Ehlert, Alex; Wilson, Patrick B.

Introduction

Warm-ups are protocols designed to prepare the body physically for subsequent activity (²⁸). Up until a few decades ago, a lack of empirical data forced coaches and athletes to rely on trial-and-error to develop warm-up protocols (²⁹), but since then, many studies have assessed the potential benefits of warm-ups. Warm-up protocols vary in design and are often context specific but tend to have 2 goals: (a) improve performance and (b) reduce risk of injury (^1,28,29,48). Warm-ups are believed to improve performance through a few key physiological mechanisms, including improved VO₂ kinetics, enhanced muscle metabolism, and postactivation potentiation (²⁹). Potential mechanisms through which warm-ups could reduce injury include enhanced stretch tolerance, greater energy absorption of the musculotendinous unit, and improved flexibility (⁴⁸).

Athletes from almost every sport would likely benefit from engaging in a proper warm-up, but it is important that sport-specific recommendations be developed because the demands of training and competition differ between sports. Because golf does not require intense sporting movements such as running or jumping, it is often perceived as a benign sport (^27,42). But, the golf swing requires near-maximal neuromuscular output at high velocities and large ranges of motion (^6,10,24,42), and golfers often accelerate the club-head to velocities of greater than 160 km·h⁻¹ in a time frame of ∼0.2 seconds (^6,24). At these high velocities, each swing imposes substantial shear, compression, and axial torsional loads on the musculoskeletal system. And, if one considers the voluminous number of swings performed during practice and competition (^42,44), it is unsurprising that golf is classified as a moderate risk factor for musculoskeletal injury, with annual injury incidence estimates ranging from 16 to 41% in amateurs to 31–90% in professionals (³⁴). There is evidence that warm‐ups and stretching may reduce injury risk during whole‐body, dynamic physical activity, although it is currently unclear whether this also applies to skill‐based sports such as golf (⁴⁸). Regardless, it is reasonable to suggest that warming-up before golf participation may reduce injury risk.

Enhancement of performance is another potential reason for golfers to warm-up. There is strong evidence that both high-load potentiating activity and longer duration dynamic stretching enhance general measures of strength, power, and athletic performance (^1,28). Although static stretching may negatively affect subsequent strength and power, the resulting increase in flexibility could benefit performance in certain activities, such as golf, that require a large range of motion (^1,28).

Owing to the golf swing's reliance on high-velocity power and rapid force production (^24,42,47) and golf's classification as a moderate risk factor for injury (^6,34,44), it is logical to hypothesize that warm-ups may improve golf performance and mitigate injury risk. Although warm-ups are often recommended by coaches, physical therapists, and researchers, there has been little consensus on which protocols will benefit golfers. Therefore, the objective of this systematic review is to gather the available data on warm-ups in golf to answer the following questions: (a) What are the current warm-up behaviors of golfers?; (b) Is there an association between warm-up behavior and injury risk?; and (c) What are the effects of various warm-up protocols on measures of golf performance? For the purposes of this review, a warm-up will be defined as any physical activity or stretching before golf play, practice, or performance testing.

Methods

Experimental Approach to the Problem

The systematic review protocol was created in-line with the Preferred Reporting Items for Systematic Review and Meta-analysis Statement (³²) and registered with the PROSPERO International Prospective register of systematic reviews (CRD42018110938).

Subjects

Eligibility Criteria

Primary source, peer-reviewed articles were eligible if they contained data related to warm-ups for golf. Conference abstracts, theses, and dissertations were not included. Specific eligibility criteria included the following:

• Types of studies: Cross-sectional, cohort, case-control, and experimental (randomized or nonrandomized).
• Subjects: Golfers of any age or skill level, including both competitive and recreational.
• Types of outcome measures: To address research question #1, studies were included if they contained data related to warm-up behaviors of golfers, including the participation rates, durations, and specific activities used. To address research question #2, studies were included if they contained data on warm-up behaviors and golf-related injury. Outcome measures could include absolute rates, odds ratios (ORs), or relative risks of injury using warm-up behaviors as an independent variable. To address research question #3, studies were included if they evaluated the effects of a warm-up intervention on golf performance or assessed the association between warm-up behaviors and performance. Outcome measures could include measures of skill (handicap and scoring average), biomechanical variables (X-factor, X-factor stretch, and rotation), or performance measures (clubhead speed [CHS], ball speed, shot distance, accuracy, and shot quality).
• Exclusion criteria: Articles were excluded if they did not meet the inclusion criteria, if they were review articles, or if a full text was not available in English.
• The review was conducted by Old Dominion University, Dept of Human Movement Science.

