2 (odds ratio [OR]=1.92) and type 4 postural patterns (OR=2.13). Being obese was positively related with type 1 postural pattern (OR=6.10), central obesity (waist circumference) was associated with type 1 postural pattern (OR=3.45), while those with higher waist-hip ratio had lower odds of showing a type 4 pattern (OR=0.52). There was also a weak direct association between female sex and a type 1 postural pattern. Regarding behavioral factors, subjects in the 2nd and the 3rd thirds of total physical activity exhibited all non-neutral postural patterns less frequently, and current smoker were more likely to present a type 4 postural pattern. Conclusion: Higher BMI and central obesity were important potential determinants of non-neutral posture among adults from the general population. Strategies focused in reducing BMI are expected to promote the prevention of all non-neutral pathologic standing postures at a population level. Keywords: Posture, Spine, Pelvis, Body Mass Index, Obesity, Population Survey. Introduction Reciprocal physiologic sagittal curvatures of the spine are arranged in order to obtain the mechanically most efficient posture (1, 2). One of the fundamental regulators of the overall sagittal alignment of the spine is pelvic morphology, mainly through influencing lumbar lordosis contours (1-5). In the presence of sagittal imbalance, pelvic 48 retroversion is a key compensatory physiological mechanism that allows for adapting overall alignment in order to restore sagittal balance (6, 7). The clinical relevance of sagittal spino-pelvic alignment as a determinant of musculoskeletal symptoms and quality of life has been shown in samples of patients with diverse spinal conditions (8-11). Among the complex set of parameters that compose standing posture, sagittal balance and pelvic tilt were identified as those most strongly associated with pain and physical disability (9, 10). Particularly, low back pain patients frequently exhibit smaller lumbar lordosis and a more vertical sacrum comparatively to control subjects (12-14). In studies of adult samples selected based on the absence of clinically relevant orthopedic conditions, sagittal spino-pelvic alignment was found to be associated with age (2, 6, 15, 16), sex (2, 17, 18) and body mass index (BMI) (3), but current evidence is still unclear and conflicting between studies. The development of a non-neutral sagittal posture is recognized as a complex occurrence, reflecting exposure to diverse interrelated factors, at the individual and contextual levels. In agreement, the association between lower socioeconomic status and clinical musculoskeletal manifestations is well-known, and probably mediated in part by behavioral characteristics (19-22). However, evidence assessing the possible link between contextual variables and upstream sagittal standing posture phenotypes is lacking. Additionally, previous evidence on standing alignment has originated from highly-selected samples of patients or healthy subjects, possibly excluding an important fraction of population variability and limiting the generalizability of findings regarding the general adult population. Recently, it was suggested that the study of non-neutral sagittal alignment etiology should be based on overall postural patterns rather than on single alignment parameters (18, 23, 24). Nevertheless, the use of postural patterns has been rare, namely regarding their population frequency and potential determinants. Using data from a population-based sample of adults, our aim was to estimate the association of sociodemographic, anthropometric and behavioral characteristics with sagittal posture, considering both individual alignment parameters and standing postural patterns. Participants and Methods Participants were evaluated as part of the EPIPorto cohort study, which comprises a sample of Portuguese adults, residents in the city of Porto. As previously described (25), recruitment was performed in 1999-2003 by random digit dialing using 49 households as the sampling frame. In each household identified, a resident aged 18 years or more was randomly selected for evaluation without replacement if a refusal occurred. A participation proportion of 70.0% was initially achieved and 1682 (67.7%) subjects were reevaluated during 2005-2008. Of these, 518 subjects consecutively evaluated by means of radiographic assessment were eligible for the present study. Subjects were excluded if they had instrumentation of the spine (n=2), hip arthroplasty (n=10), and inflammatory spinal arthropathies (n=17). The final sample included 489 subjects