Risk Factors and Risk Profiles for Persistent Neck Pain in Young Adults: Results From the HUNT Study

Little research exists about risk factors for neck pain in adolescents and young adults. Knowledge of risk factors associated with development of neck pain can inform practice and can further enable preventive initiatives with the aim of reducing this prevalent and costly disorder. By investigating risk profiles as well as single risk factors, it might be possible to identify behavioural patterns and profiles of adolescent most at risk of developing neck pain in young adulthood.

The objective of this study is to investigate potential risk factors and risk profiles for persistent neck pain in adolescence.

This is a prospective cohort study using data from The North Trøndelag Health study (HUNT). The HUNT study is a large population survey conducted in the North Trøndelag County in Norway, consisting of four surveys conducted with 11 years intervals. All inhabitants above 13 years old were invited to participate in either Young HUNT (13-19 years of age) or HUNT (20 years and above).

In this cohort study, data from Young HUNT 3 (2006-2008) and HUNT 4 (2017-2019) will be used to investigate participants from adolescence through young adulthood. The HUNT3 assessment was conducted during school hours, and the students had to answer a comprehensive questionnaire regarding biological, psychological and social factors. In addition, height and weight were measured. For participants not attending school, the questionnaire was sent to their home address, and participants were invited to do the physical measurements in a local field station.The same participants were followed up 11 years later (HUNT4). The participants were sent an invitation letter home and asked to fulfill a questionnaire.

Potential risk factors (from HUNT3):

Gender, body mass index, pain in other body regions, number of pain sites, sleeping problems, physical activity level, depression and anxiety, self esteem, resilience, loneliness and family economy.

Statistical analysis:

All statistical analyses will be conducted using IBM SPSS statistics for Windows version 25 and STATA statistical software system version 15. All statistical test will be two sided and the nominal p-value will be reported. 95% confidence intervals will be reported. Preliminary analyses of frequencies, missing data and normality will be conducted. The assumption of normal distribution will be investigated using histograms and QQ-plots.

Missing data:

Participants with missing data on the main outcome will be excluded from the analyses. Missing data on exposure variables will be handled by a model-based imputation (multiple imputation), unless the missing exceeds 20% and missing at random can be assumed. Exposure variables with more than 20% missing data will be excluded.

Multicollinearity:

Before conducting regression analyses, multicollinearity between the independent variables will be assessed. If independent variables correlates with a coefficient of stronger than 0.7, the risk factor that have the highest correlation with the outcome will be used in the multivariate analysis.

Logistic regression:

The potential risk factors listed above will individually be analysed in univariate models with persistent neck pain as the dependent variables. Variables with a p-value of ≤ 0.10 in univariate analysis will be included in multivariate analyses. Variables with a p-value of ≤ 0.05 will remain in the multivariate model, and the remaining variables no longer be statistically relevant will be removed from the final model. The results will be reported as odds ratios (OR) with corresponding 95% confidence intervals (95% CI).

Risk Matrices:

To identify risk profiles for persistent neck pain, risk matrices will be developed. The odds computed in the final regression model developed will be transformed into probabilities, and the results will be arranged in a risk matrix.