Procedures

Search Strategy

Three electronic databases (PubMed, SPORTDiscus, and Web of Science) were searched in October 2018 by 3 reviewers independently and a consensus reached through discussion. An overview of the search terms and strategy is displayed in Table 1. Titles and abstracts were screened for relevance first, followed by full-text reading when a decision was unclear. Reference lists of selected full texts were manually screened to ensure articles were not missed during the search.

Table 1

Study Selection and Data Extraction

Abstracts and titles were first independently assessed by all reviewers and coded as “yes,” “no,” or “maybe” in a custom Microsoft Excel spreadsheet. Studies coded with “maybe” were read in full and further screened for eligibility using the inclusion criteria. Rationales for inclusion or exclusion were included for each study retrieved. Reference lists of all full texts were manually searched, and any relevant articles were retrieved and read for potential inclusion. A customized Microsoft Excel spreadsheet was created and used for data extraction. Data extracted from each article included research design, sample size, subject characteristics, independent and dependent variables, and results related to each research question. Details about the warm-up protocols from each study were extracted when sufficient explanations were provided.

Study Quality and Risk of Bias

Study quality of all randomized or counterbalanced experimental designs was assessed using the 11-item PEDro Scale (⁹). Ten of the 11 items on the PEDro Scale are scored as 1 or 0, and a sum score was calculated for each intervention. Sum scores of 0–3 were considered poor, 4–6 were judged as fair, 7–8 as good, and 9–10 as excellent. Studies were not excluded based on quality scores.

Risk of bias was assessed with the Cochrane Collaboration Risk of Bias Tool (²²). The tool is used to assess the likelihood of bias related to allocation and randomization of groups, blinding of subjects and assessors, missing data, and reporting biases, along with any additional sources of bias that may be present. Both the quality and risk of bias were assessed by all 3 reviewers, and consensus reached through discussion.

Synthesis of Results

Articles were grouped into 3 sections depending on the research question they addressed. Some studies were included in more than one section when the data were relevant to multiple questions. Results were organized by outcome variable when applicable. Prevalence data were rounded to the nearest whole number in the text for ease of reading.

Results

The search yielded 274 total hits, with 176 remaining after removal of duplicates. Five articles were identified through reference list searches, yielding a total of 181 articles that were screened by reading the title and abstract. Of these, 129 were rejected based on title and abstract, leaving 52 that were read in full. Full-text screening led to an additional 29 studies being removed; 18 were review articles, 10 did not meet the definition of a warm-up, and 1 analyzed the same sample and data as another article. Ultimately, 23 articles met the inclusion criteria. Nine of the articles were observational studies with data either describing warm-up behaviors or associating warm-up behaviors with injury. The remaining 14 articles were experimental trials that assessed the effects of at least 1 warm-up protocol on measures of performance. An overview of the search, screening, and selection process is outlined in Figure 1.

Figure 1

Research Question #1: Descriptive Data on Golfer Warm-up Behaviors

Study Overview

Seven studies included descriptive data on warm-ups (Table 2). Six studies collected data with retrospective questionnaires, 5 being in written form (^{15,16,19,35,41}), whereas the sixth was orally administered (¹²). The questionnaires asked for subjects to recall information from either the past 12 months (^12,15,16,19) or the past 3 years (^35,41). One additional study observed the warm-up behaviors at golf facilities (¹¹). A wide range of ages were observed in all but one study (³⁵), which included only older golfers (age: 69.9 ± 8.7 years). The subjects were primarily recreational golfers, although 1 study included a subset of 60 professionals (¹⁹).

