Summary of studies on education and the frequency of back pain
| First author, year, country and reference | Type of study | Subjects (n)/(response rate) | Definition of education3-150 | Measure of outcome | Association3-151 | Point estimates | Major weaknesses3-152 | |||||||||
| Major studies | ||||||||||||||||
| Latza (2000) Germany60 | Cross sectional | Population based sample of adults aged 25–74 years in Lübeck, Germany, in 1991–1992 (Lübeck Survey on Back Pain) - (n=2731) / (80.9%) | (1) (2) (3) | Low (8–9 years)§ Intermediate (10–12 years) High (⩾13 years) | →Point prevalence of severe back pain (self report) |
OR (95% CI)
(2) 0.24 (0.09, 0.67) (3) 0.35 (0.21, 0.58) | ||||||||||
| Inverse | L | |||||||||||||||
| Barnekow-Bergkvist (1998) Sweden63 | Cross sectional | Sample representative of 16 year old boys and girls in the first year of upper secondary school in Sweden, assessed 18 years later (n=238) / (65%) | (1) (2) | <14 years§
⩾14 years | →Period prevalence of disabling low back pain (self report—at least once per month— preceding 12 months) | OR (95% CI) | ||||||||||
| No | Men | (2) 1.53 (0.28, 8.48) | S Z L | |||||||||||||
| Women | (2) — — — | |||||||||||||||
| Heistaro (1998) Finland36 | Cross sectional | Independent random samples of 30–59 year old, drawn from the population register of the eastern provinces of North Karelia and Kuopio, Finland, every five years from 1972 to 1992 (North Karelia Project) - (n=29 043) / (69–96%) | Years of schooling tertiles: | →Period prevalence of back pain (self report - preceding month) | OR (95% CI) | |||||||||||
| (1) (2) | Low and middle§
High | Inverse | (2) 0.77 (0.73, 0.81) | S L | ||||||||||||
| Leino-Arjas (1998) Finland29 | Cross sectional | Random sample of the occupationnally active Finnish population aged 20–64 years, in 1988–1990 (n=7544) / (73–82%) | (3) (2) (1) | Basic (⩽9 years) Secondary (10–12 years) Higher (>12 years)§ | →Period prevalence of back pain (self report - preceding month) | OR (95% CI) | ||||||||||
| Inverse (only among men) | Men | (2) 1.38 (1.14, 1.68) (3) 1.80 (1.47, 2.19) | L | |||||||||||||
| Women | (2) 1.12 (0.94, 1.34) | |||||||||||||||
| (3) 1.05 (0.87, 1.27) | ||||||||||||||||
| →Period prevalence of back pain leading to medical consultation (self report - preceding 12 months) | Inverse | Men | (2) 1.47 (1.11, 1.94) | L | ||||||||||||
| (3) 1.91 (1.46, 2.51) | ||||||||||||||||
| Women | (2) 1.36 (1.03, 1.80) | |||||||||||||||
| (3) 1.60 (1.21, 2.11) | ||||||||||||||||
| Hurwitz (1997) United States58 | Cross sectional | Multiple random sample of the civilian non-institutionalised population of the US aged 18 years and over (1989 National Health Interview Survey) - (n=84 572) / (94.9%) | (1) (2) (3) | < High school graduate§ High school graduate > High school graduate | →Period prevalence of disabling back pain (self report - preceding 2 weeks) | OR (95% CI) | ||||||||||
| Inverse | (2) 0.92 (0.87, 0.98) (3) 0.84 (0.78, 0.90) | L G | ||||||||||||||
| →Period prevalence of non-disabling back pain (self report - preceding 2 weeks) | No | (2) 0.98 (0.91, 1.06) (3) 1.02 (0.94, 1.11) | L G | |||||||||||||
| Stronks (1997) The Netherlands59 | Cross sectional | Cohort of 25–64 year old non-institutionalised people with Dutch nationality in a South East region of the Netherlands (LS-SEHD) - (n=13 391) / (70.1%) | (4) | Primary school | →Point prevalence of 23 chronic conditions, including low back pain (self report) | OR (95% CI) | ||||||||||
