Background/aims There is a paucity of prospective cohort studies investigating the incidence of low back pain (LBP) in rowing. We investigated (1) the prevalence and incidence of LBP among international-level rowers in New Zealand, (2) the relationship between training volume and LBP and (3) the effect of LBP on rowers’ ability to train and compete.
Methods This was a prospective cohort study of 76 New Zealand representative rowers, including 46 men (mean age 22, SD=4) and 30 women (mean age 21, SD=4). Data were collected using an online questionnaire repeated monthly for 12 months.
Results The prevalence of LBP ranged from 6% to 25% throughout the year. The incidence of episodes of LBP was 1.67 per 1000 exposure-hours. A total of 72 episodes of LBP were reported by 40 rowers (53%) during 12 months. Of these, 45% had an incidental effect on training. 29% minor, 18% moderate and 9% had a major effect as determined by the length of time the training was interrupted. There was a high correlation between new LBP and total training hours per month (r=0.83, p<0.01). A previous history of LBP was a risk factor in developing new LBP (OR 2.06, 95% CI 1.22 to 3.48, p=0.01). Age was also a risk factor, with the likelihood of developing LBP increasing for every year (OR 1.08, 95% CI 1.01 to 1.15, p=0.02).
Conclusions LBP is common among New Zealand representative rowers. There is a high correlation between training load and the development of LBP.
- Back injuries
Statistics from Altmetric.com
Rowing is a popular sport that requires outstanding physiological attributes and places high demands on the musculoskeletal system, particularly the lumbar spine.1 ,2 Low back pain (LBP) is reported to be the most common injury in rowing, affecting between 25% and 81% of athletes per year3 ,4 and accounting for 15–25% of all injuries in the sport.5 ,6 However, there are no prospective cohort studies specifically investigating the prevalence of LBP among rowers, with previous epidemiological studies having either retrospective, cross-sectional or case-series design.3 ,7–15 A number of sports-specific risk factors for LBP in rowing have been proposed.16 High volumes of training, load levels and ergometer training may also represent potential mechanisms for strain to the lumbar spine structures.17–19 Prospective studies are needed to establish a link between LBP and possible risk factors.
We aimed to investigate (1) the prevalence and incidence of LBP among international-level rowers in New Zealand, (2) the association between training volumes and LBP and (3) the extent to which LBP affects the rowers’ ability to train and compete.
This was a prospective cohort study conducted over 12 months. Each month, an online survey tool (http://www.surveymonkey.com) distributed an electronic questionnaire (see online supplement) to 76 national representative rowers. The study was approved by the Auckland University of Technology (AUT) Ethics Committee, and all participants received verbal and written information about the study and gave written informed consent. New Zealand Rowing was supportive of the study and helped facilitate recruitment.
All rowers selected to represent New Zealand at a World Rowing Championship event in 2011 were invited to participate in the study, including members of the junior (under 18 years) (n=13), under 23 (n=30) and senior teams (n=33).
The questionnaire was developed following a review of the relevant literature. The initial questionnaire administered collected information on the following potential risk factors: age, gender, height, weight, age at commencement of competitive rowing, current level of competition, rowing discipline (sweep oar rowing or sculling) and previous history of LBP. The initial and follow-up questionnaires collected information on the presence and severity of LBP in the previous 4 weeks, as well as the volume and type of training completed.
To determine the relationship between possible individual risk factors and experiencing a new episode of back pain during the 12-month period, a set of physical screening tests relevant to rowing were performed at the start of the data collection period. These tests comprised of Movement Competency Screen (MCS) score,20 Active Straight Leg Raise21 and trunk strength-endurance ratio.22 Body mass index (BMI) was calculated from the questionnaire data above.
LBP was defined as pain, ache or discomfort in the low back with or without referral to the buttocks or legs8 ,19 that has been present for more than 1 week10 ,13 ,14 and/or interrupted at least one training session. The period of 1 week was selected to differentiate a low back injury from that of short-term postexercise soreness that is common among elite athletes. We also determined whether participants with LBP were experiencing a new or an on-going episode. An episode was considered new if symptoms were not present in the previous month's questionnaire. New injuries were taken to include reinjury and exacerbations. Injury severity measures were based on previous criteria used in studies of rowers,11 ,12 which were altered to include training modification through either type, intensity or duration from the planned training session. Injuries resulting in no time loss or modification of rowing participation were classified as incidental; injuries causing modified participation in any training sessions for a period of less than 1 week were classified as minor; injury causing absence from or modification of rowing training for more than 1 week but less than 1 month were classified as moderate; and those causing absences or modification of participation for more than 1 month were classified as major injuries.