Table 2

Warm-up Frequency and Duration

The primary findings are displayed in Table 2. Four studies described warm-up behaviors before golf play (^11,12,15,16). In these studies, between 35 and 71% of golfers reported to never, rarely, or seldomly warm-up before play (^12,15,16), and Fradkin et al. (¹¹) observed that 46% of golfers did not perform any sort of warm-up before golf participation. Fradkin et al. (¹²) reported that only 4 and 8% of golfers answered that they always and often warmed up before playing, respectively. Although exact percentages were not explicitly stated in 2 studies (^15,16), bar graphs indicated that about 25–30% of golfers often or always warmed up before playing. Fradkin et al. (¹²) further compared warm-up behaviors by age and skill level; golfers aged 35–55 years were most likely to report warming-up, whereas those who were 56 years or older were least likely. Moderately skilled golfers (handicap of 16–27) were more likely to warm-up than novices (>27 handicap) or those with lower handicaps. Gosheger et al. (¹⁹) also found that golfers who warmed up for greater than 10 minutes had a better average handicap than those who did not (14.3 vs. 22.0). Two studies asked about warm-up behaviors before practicing golf, finding that 80 (¹⁵) and 63% (¹⁶) of golfers reported to never or seldom warm-up before practicing.

Four studies analyzed warm-up durations (^16,19,35,41). Gosheger et al. (¹⁹) found that 17% of recreational golfers and 42% of professionals warmed up for greater than 10 minutes. Palmer et al. (³⁵) reported that 17 and 36% of golfers spent less than 1-minute warming-up and stretching, respectively, and 75% spent less than 5-minute total warming-up. Fradkin et al. (¹⁶) found similar results in that the most common warm-up duration was less than 30 seconds (36%), followed by 30–59 seconds (23%) and 60–120 seconds (21%). Finally, Silva et al. (⁴¹) reported longer warm-up durations of just under 10 minutes, of which about 4 minutes was stretching.

Warm-up Activities

Three studies assessed the number of activities the golfers performed during warm-ups (^11,15,16). Of the golfers who used warm-ups, most performed 1 activity, whereas 17–18% performed 2 (^11,15,16). Golfers who reported a golf-related injury in the past year were less likely to perform more than 1 warm-up activity before play (10%) or practice (11%) (¹⁶).

For golfers who reported warming-up, the most common activities were air swings and stretching, with 23–89% and 23–90% of golfers including them in their warm-up, respectively (^11,12,15,16). Only 1 study differentiated between types of stretches (¹¹), finding that 17 and 5% of warm-ups included dynamic and static stretching, respectively. Aerobic exercise was the least commonly reported activity, as 0–2% of golfers included it in their warm-ups (^11,12,15,16). Although only discussed in 1 study (¹²), 27% of golfers reported ball hits as part of their warm-up.

Factors Influencing Warm-up Behaviors

Two studies asked golfers to expand on factors influencing warm-up behaviors (^12,15). Most believed that warming-up could reduce injury risk (79%) or improve performance (67%) (¹⁵). Fradkin et al. (¹²) asked golfers to provide reasoning behind why they typically do or do not perform a warm-up. The most common reasons for warming-up included as follows: to play better (75%), to prevent injury (27%), because everyone else does (13%), or because they had been told to (5%). Reasons that golfers chose not to warm-up included they do not need to (39%), they did not have enough time (36%), they could not be bothered to (34%), they do not know how (10%), no one else does (7%), or they do not believe warm-ups work (4%). In terms of knowledge, 84 and 85% of golfers reported not knowing how to properly warm-up to reduce injury risk and improve performance, respectively (¹²). They also found that those who reported knowing how to warm-up were 3 times more likely to perform one than those who did not.

Research Question #2: Warm-up Behaviors and Golf-Related Injury

Study Overview

Seven studies analyzed or described the association between warm-up behaviors and golf-related injury (^{15,16,19,30,31,35,41}). An overview of the injury rates from each study is displayed in Table 3. All were observational designs, with 6 using retrospective questionnaires to identify self-reported warm-up behaviors and golf-related injury over the past 1–3 years (^{15,16,19,30,35,41}). The only prospective study (³¹) was a questionnaire-based follow-up to one of the other studies (³⁰).