| (3) | Lower general and vocational education | Inverse (only among men) | Men | (2) 1.34 (1.15, 1.56) | L O | |||||||||||
| (3) 1.28 (1.12, 1.47) | ||||||||||||||||
| (2) | Intermediate vocational and intermediate/higher general education | (4) 1.49 (1.26, 1.77) | ||||||||||||||
| Women | (2) 1.00 (0.83, 1.21) | |||||||||||||||
| (1) | Higher vocational college and university§ | (3) 0.97 (0.82, 1.14) | ||||||||||||||
| (4) 1.12 (0.92, 1.36) | ||||||||||||||||
| Harreby (1996) Denmark62 | Longitudinal (25 years) | All 14 year old pupils in Helsingør, Denmark, in 1965 (n=481) / (83%) | Not stated | →Period prevalence of severe low back pain (pain lasting for more than 30 days - self report - preceding 12 months) | No | E P | ||||||||||
| Liira (1996) Canada57 | Cross sectional | Household-based population survey of all residents aged 16 to 64 years living outside institutions and Indian reserves in Ontario, Canada (Ontario Health Survey) - (n=18 920) / (77.5–87.5%) | (1) (2) (3) | Primary or some secondary Completed secondary or some post-secondary Completed post-secondary | →Point prevalence of long term disabling back problems (self report) | Prevalence of outcome (95% CI) | ||||||||||
| Inverse | (1) 10.5% (9.3, 11.7) (2) 7.3% (6.3, 8.2) (3) 5.7% (4.7, 6.7) | L M G | ||||||||||||||
| Croft (1994) United Kingdom56 | Cross sectional | Large sample of the British adult population aged 18 and more, living in private households (Health and Lifestyle Survey) - (n=9003) / (73.5%) | (5) | None | →Period prevalence of back pain (self report - preceding month) | OR (95% CI) | ||||||||||
| (4) | General certificate of education-Ordinary (O level) | Inverse (only among women) | Men | (2) 0.77 (0.31, 1.87) | L | |||||||||||
| (3) 1.18 (0.84, 1.65) | ||||||||||||||||
| (3) | Certificate of education-Secondary (A level) | (4) 0.95 (0.68, 1.31) | ||||||||||||||
| (5) 1.19 (0.89, 1.59) | ||||||||||||||||
| (2) | Professional qualification | Women (2) 0.83 (0.52, 1.30) | ||||||||||||||
| (1) | University degree§ | (3) 0.97 (0.62, 1.53) | ||||||||||||||
| (4) 1.26 (0.89, 1.81) | ||||||||||||||||
| (5) 1.45 (1.02, 2.05) | ||||||||||||||||
| Park (1993) United States55 | Cross sectional | Household survey designed to provide estimates representative of the US civilian non-institutionalised employed population aged 18 and over (1988 National Health Interview Survey) - (n=27 408) / (87%) | (1) | <12 years | →Period prevalence of back pain every day for a week or more (self report - preceding 12 months) | Prevalence of outcome | ||||||||||
| (2) | 12 years | Inverse | (1) 21.3% | L M G | ||||||||||||
| (3) | 13–15 years | (2) 20.1% | ||||||||||||||
| (4) | >15 years | (3) 19.1% | ||||||||||||||
| (4) 15.8% | ||||||||||||||||
| Von Korff (1993) United States61 | Longitudinal (3 years) | Persons with no prior history of back pain at baseline among a probability sample of adult enrollees of a large HMO (n=271) / (85%) | (1) | High school or less§ | →Incidence of back pain (self report) | OR (statistical significance) | ||||||||||
| (2) | Some college | No | (2) 0.59 (NS) | Z G | ||||||||||||
| (3) | College graduate | (3) 1.01 (NS) | ||||||||||||||
| Viikari-Juntura (1991) Finland18 | Longitudinal (22 years) | Selected respondents of the 1985 follow up of children who lived in the Helsinki Metropolitan region (Healthy Child Study) - (n=154) / (90%) | (3) | Elementary or less | →Period prevalence of severe low back symptoms (self report - preceding 12 months) | OR (95% CI) | ||||||||||
| (2) | Intermediate secondary | Inverse (only among women) | Men | (2) — — — | S Z | |||||||||||