that were invited to visit the Department of Hygiene and Epidemiology of the University of Porto Medical School, where evaluations were carried out by trained health professionals following a standardized protocol. The sample size allowed for the estimation of a 5% prevalence of a type 1 postural pattern, with a 2% precision and an 80% power, at a 95% confidence level. The local ethics committee of Hospital S. João approved the study protocol. All participants gave written informed consent to participate in the study. Sagittal spino-pelvic alignment Radiographic data collection consisted of 36-inch standing sagittal radiographs obtained in free-standing posture, performed by one of eight radiology technicians according to standard operating procedures. All radiographic films were digitized using a Vidar scanner (Vidar Systems Corp., Herndon, VA, USA) with 75 dpi resolution and 12 gray levels, and then analyzed using a spine dedicated software with a precision of 0.1º in angles and 0.1 mm in distances (Sectra Imtec AB, Linköping, Sweden). Radiographic spino-pelvic measures were recorded by a single physical therapist who was trained to the effect and remained blind regarding sociodemographic, anthropometric and behavioral participants’ characteristics. Measured parameters included lumbar lordosis, number of vertebrae in lumbar lordosis, sagittal vertical axis (negative when posterior to the posterosuperior corner of S1), pelvic incidence, pelvic tilt and sacral slope (Figure 1). Lumbar lordosis and the number of vertebrae in lumbar lordosis were assessed taking into account the point where the orientation of the spinal curvatures changed (inflexion point) (4). Since pelvic parameters are geometrically related (5), pelvic tilt-pelvic incidence ratio was computed to indicate the extent of pelvic retroversion for a given pelvic incidence. Sagittal postural pattern was then determined for each participant in respect to the classification of Roussouly et al (4), based primarily on sacral slope, and then, on the number of vertebrae in lumbar lordosis, as recently suggested (12). Specifically, a type 1 was assigned if sacral slope less than 35.5º and number of vertebrae in lumbar lordosis less or equal to three; type 50 2 if sacral slope less than 35.5º and number of vertebrae in lumbar lordosis more than three; type 3 if sacral slope between 35.5º and 44.4º, and a type 4 if sacral slope more or equal than 44.5º. Sociodemographics Age was categorized in three groups: less than 40, 40-64 and 65 years or over. Education was recorded as completed years of schooling and aggregated in three categories: up to 4, 5-9, 10 years or higher. Occupations were classified by major professional groups, according to the National Classification of Occupations (version 1994) (26) and grouped in three categories: blue collar, lower white collar and upper white collar. A more detailed description of each occupational group has been already published (27). Housewives (n=46) were analyzed separately and six students were not considered in the occupational analysis. Anthropometrics Anthropometric data were obtained with participants in light indoor clothing and barefoot. Body weight was measured to the nearest 0.1 kg using a digital scale (SECA, Columbia, USA), and height was measured to the nearest centimeter with a wall stadiometer (SECA, Hamburg, Germany) in standing and sitting (standard stool) positions. BMI was calculated as weight (kilograms) divided by squared height (meters), and categorized in three groups: normal weight (BMI less than 25.0 kg/m2), overweight (BMI 25.0 to 29.9 kg/m2) and obese (BMI equal to or above 30.0 kg/m2) (28). Waist circumference was measured midway between the lower limit of the rib cage and the iliac crest and hip circumference measured over the femoral trochanters, both to the nearest centimeter. Participants were in standing and a flexible and nondistensible tape was used to avoid exerting pressure on tissues. Waist-hip ratio was computed. Central obesity was considered present if waist circumference was equal to or above 102.0 cm for men and 88.0 cm for women, and alternatively, if waisthip ratio was equal to or above 0.95 for men and 0.85 for women (29). Behavioral characteristics Physical activity was assessed utilizing a previously validated questionnaire (30). Standard metabolic energy equivalent task (MET) values were used to calculate energy expenditure compared with the resting metabolic rate (31). Daily activities were classified as very light, light, moderate and heavy intensity, corresponding to an average of 1.5, 