Table 3

Sample sizes ranged from 61 to 1,634 golfers and were mixed-sex, except for one that recruited only amateur female golfers (¹⁵). Most golfers were amateur or recreational, although Gosheger et al. (¹⁹) included a subset of professionals (n = 60). Most studies included a wide range of ages except for Palmer et al. (³⁵), who recruited older golfers (age: 69.9 ± 8.7 years).

Injury Overview

Overall, 13–60% of golfers reported at least 1 injury (Table 3). Professional golfers reported the highest prevalence of injuries, with 36 of 60 professionals reporting a total of 110 injuries (3.06 injuries per golfer) (¹⁹). The second highest rate was in the sample of older golfers, with 50% reporting a musculoskeletal condition in the previous 3 years (³⁵). In addition, McHardy et al. (³⁰) found that golfers older than 40 years were at a higher risk of injury, whereas young golfers (<20 years) were at the lowest risk. Gosheger et al. (¹⁹) also reported that most injuries were minor (52%) and resulted in symptoms for less than 1 month (57%), although nearly a quarter of golfers had symptoms for 1–6 months and between a 1-week and 1-month lay-off from golf. An additional 19% had symptoms that persisted for greater than 6 months. McHardy et al. (³⁰) found that 7% of injured golfers sought treatment for their injury, and over half (55%) had to temporarily halt golf participation. Only 1 study analyzed sex differences and found no difference in injury rates (³⁰).

Warm-up and Injury

An overview of the results is found in Table 4. Gosheger et al. (¹⁹) reported that golfers who spent at least 10 minutes of warming-up were less likely to report an injury than those who did not (0.43 vs. 1.02 injuries/golfer). Two studies assessed warm-up duration based on injury status and found that golfers with a previous musculoskeletal condition warmed up longer than those who did not report an injury (^35,41).

Table 4

The remaining studies assessed warm-up behaviors as a risk factor for injury using chi-square analysis and ORs (^15,16,30,31). Fradkin et al. (¹⁵) reported that warm-up behaviors performed before both play and practice were significantly associated with lower injury risk. Adjusted ORs indicated that never or seldomly warming-up had a strong association with risk of injury when performed before play (OR: 45.2; 95% confidence interval [CI]: 13.5–151.7) and practice (OR: 3.4; 95% CI: 1.0–12.2). A second study (¹⁶) partially supported these results, finding that warm-up behaviors were significantly associated with injury before play and practice. The adjusted ORs indicated that not warming-up before play was strongly associated with injury, but to a lesser degree than the previous study (OR: 3.2; 95% CI: 1.0–8.5). Surprisingly, McHardy et al. (³⁰) found that range of motion activities before play were associated with increased injury compared with no warm-up (OR: 1.6; 95% CI: 1.2–2.2), although the 1-year follow-up found no significant association between warm-up behaviors and injury (³¹). Overall, the results on warm-ups and injury risk were mixed and inconclusive.

Research Question #3: Warm-up and Performance

Study Overview

Fourteen studies assessed the effects of warming-up on performance. All studies were experimental, 8 being counterbalanced cross-over designs (^{8,17,18,25,33,38,43,45}), 2 using within-subject repeated measures (^3,20), and 4 being randomized controlled trials with parallel groups (^7,14,21,37). A summary of the PEDro scale scoring is displayed in Table 5. Overall, 3 studies were of poor quality, 8 were fair, and 1 was good.

Table 5

In addition, the overview of responses using the Cochrane Risk of Bias Assessment is displayed in Figure 2. The 2 repeated-measures designs (^3,20) were excluded from the PEDro scale and Cochrane risk of bias scoring, as the assessments are designed for randomized or counterbalanced designs (^9,22).

Figure 2

An overview of the specific warm-up protocols used in each study can be seen in Table 6, whereas findings are displayed in Table 7. Outcomes were divided into biomechanical variables, measures of shot quality, shot accuracy, CHS or clubhead velocity (CHV), carry and total shot distance, and other performance variables.