| (1) | High school§ | (3) — — — | ||||||||||||||
| Women | (2) 5.86 (0.28, 121.19) | |||||||||||||||
| (3) 35.45 (1.56, 804.55) | ||||||||||||||||
| Leigh (1989) United States54 | Cross sectional | Probability sample of US workers at least 16 years old working more than 20 h/week (Quality of Employment Survey) - (n=1414) / (weighted) | (4) | ⩽8 years | →Period prevalence of spinal pain (self report - preceding 12 months) | OR (95% CI) | ||||||||||
| (3) | 9–11 years | Inverse | (2) 1.05 (0.65, 1.72) | L G | ||||||||||||
| (2) | 12 years | (3) 1.45 (0.75, 2.81) | ||||||||||||||
| (1) | ⩾13 years§ | (4) 2.18 (1.04, 4.63) | ||||||||||||||
| Deyo (1987) United States52 | Cross sectional | Probability sample representing the civilian, non-institutionalised US population aged 25 and over (NHANES II) - (n=10 404) / (weighted) | (1) | Elementary or none | →Lifetime prevalence of low back pain for ⩾2 weeks (self report) | Prevalence of outcome | ||||||||||
| (2) | High school | Inverse | (1) 17.3% | L M G | ||||||||||||
| (3) | College | (2) 14.4% | ||||||||||||||
| (3) 11.2% | p<0.0001 | |||||||||||||||
| Pincus (1987) United States11 | Cross sectional | Individuals selected from the 1976 cohort of the Health Interview Survey to be representative of the US working age population (18–64 years) - (n=5652) / (weighted) | (4) | ⩽8 years | →Point prevalence of back stiffness or deformity (self report - “as told by a physician”) | OR (95% CI) | ||||||||||
| (3) | 9–11 years | Inverse | (2) 1.0 (1.0, 1.1) | L | ||||||||||||
| (2) | 12 years | (3) 1.5 (1.3, 1.6) | ||||||||||||||
| (1) | ⩾13 years§ | →Point prevalence of other back problems (self report - “as told by a physician”) | (4) 2.4 (2.1, 2.6) | |||||||||||||
| Inverse | (2) 1.3 (1.2, 1.3) | L | ||||||||||||||
| (3) 1.8 (1.7, 1.9) | ||||||||||||||||
| (4) 2.0 (1.9, 2.2) | ||||||||||||||||
| Saraste (1987) Sweden53 | Cross sectional | Random, geographically stratified 1:1000 sample of the Swedish working population aged 30–59 years (n=2872) / (?) | Pre-high school | →Lifetime prevalence of low back pain with disability (self report) | Per cent with post-high school (low back pain subjects/controls) | |||||||||||
| High school | ||||||||||||||||
| Post-high school | Inverse (only among men aged 30–39) | Age | 30–39 | 40–49 | 50–59 | L R M | ||||||||||
| Men | 19/30 | 13/24 | 11/17 | |||||||||||||
| Women | 22/27 | 15/19 | 10/12 | |||||||||||||
| Reisbord (1985) United States51 | Cross sectional | Adult enrollees in a health insurance plan (n=2792) / (weighted) | (1) | Less than high school§ | →Period prevalence of frequent back pain (self report - preceding 12 months) | OR (95% CI) | ||||||||||
| (2) | High school | Inverse | (2),(3) 0.66 (0.57, 0.78) | L | ||||||||||||
| (3) | Over high school | |||||||||||||||
| Cunningham (1984) United States4 | Cross sectional | Multistage, stratified probability sample of persons aged 25–74 in the US (NHANES I Survey) - (n=6913) / (weighted) | (2) | <12 years | →Point prevalence of back abnormalities (physician observed) | OR (95% CI) | ||||||||||
| (1) | ⩾12 years§ | Inverse | (2) 1.16 (1.10, 1.23) | L G | ||||||||||||
| →Period prevalence of back symptoms (self report - preceding 12 months) | Inverse | (2) 1.85 (1.75, 2.00) | L G | |||||||||||||
| Nagi (1973) United States50 | Cross sectional | Probability sample of all persons 18–64 year old residing within a Standard Metropolitan Statistical Area in Columbus, Ohio (n=1135) / (94%) | (1) | ⩽8 years | →Point prevalence of persistent back pain (self report) | Prevalence of outcome | ||||||||||