2.5, 5.0 and 7.0 MET, respectively (32). Energy expenditure was estimated through multiplication of the corresponding MET value by the time spent in 51 each activity. Total physical activity comprised all activities during the entire day (sleep, work, household chores and leisure time activities), while leisure-time physical activity included leisure time activities only (sedentary activities such as watching television and different types of exercises). Sex-specific thirds were created for total and leisuretime physical activity. Participants also self-reported the average time spent per day in sitting position during leisure time, and the median of the distribution was used to classify exposure. Participants were classified as never-smokers, former smokers (having quitted smoking for at least six months) and current smokers (including daily and occasional smokers) (33). Tobacco consumption over the life course was calculated as the result of multiplying the number of cigarette packs smoked in a year by the number of smoking years. Categorization was then performed: never smokers, lower and higher tobacco consumption (cut off based on the median of the distribution among those who had ever smoked). Statistical analysis Descriptive data for radiographic spino-pelvic parameters were reported as median (25th percentile; 75th percentile). The statistical significance of differences in radiographic parameters between categories of participants’ characteristics were assessed through Mann-Whitney or Kruskal-Wallis tests (as appropriate), and between types of sagittal postural patterns were assessed through the Kruskal-Wallis test. Proportions of types of sagittal postural patterns within categories of participants’ characteristics were compared using the χ2 test. Crude and adjusted odds ratios (OR) with 95% confidence intervals (95%CI) were estimated by multinomial logistic regression to quantify the associations of sociodemographic, anthropometric and behavioral characteristics with the non-neutral sagittal postural patterns, having type 3 (neutral) as reference. Associations were additionally adjusted for age, sex, education, BMI, total physical activity and smoking status. Results Table 1 describes individual radiographic parameters according to categories of each sociodemographic, anthropometric and behavioral characteristic. Median pelvic incidence was higher in less educated participants and among blue-collar workers and obese participants (although in the latter two not statistically significant). Sagittal balance was more anteriorly displaced in older adults, and among those with lower 58 28. National Institutes of Health, National Heart Lung and Blood Institute. Clinical Guidelines on the Identification, Evaluation, and Treatment of Overweight and Obesity in Adults--The Evidence Report. National Institutes of Health. Obes Res. 1998;6(Suppl 2):51S-209S. 29. World Health Organization. Waist circumference and waist–hip ratio: report of a WHO expert consultation, Geneva, 8–11, December 2008. Geneva, Switzerland: World Health Organization. 2011. 30. Camões M, Severo M, Santos AC, Barros H, Lopes C. Testing an adaptation of the EPIC physical activity questionnaire in Portuguese adults: a validation study that assesses the seasonal bias of self-report. Ann Hum Biol. 2010;37(2):185-97. 31. Blair SN, Haskell WL, Ho P, Paffenbarger RS Jr, Vranizan KM, Farquhar JW, et al. Assessment of habitual physical activity by a seven-day recall in a community survey and controlled experiments. Am J Epidemiol. 1985;122(5):794-804. 32. Subcommittee on the Tenth Edition of the Recommended Dietary Allowances, Food and Nutrition Board, Commission on Life Sciences, National Research Council. Recommended dietary allowances. 10th ed. Washington, DC: National Research Council - National Academy Press; 1989. 33. World Health Organization. Guidelines for controlling and monitoring the tobacco epidemic. Geneva, Switzerland: World Health Organization. 1998. 34. Smith AJ, O'Sullivan PB, Beales DJ, de Klerk N, Straker LM. Trajectories of childhood body mass index are associated with adolescent sagittal standing posture. Int J Pediatr Obes. 2011;6(2-2):e97-106. 35. Strand BH, Cooper R, Hardy R, Kuh D, Guralnik J. Lifelong socioeconomic position and physical performance in midlife: results from the British 1946 birth cohort. Eur J Epidemiol. 2011;26(6):475-83. 36. Kuh D, Bassey EJ, Butterworth S, Hardy R, Wadsworth ME, Musculoskeletal Study Team. Grip strength, postural control, and functional leg power in a representative cohort of British men and women: associations with physical activity, health status, and socioeconomic conditions. J Gerontol A Biol Sci Med Sci. 2005;60(2):224-31. 