Table 6

Table 7

Biomechanical Variables

Four studies included at least one biomechanical measure of performance (^20,21,33,37). Moran et al. (³³) found that a dynamic warm-up resulted in straighter swing paths than static stretching and no stretching, but no differences were observed in club face angle at impact. Green et al. (²⁰) measured the effects of a 500-m walk on several biomechanical variables in a subset of their sample (n = 6) and found that right knee rotation decreased by about 2.5°, but no other changes were observed. A comparison of a dynamic, rotation-specific golf warm-up and a sham warm-up found that both significantly increased X-factor stretch, but neither influenced X-factor (²¹). Finally, trigger point therapy in combination with a medicine ball warm-up clinically improved backswing hip rotation while the control group declined (³⁷).

Shot Quality

Five studies included a subjective measure of shot quality (^{8,17,18,43,45}). The first method involved asking golfers whether they struck the ball well or not as a “Yes” or “No” after each shot. An outcome score was calculated as the number of “Yes” answers out of 10 shots (^17,18,43,45). The second method asked golfers to rate the quality of each strike from 0 to 10, and an average shot quality score was calculated (⁸). Gergley (¹⁷) found a significant decline in shot quality when static stretching was added to an active dynamic warm-up (−16.3%). A follow-up to that study using the same warm-up protocol also found an acute decline in shot quality (−31.3%), which remained impaired at 15, 30, 45, and 60 minutes after stretching (¹⁸). Tilley and Macfarlane (⁴⁵) found that a functional resistance-band warm-up significantly increased shot quality compared with an active dynamic warm-up (+13.4%) and a weightlifting warm-up (+12.1%). Dynamic stretching tended to result in greater reports of shot quality than static stretching, although it did not reach significance (⁴³). Finally, Coughlan et al. (⁸) reported that a dynamic warm-up resulted in significantly higher shot quality than a golf club-only warm-up (+40.0%) and no warm-up (+40.0%).

Four studies also assessed objective measures of shot quality as distance between the impact point and the center of the clubface (^20,33) and smash factor ratio (ball velocity/CHV) (^37,45). Moran et al. (³³) reported that dynamic stretching resulted in more central impact points than static stretching (Δ0.7 cm), whereas Green et al. (²⁰) found no changes after a 500-m walk. A functional resistance-band warm-up resulted in greater smash factor ratio compared with an active dynamic warm-up (+1.5%) and an active dynamic warm-up with the addition of weightlifting (+0.7%) (⁴⁵), although Quinn et al. (³⁷) found no significant differences between trigger point therapy with either stretching or medicine ball exercises and a control group.

Accuracy

Accuracy following warm-ups was assessed in 8 studies (^{3,17,18,20,25,37,43,45}), defined as shot displacement from a target line. When static stretching was added to an active dynamic warm-up, it significantly impaired accuracy (^17,18), and it remained impaired 15, 30, and 45 minutes later (¹⁸). A dynamic stretching warm-up resulted in greater accuracy than static stretching (⁴³), and trigger point therapy with either stretching or medicine ball exercises had greater improvements in accuracy than a control group (³⁷). However, no differences in accuracy were observed in the remaining 4 studies (^3,20,25,45).

Clubhead Speed/Velocity

Eight studies assessed CHS or CHV (^{8,14,17,18,33,37,38,45}). Moran et al. (³³) found that dynamic stretching resulted in greater CHS than static stretching (+1.9 m·s⁻¹) and no stretching (+1.7 m·s⁻¹). Fradkin et al. (¹⁴) reported a greater CHS with a multicomponent golf warm-up group compared with a control group that did not warm-up. Two studies observed the effects of adding static stretching to an active dynamic warm-up and found it impaired CHS acutely (^17,18) and after 15 and 30 minutes, but not at 45 or 60 minutes (¹⁸). Read et al. (³⁸) found that the addition of postactivation potentiation in the form of 3 countermovement jumps resulted in increased CHS (+2.2%). Similarly, Tilley and Macfarlane (⁴⁵) found increases in CHS after a functional resistance-band warm-up and weightlifting warm-up compared with an active dynamic warm-up, but it did not reach statistical significance. Finally, Coughlan et al. (⁸) found significantly higher CHS following the combination of a club and dynamic warm-up compared with a control condition (+1.1%).