| (2) | 9–11 years | Inverse | (1) 29.5% | L M G | ||||||||||||
| (3) | 12 years | (2) 24.2% | ||||||||||||||
| (4) | ⩾13 years | (3) 16.2% | ||||||||||||||
| (4) 11.8% | p<0.01 | |||||||||||||||
| Other studies | ||||||||||||||||
| Bergenudd (1988, 1994) Sweden70 71 | Longitudinal (45 years) | Residents of Malmö (Sweden) who participated to the Malmö Longitudinal Study in 1938 and volunteered for a health examination in 1983 (n=575) / (69%) | Not stated | →Point prevalence of back pain (examination) | Inverse | S E M P G | ||||||||||
| Jacobsson (1992) Sweden19 | Cross sectional | Persons living in the Malmö area, aged 50–70 years, randomly selected from population records (Malmö Food Health Survey) - (n=502) / (56%) | ⩽8 years | →Point prevalence of low back pain (examination or self report) | Prevalence of low education (low back pain subjects v controls) | |||||||||||
| >8 years | ||||||||||||||||
| Inverse | 81% v56% | L R M G | ||||||||||||||
| Armenian (1989) Lebanon69 | Cross sectional | Stratified random sample of 2752 households in Beirut (Population Laboratory Survey) - (n=13 735) / (?) | (2) | Low | →Lifetime prevalence of “chronic” back pain (self report) | OR (95% CI) | ||||||||||
| (1) | Middle and high§ | No | (2) 0.97 (0.74, 1.29) | L R O G | ||||||||||||
| Roncarati (1988) United States68 | Cross sectional | Subjects in sport medicine, physical therapy clinics, high schools and universities (n=674) / (?) | Unclearly stated | →Point prevalence of traumatic or intrinsic low back pain (differential diagnosis by an orthopaedist) | Direct | S E L R M P G | ||||||||||
| Biering-Sørensen (1986) Denmark67 | Longitudinal (12 months) | Inhabitants of Glostrup (Denmark) 30–60 year old invited to take part in a free general health survey (n=928) / (82%) | Not stated | →One year incidence of low back pain (self report) | No | S E P G | ||||||||||
| →One year period prevalence of low back pain (self report) | No | S E P G | ||||||||||||||
| →Lifetime prevalence of low back pain (self report) | No | S E P G | ||||||||||||||
| →Recurrence/persistence rate of low back pain (self report) | No | S E P G | ||||||||||||||
| Svensson (1982) Sweden66 | Cross sectional | Random sample of 40–47 year old men selected from the census register of the city of Göteborg, Sweden (n=716) / (76%) | Primary school | →Lifetime prevalence of low back pain (insurance data) | No | S L P G | ||||||||||
| Middle school | ||||||||||||||||
| Upper school | ||||||||||||||||
| Adult continuation school | ||||||||||||||||
| University or others | ||||||||||||||||
| Magora (1970) Israel65 | Cross sectional | Subjects chosen randomly from eight selected occupations (n=3316) / (?) | Unclearly stated | →Lifetime prevalence of low back pain for more than 3 days (self report) | Direct | S E L R M P G | ||||||||||
↵3-150 Reference category indicated by §.
↵3-151 No association: there was no statistically significant association. Direct association: statistically significant association where back pain frequency increased with increasing educational status. Inverse association: statistically significant association where back pain frequency decreased with increasing educational status.
↵3-152 Major weaknesses pertaining to: S: selection of subjects, Z: sample size, E: measure of education, L: length of follow up, R: drop out rate, O: measure of outcome, M: multivariate analysis, P: point estimates, G: consideration of effect modification by gender.