37. Houston DK, Stevens J, J. C. Abdominal fat distribution and functional limitations and disability in a biracial cohort: the Atherosclerosis Risk in Communities Study. Int J Obes (Lond). 2005;29(12):1457-63. 38. Backholer K, Wong E, Freak-Poli R, Walls HL, Peeters A. Increasing body weight and risk of limitations in activities of daily living: a systematic review and metaanalysis. Obes Rev. 2012;13(5):456-68. 59 39. Rantanen T, Era P, Heikkinen E. Physical activity and the changes in maximal isometric strength in men and women from the age of 75 to 80 years. J Am Geriatr Soc. 1997;45(12):1439-45. 40. Langsetmo L, Hitchcock CL, Kingwell EJ, Davison KS, Berger C, Forsmo S, et al. Physical activity, body mass index and bone mineral density-associations in a prospective population-based cohort of women and men: the Canadian Multicentre Osteoporosis Study (CaMos). Bone. 2012;50(1):401-8. 41. Hongo M, Miyakoshi N, Shimada Y, Sinaki M. Association of spinal curve deformity and back extensor strength in elderly women with osteoporosis in Japan and the United States. Osteoporos Int. 2012;23(3):1029-34. 42. Sinaki M, Itoi E, Rogers JW, Bergstralh EJ, Wahner HW. Correlation of back extensor strength with thoracic kyphosis and lumbar lordosis in estrogen-deficient women. Am J Phys Med Rehabil. 1996;75(5):370-4. 43. Endo K, Suzuki H, Nishimura H, Tanaka H, Shishido T, Yamamoto K. Sagittal lumbar and pelvic alignment in the standing and sitting positions. J Orthop Sci. 2012;[Epub ahead of print]. 44. Lord MJ, Small JM, Dinsay JM, Watkins RG. Lumbar lordosis. Effects of sitting and standing. Spine. 1997;22(21):2571-4. 45. Rapuri PB, Gallagher JC, Smith LM. Smoking is a risk factor for decreased physical performance in elderly women. J Gerontol A Biol Sci Med Sci. 2007;62(1):93100. 46. Strand BH, Mishra G, Kuh D, Guralnik JM, Patel KV. Smoking history and physical performance in midlife: results from the British 1946 birth cohort. J Gerontol A Biol Sci Med Sci. 2011;66(1):142-9. 47. Wüst RC, Morse CI, de Haan A, Rittweger J, Jones DA, Degens H. Skeletal muscle properties and fatigue resistance in relation to smoking history. Eur J Appl Physiol. 2008;104(1):103-10. 48. Wong PK, Christie JJ, Wark JD. The effects of smoking on bone health. Clin Sci. 2007;113(5):233-41. 49. Ensrud KE, Black DM, Harris F, Ettinger B, Cummings SR. Correlates of kyphosis in older women. The Fracture Intervention Trial Research Group. J Am Geriatr Soc. 1997;45(6):682-7. 50. Schneider DL, von Mühlen D, Barrett-Connor E, Sartoris DJ. Kyphosis does not equal vertebral fractures: the Rancho Bernardo study. J Rheumatol. 2004;31(4):74752. 60 Table 1 – Median (25th percentile; 75th percentile) of radiographic sagittal spino-pelvic parameters according to categories of sociodemographic, anthropometric and behavioral characteristics. Characteristic All participants Pelvic incidence P Pelvic tilt P Sacral slope P Lumbar lordosis P Sagittal vertical axis P Pelvic tilt-pelvic incidence ratio P n (%) All participants (n=489) 54.1 (46.6; 62.6) 15.8 (11.2; 21.1) 38.7 (32.8; 43.7) 61.6 (54.5; 69.2) -15.9 (-38.8; 6.2) 0.29 (0.23; 0.36) Sociodemographic Age – years (n=489) < 40 57 (11.7) 55.4 (44.9; 63.8) 0.139 15.2 (9.0; 21.0) 0.001 39.8 (34.2; 45.2) 0.167 64.0 (57.0; 71.2) 0.184 -39.6 (-59.7; -20.1) < 0.001 0.27 (0.19; 0.33) < 0.001 40-64 238 (48.7) 51.8 (45.8; 61.8) 14.9 (10.6; 19.6) 39.1 (32.9; 43.2) 61.5 (54.5; 68.3) -26.3 (-47.0; -7.5) 0.28 (0.22; 0.35) ≥ 65 194 (39.7) 55.4 (47.5; 63.1) 17.0 (12.5; 22.6) 37.7 (32.0; 43.9) 61.2 (53.2; 68.8) 2.1 (-18.2; 23.0) 0.32 (0.25; 0.38) Sex (n=489) Men 178 (36.4) 54.1 (47.0; 62.3) 0.693 15.9 (11.3; 19.9) 0.586 38.7 (32.6; 44.3) 0.838 61.0 (54.5; 68.7) 0.603 -14.9 (-36.9; 6.68) 0.508 0.29 (0.23; 0.35) 0.602 Women 311 (63.6) 54.1 (45.9; 63.1) 15.6 (11.1; 21.8) 38.8 (33.2; 43.3) 62.1 (54.6; 69.4) -16.8 (-39.9; 5.7) 0.29 (0.22; 0.36) Education – years (n=489) ≤ 4 184 (37.6) 56.4 (48.3; 65.0) 0.006 17.0 (12.5; 22.8) 0.001 39.6 (32.9; 44.1) 0.272 61.6 (54.1; 71.0) 0.567 -6.3 (-26.3; 15.5) < 0.001 0.31 (0.24; 0.38) 0.015 5-9 137 (28.0) 52.6 (44.5; 59.4) 15.2 (10.6; 19.6) 37.4 (31.5; 43.5) 60.9 (52.6; 68.3) -24.3 (-42.8; 3.0) 0.28 (0.22; 0.35) ≥ 10 168 (34.4) 52.5 (45.0; 63.0) 15.3 (9.7; 20.2) 39.2 (33.5; 43.6) 63.2 (55.2; 68.6) -26.3 (-48.3; -5.9) 0.29 (0.22; 0.36) Occupation (n=483)* Blue collar 147 (30.4) 56.1 (48.7; 64.9) 0.052 16.7 (12.9; 22.9) 0.016 39.7 (33.7; 44.6) 0.361 63.3 (54.6; 70.3) 0.414 -8.1 (-32.1; 12.0) 0.002 0.32 (0.24; 