Carry and Total Shot Distance

Shot distance was assessed through carry distance or total distance. Bunker et al. (³) reported a significantly enhanced carry distance after whole-body vibration. Langdown et al. (²⁵) found a small but insignificant increase in carry distance when comparing dynamic and resistance-band conditions with a control, but there were no significant differences between the dynamic and resistance-band conditions.

Total distance was assessed in 8 studies (^{3,7,17,18,20,37,43,45}). Costa et al. (⁷) found that a stretching-only group had a significant decrease in distance at week 2 measurements, whereas the addition of spinal manipulative therapy resulted in slight increases over time. The addition of static stretching to an active dynamic warm-up impaired distance acutely (^17,18) and at 15, 30, 45, and 60 minutes after stretching time points (¹⁸). The addition of whole-body vibration after a typical warm-up led to a significant increase in distance (³). Similar findings were observed with the addition of a 500-m walk in more competitive golfers, although the less experienced golfers had no significant changes (²⁰). When comparing warm-up protocols, Sorbie et al. (⁴³) found greater distance after dynamic stretching than static stretching. A dynamic warm-up with the addition of functional resistance-band condition also resulted in greater distance than when weight-training exercises were added (+4.8%) as well as the dynamic warm-up alone (+5.6%) (⁴⁵). The only study to find no significant changes in distance was Quinn et al. (³⁷), who found similar distances between groups administered trigger point therapy with stretching, trigger point therapy with medicine ball exercises, and a control group.

Other Shot Performance Variables

Ball speed or velocity immediately following impact was assessed in 3 studies (^3,25,33). Moran et al. (³³) found greater ball speed after dynamic stretching compared with static stretching (+3.5 m·s⁻¹). Bunker et al. (³) assessed ball speed after a standard warm-up and then again after whole-body vibration, finding significant improvements following a whole-body vibration protocol. Langdown et al. (²⁵) found no differences in ball velocity after dynamic or resistance-band warm-ups, but both were greater than a control condition with no warm-up.

Spin rate and launch angle after impact was assessed in 2 studies (^3,25). No significant differences in the spin rate were observed in either study. A dynamic warm-up condition did result in greater launch angle compared with no warm-up, but no other differences were observed (²⁵).

Discussion

The purpose of this review was to summarize the current state of the literature regarding warm-ups for golf and to parlay the findings into practical recommendations as well as ideas for future research. Twenty-three total articles were reviewed (9 observational and 14 experimental) to answer the following 3 research questions: (a) What are the current warm-up behaviors of golfers?; (b) Is warm-up behavior associated with golf-related injury?; and (c) What are the effects of warm-up protocols on measures of golf performance?

Overall, data from 7 observational studies suggest that many golfers do not consistently warm-up before play or practice. Those who did warm-up tended to perform only one type of activity for short durations. The most common types of activities were air swings and stretching of some form. Although aerobic activity is recommended before physical activity (²⁹), 2% or less of golfers included it before golf participation (^11,12,15).

Professional golfers likely have the highest risk of injury (^19,39,44), because of the high swing velocities and the large practice and competition loads that leave them susceptible to overuse injury (^19,39,44). Despite the increased risk, little data exist on warm-up behaviors of competitive or professional golfers. One study in this review found that a small subset of professional golfers warmed up for greater durations than recreational or amateur golfers (¹⁹). Clearly, more data are needed to determine the warm-up behaviors of competitive golfers before practice and competition.

Another group that is considered at high risk of injury is the large population of older golfers (^4,44), mainly due to a cumulative stress from repetitive swings combined with physiological changes that accompany aging, such as reductions in strength, power, and coordination (^4,36,44). Despite the higher risk of injury, the available data indicate that older golfers are less likely to warm-up than their younger counterparts (¹²). Given the large number of older adults who play golf (⁵) and their higher risk of injury (^4,19,44), future attempts should be made to evaluate whether warm-ups may mitigate injury risk and explore options to increase warm-up compliance.