A training diary was developed and participants were asked to complete this for the duration of their involvement in the study to aid accurate recall of their training programme.
Each time the questionnaire was administered, the prevalence of LBP was determined by dividing the number of rowers with LBP by the number of questionnaire respondents. The incidence of LBP was expressed as the number of new episodes of LBP per 1000 h of rowing exposure.23 ,24 Pearson correlations were used to determine the relationship between training volume and the incidence of LBP. A multivariate logistic regression model was used to determine the relationship between potential risk factors and experiencing a new episode of back pain. Variables (listed in tables 1–3 and 5) were retained in the final logistic regression model if they reached a statistical threshold of 0.10 or were deemed to be of clinical significance. All analyses were performed using Statistical Programme for Social Science (SPSS) software (SPSS V.19, IBM Corporation, New York, USA).
Rowing New Zealand contacted the 102 athletes who met the inclusion criteria, having represented New Zealand at rowing during the 2011 season. Of these, 76 consented to take part in the study where 42 were currently senior elite athletes and the remainder were age group representatives. With respect to the compliance rate, 59 participants completed all 12 questionnaires (78%) and of the total 912 questionnaires sent out over the year, 817 (90%) were completed. Of those participants who did not complete all the questionnaires, 6 stopped rowing during the season due to non-selection in teams (1), injury (3) and personal reasons (2). The remaining 11 participants who did not complete all 12 questionnaires continued rowing during the season and either withdrew from the study (4) or missed one or more questionnaires (7). The demographic characteristics of the participants are presented in table 1.
Injury prevalence, incidence and severity
Injury prevalence and severity
The prevalence of LBP ranged from 6% to 25% over the 12-month period of the study (figure 1).
Forty participants (52.6%) experienced at least one episode of LBP during the 12-month period. No statistically significant differences were observed between rowers with and without LBP in age, height, weight, BMI, the age they began rowing or the number of years they had been rowing (table 2).
χ2 Tests also revealed no statistically significant difference between these groups for either gender (χ2=0.10, p=0.92), rowing discipline (χ2=0.46, p=0.50) or competition level (χ2=0.43, p=0.51; table 3).
There was no difference in the number of low back injuries reported over the 12-month period between the elite and age group participants (χ2=1.59, p=0.21).
Of the 40 rowers who experienced LBP during the study, 21 (52.5%) reported multiple episodes. Thirty rowers (75%) reported having LBP in two or more consecutive questionnaires, indicating their LBP lasted for more than 1 month. Forty rowers (52.6%) reported experience of at least one episode of LBP prior to the start of this study. Injury severity is presented in figure 2.
A total of 72 episodes of LBP were reported over the 12-month period (table 4). The incidence of LBP was 1.67 episodes per 1000 h of rowing exposure (water and ergometer combined).
Relationship between training load and LBP
There were high positive correlations between the number of new episodes of LBP and total training hours per month (r=0.83, p<0.01; figure 3), the number of ergometer training hours per month (r=0.80, p<0.01), the average training hours per participant per month (r=0.73, p<0.01) and the average number of kilometres rowed per participant per month (r=0.71, p=0.01).
Relationship between potential risk factors and LBP
The relationships between potential risk factors and the odds of developing an episode of new LBP are reported in table 5. A linear mixed effects model was fit to identify significant risk factors for repeated episodes of LBP within the 12-month study (see online supplementary table 1). Active straight leg raising and trunk endurance ratios did not have any association with injury (p>0.1) and were not included in the multivariate model. Results observed were similar to those reported in table 5; with age of the rower and history of LBP the only significant predictors of new episodes of LBP.
Key strengths of this study were (1) the population of elite rowers who agreed to participate and (2) the high response rate to the questionnaires. Seventy-five per cent of the rowers who were eligible for the study agreed to participate. In addition to this, 90% of the possible 912 questionnaires were completed and 78% of the participants completed all 12 questionnaires over the year. These are substantial new data to assist rowing sports medicine and physiotherapy.