0.37) 0.048 Lower white collar 133 (27.5) 53.1 (43.8; 62.3) 15.0 (10.7; 20.0) 38.9 (32.7; 43.1) 61.1 (54.1; 70.1) -19.8 (-47.7; 9.3) 0.28 (0.22; 0.35) Upper white collar 157 (32.5) 52.6 (46.1; 61.2) 15.3 (10.1; 20.1) 38.0 (33.1; 43.6) 62.1 (54.9; 67.8) -25.2 (-43.2; -0.2) 0.29 (0.22; 0.36) Housewives 46 (9.5) 53.9 (44.7; 58.4) 16.1 (11.1; 21.1) 37.1 (29.8; 42.9) 58.8 (49.2; 68.6) -11.7 (-33.2; 7.9) 0.32 (0.21; 0.37) Anthropometric BMI (n=489)† Normal weight 170 (34.8) 52.2 (46.5; 59.4) 0.141 14.2 (8.38; 18.1) < 0.001 39.2 (34.5; 43.5) 0.503 61.9 (54.9; 69.5) 0.644 -25.2 (-44.8; 4.1) 0.067 0.26 (0.19; 0.33) < 0.001 Overweight 211 (43.1) 54.4 (45.9; 63.3) 16.5 (12.2; 22.4) 38.3 (32.1; 43.9) 61.4 (53.7; 68.4) -14.7 (-36.6; 8.1) 0.32 (0.26; 0.37) Obese 108 (22.1) 56.5 (46.6; 64.4) 17.1 (12.3; 22.4) 38.6 (31.3; 42.6) 61.8 (54.7; 71.2) -13.1 (-34.7; 4.4) 0.31 (0.24; 0.38) Waist circumference (n=487)‡ No central obesity 280 (57.5) 53.5 (45.5; 61.5) 0.228 15.1 (10.0; 19.5) < 0.001 39.4 (33.7; 44.0) 0.101 62.2 (54.9; 69.3) 0.566 -22.2 (-42.9; 1.8) < 0.001 0.28 (0.21; 0.34) < 0.001 Central obesity 207 (42.5) 54.6 (47.2; 63.6) 17.2 (12.5; 22.5) 38.2 (31.1; 42.9) 61.3 (53.3; 68.6) -10.5 (-32.1; 11.4) 0.32 (0.25; 0.39) Waist-hip ratio (n=487)§ No central obesity 132 (27.1) 54.0 (46.7; 63.2) 0.931 14.6 (9.7; 20.0) 0.008 40.1 (34.8; 44.9) 0.012 63.8 (54.9; 71.3) 0.042 -32.0 (-56.9; -7.4) < 0.001 0.27 (0.20; 0.33) < 0.001 Central obesity 355 (72.9) 54.0 (46.3; 62.4) 16.2 (11.8; 21.5) 38.3 (32.0; 43.2) 61.1 (54.3; 67.8) -11.7 (-34.5; 9.0) 0.31 (0.24; 0.37) Behavioral Total PA (n=430)|| 1st third 133 (30.9) 53.9 (45.1; 60.1) 0.763 15.4 (11.4; 20.9) 0.991 37.9 (31.1; 44.0) 0.528 61.8 (53.3; 68.0) 0.397 -10.5 (-36.9; 10.8) 0.010 0.30 (0.23; 0.37) 0.543 2nd third 158 (36.7) 54.8 (46.6; 62.3) 15.8 (11.2; 20.0) 39.0 (33.7; 43.4) 62.1 (55.1; 69.5) -15.6 (-39.7; 7.5) 0.28 (0.22; 0.35) 3rd third 139 (32.3) 53.0 (45.9; 63.6) 15.6 (10.8; 20.9) 39.0 (32.7; 43.5) 61.4 (54.7; 68.6) -23.7 (-43.5; -7.2) 0.30 (0.23; 0.36) Leisure-time PA (n=483)¶ 1st third 135 (28.0) 53.2 (47.0; 63.7) 0.345 16.1 (11.8; 21.0) 0.117 40.0 (32.8; 43.9) 0.597 63.0 (54.9; 71.4) 0.274 -21.6 (-43.5; -3.3) 0.038 0.30 (0.23; 0.35) 0.226 2nd third 188 (38.9) 52.9 (45.8; 60.9) 15.2 (10.5; 20.6) 38.4 (32.5; 43.4) 61.6 (53.3; 68.4) -19.2 (-40.5; 4.7) 0.29 (0.21; 0.36) 3rd third 160 (33.1) 55.5 (46.8; 63.0) 16.3 (11.5; 21.8) 38.6 (33.4; 44.1) 61.3 (53.4; 67.6) -11.6 (-34.7; 10.7) 0.30 (0.24; 0.36) Sitting position (n=487)** Below or at the median 248 (50.9) 53.2 (45.9; 63.3) 0.851 15.3 (10.8; 20.2) 0.140 39.5 (33.3; 43.9) 0.187 62.3 (55.4; 70.4) 0.060 -21.6 (-43.4; -1.5) 0.002 0.29 (0.22; 0.35) 0.065 Above the median 239 (49.1) 54.9 (47.0; 62.1) 16.2 (11.4; 21.6) 38.1 (32.0; 43.4) 61.1 (52.5; 67.5) -12.0 (-34.5; 9.9) 0.30 (0.23; 0.36) Smoking status (n=488)†† Never 297 (60.9) 54.4 (46.6; 62.7) 0.611 16.2 (11.3; 21.6) 0.564 39.2 (32.9; 43.6) 0.401 61.2 (54.7; 69.3) 0.730 -13.5 (-37.8; 9.8) < 0.001 0.30 (0.23; 0.36) 0.430 Former 115 (23.6) 51.0 (47.0; 62.2) 15.5 (11.3; 19.2) 37.9 (32.1; 43.3) 61.9 (54.5; 67.6) -12.6 (-34.5; 3.6) 0.29 (0.22; 0.34) Current 76 (15.6) 55.1 (45.0; 63.8) 15.4 (10.8; 20.1) 39.9 (33.5; 45.0) 63.7 (52.8; 71.1) -32.7 (-48.2; -14.0) 0.27 (0.22; 0.36) Tobacco consumption (n=481)‡‡ None 297 (61.7) 54.4 (46.6; 62.7) 0.700 16.2 (11.3; 21.6) 0.470 39.2 (32.9; 43.6) 0.578 61.2 (54.7; 69.3) 0.736 -13.5 (-37.8; 9.8) 0.110 0.30 (0.23; 0.36) 0.383 Lower 92 (19.1) 50.6 (44.6; 63.7) 16.2 (11.0; 20.0) 37.7 (32.1; 42.9) 61.7 (53.1; 69.1) -21.6 (-43.6; 2.5) 0.29 (0.22; 0.36) Higher 92 (19.1) 54.7 (47.2; 62.1) 15.3 (11.2; 18.9) 38.8 (32.9; 44.9) 63.2 (54.4; 69.3) -24.6 (-39.2; -0.8) 0.28 (0.22; 0.35) BMI, body mass index; PA, physical activity. *Six students were not considered in the analysis. †Normal weight: BMI < 25.0 kg/m2; Overweight: 25.0 kg/m2 ≤ BMI < 30.0 kg/m2; Obese ≥ 30.0 kg/m2. ‡Central obesity defined as waist circumference ≥ 102.0 cm in men and ≥ 88.0 cm in women. §Central obesity defined as waist-hip ratio ≥ 0.90 in men and ≥ 0.85 in women. ||Thirds of total PA (METS-h/day): < 1.39, 1.39-1.57, > 1.57 for men and < 1.42, 1.42-1.50, > 1.50 for women. ¶Thirds of leisure-time PA (METS-h/day): < 4.50, 4.50-7.23, > 7.23 for men and < 3.00, 3.00-5.92, > 5.92 for women. 61 **Median equal to 2.50 hours h/day. ††Never: participant that never smoked; Former: participant that stopped smoking for at least 6 months; Current: participant that smoke daily or occasionally. ‡‡None: participant that never smoked; Lower: participant that smoked ≤ 7547.03 cigarette packs over life; Higher: participant that smoked > 7547.03 cigarette packs over life. 62 Table 2 – Radiographic sagittal spino-pelvic parameters in types of postural patterns.