Although warm-up behaviors were generally poor, most golfers believed that warming-up is beneficial (¹⁵). Improving performance and reducing risk of injury were also the 2 most common reasons given for why golfers chose to warm-up (¹²). Golfers who did not warm-up typically stated that they did not need to, did not have the time, or could not be bothered. In addition, most golfers did not know how to properly warm-up in a way that could improve performance or injury risk (¹²). However, golfers were 3 times more likely to warm‐up prior to golf play if they reported to know how to warm‐up properly. Given this information, it is possible that evidence-based education on the benefits of warming-up, as well as instruction on how to warm-up in a time-efficient manner, may improve the quantity and quality of warm-up behaviors.

Most of the data were collected using retrospective questionnaires. This is a concern because questionnaires relying on recall are often unreliable, especially if people are asked to remember information from distant memory (⁴⁰). Self-report may also elicit different answers depending on the questionnaire format, order, and wording of the items (⁴⁰). For example, many of the studies did not clearly define the types of warm-up activities, and only one study differentiated between static and dynamic stretching (¹¹). In addition, most of the studies were primarily concerned with golf-related injury and not warm-up behaviors specifically and, as a result, included only 1 or 2 basic descriptors of warm-ups such as average duration.

Moving forward, researchers should clearly define warm-up variables and consider using physical observations and longitudinal methods. Likewise, more data are needed on the warm-up behaviors of competitive and professional golfers, as this was seen in only one study (¹⁹). This may be particularly important because more elite golfers produce high levels of CHS, experience more stressful loads on the musculoskeletal system (^23,26), and have the highest risk of injury (^4,44). Finally, future research may address whether educational or instructional programs lead to improved warm-up behaviors.

Based on the studies identified in this review, the current data on the association between warm-up behaviors and injury in golf are conflicting. Gosheger et al. (¹⁹) found that golfers who reported to warm-up less than 10 minutes also reported more injuries than those who warm-up for longer durations. However, Palmer et al. (³⁵) and Silva et al. (⁴¹) found opposite associations, as those who reported a history of musculoskeletal conditions warmed up for a greater duration than those who did not. Similarly, a mix of results was observed when analyzing warm-up behaviors as a risk factor for injury. A retrospective study of Australian golfers found that performing range of motion activities was associated with a higher risk of injury (³⁰). The prospective follow-up to this study found no association between warm-up behaviors and injury over the subsequent year (³¹). Interestingly, 2 studies by Fradkin et al. (^15,16) found the opposite, with failure to warm-up significantly associated with increased injury risk.

All the studies were observational, meaning that no firm causal inferences can be made from the data. For example, Palmer et al. (³⁵) and Silva et al. (⁴¹) found that those with an injury history also reported longer warm-up durations, but it is unknown whether the warm-up behaviors contributed to the injuries or whether the golfers chose to spend more time warming-up as a response to previous or current injuries. Conflicting results between the McHardy and Fradkin studies (^15,16,30,31) also make it unclear whether warm-up behaviors are a significant risk factor for injury. As mentioned previously, there are limitations to self-report data that may have impacted the findings (⁴⁰). It is also possible that the way injury was defined (Table 3), how the questions were asked, and how the data were analyzed may have contributed to the differences between the 2 authors. More clear definitions on what constitutes a warm-up or range of motion activity would allow for a more straightforward interpretation. In sum, due in large part to a lack of longitudinal and experimental research, it is unclear whether warm-up behaviors significantly reduce, increase, or have a null effect on the risk of golf-related injury.

Several steps should be taken to improve the quality of data on warm-up behaviors and injury risk. First, clear and more consistent data collection procedures and definitions of variables are needed for comparative purposes and to allow for the aggregation of data in meta-analyses. As it currently stands, the ways in which researchers have defined injuries have varied substantially, making it difficult to compare results between studies. In addition, as mentioned previously, the definition of what constitutes a warm-up is often vague or undefined, and different warm-up activities are often grouped together into broad categories. It is possible that different warm-up activities (dynamic stretching, aerobic activity, and resistance exercise) may have different effects on injury risk through different mechanisms and should be studied as isolated variables.