Prevalence, incidence and severity
The prevalence of participants experiencing LBP ranged from a maximum of 25% to a low of 6% throughout the year. This finding is similar to a Norwegian cross-sectional study which reported an LBP prevalence of 25.3% among club-level rowers during their peak season.8 However, it is lower than the prevalence reported by Stutchfield and Coleman,13 who found that 42% of Scottish University-level rowers had LBP and by Perich et al25 who reported a prevalence of 47.5% among Australian adolescent female rowers. As the prevalence of LBP varies throughout the season, comparison between studies is difficult.
Forty participants (52.6%) experienced at least one new episode of LBP over the 12-month period. This is very similar to the prevalence of LBP of 55.3% over a 12-month period among Norwegian national elite level rowers.8 It is somewhat higher than the 32% prevalence of LBP among American rowers reported over their intercollegiate rowing careers11 and that of 21.3% among elite junior rowers during a season.12 This is also high in comparison to the 12-month prevalence of LBP in the general population.26 In a systematic review of LBP epidemiological literature, Hoy et al26 found a mean 12-month prevalence of 38.1% among the studies they reviewed and Bahr et al8 reported a 12-month prevalence of 47.5% among the non-athletic controls in their study.
Forty (52.6%) of the participants reported having had experienced at least one episode of LBP previously. When making comparisons to previous studies, there is a large variation between 32% and 81% in the reported lifetime prevalence of LBP among rowers.8 ,13 ,14 ,8 The only study that reports lifetime prevalence among elite level rowers found this to be 63.3%.8 It is well known among the rowing population that LBP is a common problem and the injury that causes the most interruption to training. Therefore, it is possible that these athletes will have a lower threshold for reporting and remembering episodes of LBP than the general population. A strength of the current study is that participants were followed up monthly over a 12-month period and therefore, recall bias will be less than that of some other studies, further increasing the reported prevalence of LBP compared to other studies.
A secondary objective of this study was to determine the incidence of LBP among New Zealand rowers. The incidence rate was 1.67 injuries per 1000 rowing-specific training hours. One previous study of injury incidence among rowers9 reported 1.31 lumbar spine and sacroiliac joint injuries per 1000 h of combined training and competing.
As the majority of low back injuries are overuse injuries, the onset of symptoms is often gradual and the point at which they become an injury is blurred.27 In addition to this, it is difficult to determine whether an episode of LBP is a new injury, a reinjury or an exacerbation of a previous-related injury due to the inability to determine the state of full recovery.28 ,29 The traditional method of classifying recurrent events as either reinjuries or exacerbations30 ,31 is difficult to apply to LBP and therefore, the incidence of LBP episodes needs to be interpreted with caution.
We determined the severity of LBP among New Zealand rowers as it is related to the effect on training and competition. Approximately three of four reported episodes of pain were either incidental (45%) or minor (29%), meaning that participants did not miss or modify any of their training sessions and/or competitions, or missed or modified training or competition for under a week. Of the remaining quarter of the injuries, 18% were moderate and caused the participants to miss or modify training and/or competition between 1 week and 1 month; and 9% were major, resulting in more than 1 month of training modification. Of the major injuries, 3 of the participants retired from rowing because of on-going LBP. This equated to 4% of the cohort, which is lower than that reported in a college rowing population.11 ,15
This distribution of injury severity is very similar to that reported among elite age group rowers12 and that of intercollegiate rowers who had experienced LBP prior to starting rowing.11 Wilson et al7 established the severity of injury by the number of training and racing hours lost due to injury. They found a number of the injured participants who completed more training hours than their non-injured counterparts.7 Therefore, time lost is a very poor indicator of injury severity.7 When an athlete suffers an injury, it is likely that he or she will substitute a planned training session for a cross-training session that can be completed without aggravating their symptoms. By including sessions modified from that planned in terms of type, time and intensity, the current study's findings reflect a more accurate picture of injury severity.28
Relationship between training load and LBP
An important finding in this study was the high correlation between new LBP and current LBP, and the total number of training hours (on and off water) and kilometres rowed per month among the participants. This supports the findings of Smoljanovic et al12 who found that junior rowers, who averaged more than seven training sessions per week, had more overuse LBP than those with fewer training sessions per week.11 Bahr et al8 and Wilson et al7 found more LBP during periods of higher competition and training load. It also supports the findings from the study by Wilson et al7 who found an association between high on-water training volume and injury rate; however, in their study this association was not statistically significant. Kinematic studies indicate that spinal motion markedly increases with prolonged rowing32 and there are important changes to lumbopelvic and spinal kinematics with increasing work intensity levels.18 ,33–35 Although studies have investigated these changes during ergometer rowing, these changes may be important in explaining the relationship between high training load and the new LBP found in the current study.