* Sagittal postural pattern Parameter Type 1 (n=24 [4.9%]) Type 2 (n=153 [31.3%]) Type 3 (n=207 [42.3%]) Type 4 (n=105 [21.5%]) P Pelvic incidence – ° 46.5 (37.5; 54.7) 44.7 (39.7; 50.0) 55.4 (50.0; 60.3) 67.5 (62.3; 75.0) < 0.001 Pelvic tilt – ° 21.0 (10.8; 27.4) 14.9 (10.7; 18.9) 15.3 (10.7; 19.9) 17.4 (14.6; 26.1) < 0.001 Sacral slope – ° 27.4 (24.7; 32.1) 30.6 (27.6; 33.4) 40.2 (38.2; 42.1) 48.4 (45.9; 51.3) < 0.001 Lumbar lordosis – ° 44.4 (38.3; 49.6) 53.7 (48.1; 58.8) 63.3 (58.4; 67.7) 73.3 (68.3; 80.0) < 0.001 Vertebrae in lumbar lordosis 3.0 (3.0; 3.0) 5.0 (4.0; 5.0) 5.0 (4.0; 5.0) 5.0 (4.0; 6.0) < 0.001 Sagittal vertical axis – mm -4.4 (-20.0; 26.6) -19.2 (-39.9; -3.9) -18.9 (-40.5; 4.8) -14.7 (-37.5; 12.3) 0.068 Pelvic tilt-pelvic incidence ratio 0.44 (0.24; 0.49) 0.33 (0.27; 0.39) 0.27 (0.21; 0.33) 0.27 (0.22; 0.34) < 0.001 *Values are reported as median (25th percentile; 75th percentile). Table 3 – Crude and adjusted associations of sociodemographic, anthropometric and behavioral characteristics with non-neutral sagittal postural patterns. Type of sagittal postural pattern 1vs. 3 2 vs. 3 4 vs. 3 Characteristic Crude Adjusted Crude Adjusted Crude Adjusted OR (95%CI) OR (95%CI)* OR (95%CI) OR (95%CI)* OR (95%CI) OR (95%CI)* Sociodemographic Age – years (n=489) < 40 1.00 1.00 1.00 1.00 1.00 1.00 40-64 0.91 (0.18-4.48) 0.32 (0.05-1.97) 0.91 (0.45-1.83) 0.80 (0.36-1.77) 0.54 (0.26-1.12) 0.45 (0.19-1.06) ≥ 65 2.14 (0.45-10.18) 1.31 (0.19-9.25) 1.34 (0.65-2.75) 1.25 (0.51-3.05) 0.95 (0.45-1.98) 0.80 (0.30-2.12) Sex (n=489) Men 1.00 1.00 1.00 1.00 1.00 1.00 Women 3.42 (0.99-11.88) 4.09 (0.81-20.7) 0.68 (0.44-1.05) 0.83 (0.49-1.40) 0.71 (0.43-1.15) 0.83 (0.46-1.51) Education – years (n=489) ≤ 4 1.00 1.00 1.00 1.00 1.00 1.00 5-9 0.83 (0.30-2.27) 2.56 (0.76-8.68) 1.18 (0.70-1.99) 1.33 (0.72-2.46) 0.73 (0.40-1.32) 0.88 (0.44-1.75) ≥ 10 0.59 (0.21-1.68) 1.11 (0.24-5.10) 1.19 (0.72-1.96) 1.46 (0.79-2.71) 0.84 (0.49-1.46) 0.86 (0.42-1.74) Occupation (n=483)† Blue collar 1.00 1.00 1.00 1.00 1.00 1.00 Lower white collar 2.80 (0.71-11.07) 6.94 (1.20-40.07) 1.00 (0.56-1.71) 1.00 (0.51-1.95) 0.69 (0.37-1.28) 0.61 (0.28-1.31) Upper white collar 1.91 (0.46-7.96) 4.85 (0.57-41.28) 1.17 (0.69-1.99) 1.22 (0.53-2.83) 0.80 (0.45-1.44) 0.95 (0.36-2.47) Housewives 7.61 (1.79-32.36) 9.25 (1.50-57.15) 0.89 (0.39-2.02) 0.84 (0.32-2.22) 0.69 (0.27-1.72) 0.68 (0.23-1.98) Anthropometric BMI (n=489)‡ Normal weight 1.00 1.00 1.00 1.00 1.00 1.00 Overweight 1.27 (0.41-3.95) 2.22 (0.55-8.97) 1.79 (1.11-2.88) 1.92 (1.13-3.27) 1.65 (0.97-2.83) 2.13 (1.16-3.91) Obese 3.35 (1.16-9.64) 6.10 (1.52-24.57) 1.13 (0.63-2.03) 1.41 (0.72-2.75) 1.22 (0.64-2.33) 1.51 (0.70-3.25) Waist circumference (n=487)§ No central obesity 1.00 1.00 1.00 1.00 1.00 1.00 Central obesity 3.05 (1.25-7.45) 3.45 (1.09-10.92) 1.20 (0.79-1.84) 1.44 (0.86-2.40) 1.03 (0.64-1.67) 1.30 (0.72-2.35) Waist-hip ratio (n=487)|| No central obesity 1.00 1.00 1.00 1.00 1.00 1.00 Central obesity 0.95 (0.37-2.40) 1.87 (0.50-7.03) 1.40 (0.86-2.29) 1.11 (0.63-1.97) 0.80 (0.48-1.33) 0.52 (0.28-0.98) Behavioral Total PA(n=430)¶ 1st third 1.00 1.00 1.00 1.00 1.00 1.00 2nd third 0.35 (0.11-1.10) 0.35 (0.10-1.16) 0.64 (0.37-1.10) 0.61 (0.35-1.06) 0.65 (0.35-1.19) 0.64 (0.34-1.19) 3rd third 0.46 (0.15-1.39) 0.69 (0.20-2.46) 0.62 (0.35-1.09) 0.64 (0.35-1.17) 0.56 (0.30-1.07) 0.59 (0.29-1.18) Leisure-time PA (n=483)** 1st third 1.00 1.00 1.00 1.00 1.00 1.00 2nd third 2.33 (0.72-7.50) 1.94 (0.53-7.09) 0.99 (0.59-1.67) 0.88 (0.49-1.56) 0.83 (0.46-1.50) 0.71 (0.37-1.38) 3rd third 1.59 (0.44-5.70) 0.75 (0.15-3.65) 1.10 (0.64-1.89) 0.98 (0.53-1.82) 1.11 (0.61-2.01) 1.07 (0.53-2.13) Sitting position (n=487)†† Below or at the median 1.00 1.00 1.00 1.00 1.00 1.00 Above the median 1.74 (0.74-4.09) 1.52 (0.54-4.31) 1.38 (0.90-2.10) 1.43 (0.88-2.34) 1.26 (0.79-2.02) 1.17 (0.67-2.05) Smoking status (n=488)‡‡ Never 1.00 1.00 1.00 1.00 1.00 1.00 Former 0.68 (0.22-2.11) 0.68 (0.13-3.61) 1.58 (0.96-2.60) 1.17 (0.65-2.13) 1.17 (0.65-2.10) 1.06 (0.52-2.16) Current 0.51 (0.11-2.34) 1.78 (0.31-10.20) 1.25 (0.68-2.30) 1.28 (0.63-2.59) 1.69 (0.90-3.19) 2.09 (0.97-4.48) Tobacco consumption (n=481)§§ None 1.00 1.00 1.00 1.00 1.00 1.00 Lower 0.85 (0.27-2.68) 0.88 (0.17-4.68) 1.58 (0.92-2.71) 1.38 (0.73-2.59) 1.15 (0.60-2.19) 1.28 (0.61-2.69) Higher 0.21 (0.03-1.65) 0.61 (0.07-5.64) 1.40 (0.80-2.43) 1.17 (0.60-2.29) 1.60 (0.88-2.90) 1.80 (0.86-3.78) OR, odds ratio; 95%CI, 95% confidence interval; BMI, body mass