Another relevant observation is the sudden increase in new episodes of LBP in the month of October. This was related to the rapid increase in training load as the athletes come out of their off-season. Previous research has noted that sudden changes in training load, in terms of intensity, duration and frequency, in adolescent athletes increases the risk of LBP.36 Educating coaches regarding this risk factor and on increasing training load gradually at this particular time of the year might reduce this increase in LBP.
A previous history of LBP was the most relevant risk factor in experiencing an episode of LBP during this study. This extends to other research on previous LBP as a predictor of future back pain in adolescent populations.37 This is of concern in the sport of rowing as many athletes start their careers as school age rowers when the spine is still developing. Perich et al25 have demonstrated that this school age group already have a higher prevalence of LBP than children of a similar age who do not row. Preventing the first episode of LBP among rowers is, therefore, the most significant factor in reducing the prevalence of these injuries in this population. As a large number of athletes entering a New Zealand rowing programme having already experienced their first episode of LBP, it could be argued that injury prevention strategies are best targeted at the younger, school aged, rowing population. A multidimensional approach consisting of physiotherapy screening, prescription of individualised ‘specific exercise’, education, and strength and conditioning sessions can be successful in reducing the incidence of LBP in schoolgirl rowers.38 ,39 However, it is interesting to note that in this study, the ability to predict injury incidence from screening using tools such as the MCS was not significant. Further work on risk factors that predict injury will be required in future.
The age of the participants was also a risk factor in developing LBP with the older participants more at risk than those younger participants. This supports the findings of previous epidemiology studies40–42 that have found LBP gradually increases with age in the general population.
Limitations of the study
This study used a non-validated questionnaire. Although it was pilot tested, this process was not vigorous and as such when gathering the exposure data the amount of cross training hours that participants completed was not included. Although data was collected on the type and number of cross training sessions completed, the time exposure was not and therefore, it could not be included in the calculation of injury incidence. This would cause the incidence rate to be overestimated in comparison to other studies that have collected this information.
Severity measures included training modification described as a change in training from that planned, in terms of type, intensity or duration. However, to what extent training was modified was not further investigated. A recent study has validated a questionnaire that assesses to what degree training is modified28 and this could be used to assess injury severity in rowers in future research.
There were 72 cases of LBP in this study. Sample size mainly depends on the expected effect of the risk factor on injury risk and to detect moderate to strong associations 20–50 injury cases are needed, whereas small to moderate associations would need about 200 injured participants.43
The prevalence of LBP among international-level rowers in New Zealand is high, varies throughout the rowing season as it is strongly correlated to the volume of rowing training performed. Episodes of LBP were highest in the month of October, the time of peak training volumes. Previous LBP and age are risk factors for developing an episode of LBP in rowers in the future.
What are the new findings?
There is a strong relationship between high training volumes and the prevalence of low back pain in elite rowers.
Previous episodes of low back pain predict future episodes of low back pain in rowers.
The age of rowers (over 22 years of age) is also a risk factor for developing low back pain.
How might it impact on clinical practice in the near future?
Sports medicine providers should monitor the volume of training and keep in discussion with coaches to reduce the prevalence of low back pain.
Because of the prevalence of overuse injuries, the Oslo Sport Trauma Research Centre Overuse Injury Questionnaire may be a useful instrument to achieve this.
Preparticipation screening for previous episodes of low back pain may enable preventative strategies to be implemented. Whether these prove successful will need to be evaluated.
As rowers age, the volume of training may need to be more closely monitored.
Contributors CN undertook the study, collected the data and contributed to 50% of writing the paper. DR supervised the study and contributed to 30% in writing the paper. PP provided statistical support and contributed to 20% in writing the paper.
Funding This research was funded by a New Zealand High Performance Sport Prime Ministers Scholarship.
Competing interests None.
Ethics approval Auckland University of Technology Ethics Committee.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.