index; PA, physical activity. *Adjusted for age, sex, education, BMI, total PA and smoking status, except in waist circumference and waist-hip ratio (not adjusted for BMI), and in tobacco consumption (not adjusted for smoking status). †Six students were not considered in the analysis. ‡Normal weight: BMI < 25.0 kg/m2; Overweight: 25.0 kg/m2 ≤ BMI < 30.0 kg/m2; Obese ≥ 30.0 kg/m2. §Central obesity defined as waist circumference ≥ 102.0 cm in men and ≥ 88.0 cm in women. ||Central obesity defined as waist-hip ratio ≥ 0.90 in men and ≥ 0.85 in women. ¶Thirds of total PA (METS-h/day): < 1.39, 1.39-1.57, > 1.57 for men and < 1.42, 1.42-1.50, > 1.50 for women. **Thirds of leisure-time PA (METS-h/day): < 4.50, 4.50-7.23, > 7.23 for men and < 3.00, 3.00-5.92, > 5.92 for women. ††Median equal to 2.50 h/day. ‡‡Never: participant that never smoked; Former: participant that stopped smoking for at least 6 months; Current: participant that smoke daily or occasionally. §§None: participant that never smoked; Lower: participant that smoked ≤ 7547.03 cigarette packs over life; Higher: participant that smoked > 7547.03 cigarette packs over life. Bold type indicates statistical significance. 63 Figure 1. Measurement technique of spinal (A) and pelvic (B) sagittal radiographic parameters. 64 Figure 2. Proportions of non-neutral sagittal postural patterns within categories of selected sociodemographic, anthropometric and behavioral characteristics. 3.5 3.8 6.7 28.1 30.3 33.5 28.1 18.1 23.7 0 20 40 60 80 100 < 40 40-64 ≥ 65 Age Type 1 Type 2 Type 4 1.7 6.8 36.0 28.6 24.2 19.9 0 20 40 60 80 100 Men Women Sex Type 1 Type 2 Type 4 6.0 5.1 3.6 27.7 33.6 33.3 24.5 18.2 20.8 0 20 40 60 80 100 ≤ 4 5-9 ≥ 10 Education Type 1 Type 2 Type 4 2.0 6.0 3.8 15.2 29.9 30.8 34.4 26.1 25.9 18.8 20.4 17.4 0 20 40 60 80 100 Blue collar Lower white collar Upper white collar Housewives Occupation Type 1 Type 2 Type 4 3.5 3.3 10.2 27.6 36.5 26.9 19.4 23.7 20.4 0 20 40 60 80 100 Normal weight Overweight Obese Body mass index Type 1 Type 2 Type 4 2.9 7.7 30.4 32.4 22.1 20.3 0 20 40 60 80 100 No Yes Central obesity (waist circumference) Type 1 Type 2 Type 4 6.8 3.2 4.3 35.3 30.4 30.2 24.1 20.9 18.7 0 20 40 60 80 100 1st third 2nd third 3rd third Total physical activity Type 1 Type 2 Type 4 3.0 6.9 4.4 31.1 30.9 31.9 23.0 19.1 23.8 0 20 40 60 80 100 1st third 2nd third 3rd third Leisure-time physical activity Type 1 Type 2 Type 4 65 Conclusions Sagittal standing posture was not consistently associated with pain or quality of life among adult men. However, pelvic incidence and sacral slope outside neutral ranges, increased sagittal balance, increased pelvic tilt and pelvic retroversion may be involved in causing severe back pain and consequently decreased quality of life among women, similarly to findings previously described in orthopaedic clinical samples. Higher body mass index and central obesity were important potential determinants of non-neutral posture among adults from the general population, where these anthropometric characteristics were consistently associated with non-neutral posture at isolated or overall standing alignment. Strategies focused in reducing body mass index are expected to promote the prevention of all non-neutral pathologic standing postures at a population level. Such strategies may have higher impact in women, among whom sagittal standing posture likely is an important determinant of relevant subjective outcomes, such as pain and quality of life. 66 67 References 1. Hootman JM, Helmick CG, Brady TJ. A public health approach to addressing arthritis in older adults: the most common cause of disability. Am J Public Health. 2012;102(3):426-33. 2. Woolf AD, Pfleger B. Burden of major musculoskeletal conditions. Bull World Health Organ. 2003;81(9):646-56. 3. Manchikanti L, Singh V, Datta S, Cohen SP, Hirsch JA, American Society of Interventional Pain Physicians. Comprehensive review of epidemiology, scope, and impact of spinal pain. Pain Physician. 2009;12(4):E35-70. 4. Vaz G, Roussouly P, Berthonnaud E, Dimnet J. Sagittal morphology and equilibrium of pelvis and spine. Eur Spine J. 2002;11(1):80-7. 5. Vialle R, Levassor N, Rillardon L, Templier A, Skalli W, Guigui P. Radiographic analysis of the sagittal alignment and balance of the spine in asymptomatic subjects. J Bone Joint Surg Am. 2005;87(2):260-7. 6. Berthonnaud E, Dimnet J, Roussouly P, Labelle H. Analysis of the sagittal balance of the spine and pelvis using shape and orientation parameters. J Spinal Disord Tech. 2005;18(1):40-7. 7. Boulay C, Tardieu C, Hecquet J, Benaim C, Mouilleseaux B, Marty C, et al. Sagittal alignment of spine and pelvis regulated by pelvic incidence: standard values and prediction of lordosis. Eur Spine J. 2006;15(4):415-22. 8. Roussouly P, Gollogly S, Berthonnaud E, Dimnet J. Classification of the normal variation in the sagittal alignment of the human lumbar spine and pelvis in the standing position. Spine. 2005;30(3):346-53. 9. Labelle H, Mac-Thiong JM, Roussouly P. Spino-pelvic sagittal balance of spondylolisthesis: a review and classification. Eur Spine J. 2011;20(Suppl 5):641-6. 10. Barrey C, Jund J, Noseda O, Roussouly P. Sagittal balance of the pelvis-spine complex and lumbar degenerative diseases. A comparative study about 85 cases. Eur Spine J. 2007;16(9):1459-67. 11. Rajnics P, Templier A, Skalli W, Lavaste F, Illes T. The importance of spinopelvic parameters in patients with lumbar disc lesions. Int Orthop. 2002;26(2):1048. 12. Endo K, Suzuki H, Tanaka H, Kang Y, Yamamoto K. Sagittal spinal alignment in patients with lumbar disc herniation. Eur Spine J. 2010;19(3):435-8. 13. Lafage V, Schwab F, Patel A, Hawkinson N, Farcy JP. Pelvic tilt and truncal inclination: two key radiographic parameters in the setting of adults with spinal deformity. Spine. 2009;34(17):E599-606. 74 87. Boldori L, Da Soldà M, Marelli A. [Anomalies of the trunk. An analysis of their prevalence in young athletes]. Minerva Pediatr. 1999;1999(51):7-8. 88. Endo K, Suzuki H, Nishimura H, Tanaka H, Shishido T, Yamamoto K. Sagittal lumbar and pelvic alignment in the standing and sitting positions. J Orthop Sci. 2012;[Epub ahead of print]. 89. Lord MJ, Small JM, Dinsay JM, Watkins RG. Lumbar lordosis. Effects of sitting and standing. Spine. 1997;22(21):2571-4. 90. De Carvalho DE, Soave D, Ross K, Callaghan JP. Lumbar spine and pelvic posture between standing and sitting: a radiologic investigation including reliability and repeatability of the lumbar lordosis measure. J Manipulative Physiol Ther. 2010;33(1):48-55. 91. Callaghan JP, McGill SM. Low back joint loading and kinematics during standing and unsupported sitting. Ergonomics. 2001;44(3):280-94. 92. O'Sullivan PB, Smith AJ, Beales DJ, Straker LM. Association of biopsychosocial factors with degree of slump in sitting posture and self-report of back pain in adolescents: a cross-sectional study. Phys Ther. 2011;91(4):470-83. 93. Rapuri PB, Gallagher JC, Smith LM. Smoking is a risk factor for decreased physical performance in elderly women. J Gerontol A Biol Sci Med Sci. 2007;62(1):93100. 94. Strand BH, Mishra G, Kuh D, Guralnik JM, Patel KV. Smoking history and physical performance in midlife: results from the British 1946 birth cohort. J Gerontol A Biol Sci Med Sci. 2011;66(1):142-9. 95. Wong PK, Christie JJ, Wark JD. The effects of smoking on bone health. Clin Sci. 2007;113(5):233-41. 96. Morse CI, Wüst RC, Jones DA, de Haan A, Degens H. Muscle fatigue resistance during stimulated contractions is reduced in young male smokers. Acta Physiol (Oxf). 2007;191(2):123-9. 97. Wüst RC, Morse CI, de Haan A, Rittweger J, Jones DA, Degens H. Skeletal muscle properties and fatigue resistance in relation to smoking history. Eur J Appl Physiol. 2008;104(1):103-10. 98. Battié MC, Videman T, Gill K, Moneta GB, Nyman R, Kaprio J, et al. 1991 Volvo Award in clinical sciences. Smoking and lumbar intervertebral disc degeneration: an MRI study of identical twins. Spine. 1991;16(9):1015-21. 99. Videman T, Battié MC, Parent E, Gibbons LE, Vainio P, Kaprio J. Progression and determinants of quantitative magnetic resonance imaging measures of lumbar disc degeneration: a five-year follow-up of adult male monozygotic twins. Spine. 2008;33(13):1484-90. 75 100. Uematsu Y, Matuzaki H, Iwahashi M. Effects of nicotine on the intervertebral disc: an experimental study in rabbits. J Orthop Sci. 2001;6(2):177-82. 101. Akmal M, Kesani A, Anand B, Singh A, Wiseman M, Goodship A. Effect of nicotine on spinal disc cells: a cellular mechanism for disc degeneration. Spine. 2004;29(5):568-75. 102. Nemoto Y, Matsuzaki H, Tokuhasi Y, Okawa A, Uematu Y, Nishimura T, et al. Histological changes in intervertebral discs after smoking and cessation: experimental study using a rat passive smoking model. J Orthop Sci. 2006;11(2):191-7. 103. Vo N, Wang D, Sowa G, Witt W, Ngo K, Coelho P, et al. Differential effects of nicotine and tobacco smoke condensate on human annulus fibrosus cell metabolism. J Orthop Res. 2011;29(10):1585-91. 104. Uei H, Matsuzaki H, Oda H, Nakajima S, Tokuhashi Y, Esumi M. Gene expression changes in an early stage of intervertebral disc degeneration induced by passive cigarette smoking. Spine. 2006;31(5):510-4.