Second, the research would be substantially improved if injuries were diagnosed—or at a minimum confirmed—by a qualified research team member, as opposed to relying solely on golfer self-reports. Obviously, this would impose additional burdens on both research subjects and investigators, but it is clear from the current literature that the reliance on self-reports (especially retrospective questionnaires) is hampering the synthesis of clear conclusions and recommendations. Similarly, more longitudinal data are needed; in some cases, golfers were asked to recall injuries from up to 3 years before (^35,41). Preferably, some of these longitudinal studies would use experimental designs so that causality can be determined, but this may be challenging to do in elite golfers. Finally, future research needs to delineate between recreational, competitive, and elite golfers. The only study identified in this review that included data on professionals did not separate them from the remainder of the sample when analyzing the association between warm-up duration and injury rates (¹⁹).

Fourteen experimental studies assessed the effects of warm-up protocols on measures of performance. The measures were not direct performance indicators (score or handicap over time), but they are commonly regarded as valid proxy measures of golf performance. Distance off the tee and CHS are considered vital components of golf success because of close associations with skill level and on-course performance (^2,13). Other measures included were accuracy and shot quality, which are important to assess because a common concern among golfers is increased CHS or distance will be accompanied by decrements in shot quality and accuracy. An ideal warm-up would enhance CHS and distance while simultaneously having a positive or at least minimal negative effect on accuracy and shot quality.

Although the warm-up protocols used, outcomes assessed, and research methodology was inconsistent, a few general themes emerged from this review. First, static stretching appears to be a suboptimal warm-up approach for improving golf performance. Static stretching was outperformed by dynamic stretching (^33,43), and the addition of static stretching to an active dynamic warm-up resulted in impaired CHS, driving distance, accuracy, and self-reported shot quality (^17,18). Moreover, performance on these metrics remained impaired for up to 30–60 minutes (¹⁸). In contrast to static stretching, dynamic warm-ups tended to enhance measures of performance compared with control conditions or club-only warm-ups (^8,25,33). Moran et al. (³³) found that dynamic stretching produced greater CHS, ball speed, and straighter swing paths than static stretching or no stretching. Similarly, Coughlan et al. (⁸) and Langdown et al. (²⁵) found that dynamic exercise warm-up conditions resulted in superior results compared with either no warm-up or a club-only warm-up. The findings are in-line with the current literature, and recommendations suggesting dynamic stretching is often superior to static stretching before strength or power activities (^1,28).

A second general theme to emerge is that resistance-type exercise may be a viable warm-up strategy. In nongolf research, resistance exercise before activity has improved strength, power, and performance through a potentiation effect (²⁹), and the results of 2 studies from this systematic review suggest these effects may extend to golf. Specifically, Tilley and Macfarlane (⁴⁵) assessed the effects of adding barbell weight-training exercises and resistance-band exercises to an active dynamic warm-up consisting of weighted air swings and practice golf shots. The addition of the weight-training exercises resulted in greater CHS, but not distance, accuracy, or smash factor ratio. Although this could potentially be an effective warm-up strategy, barbells are not readily available at most golf facilities. Instead, the addition of resistance-band exercises may be a viable alternative. In fact, the investigation by Tilley and Macfarlane (⁴⁵) showed the addition of resistance-band exercises produced greater distance, smash factor ratio, and self-reported shot quality in comparison with an active dynamic warm-up of weighted air swings and practice golf shots alone or when weight-training exercises were added to the dynamic warm-up. In terms of CHS, both weight-training and resistance-band exercises resulted in greater performance than a dynamic warm-up alone. The second study that showed a benefit of resistance-type exercise was from Langdown et al. (²⁵), who reported that both resistance-band and dynamic exercise warm-up protocols were more effective at increasing ball velocity than a control condition.

Clearly, more studies are required to make firm conclusions about the benefits and risks of resistance exercise before golf, but the limited current research suggests it has a positive or neutral effect on most measures of performance. On a practical note, resistance-band exercises may be a feasible addition to a warm-up, as they are easy to transport and can be used anywhere. Finally, the addition of 3 countermovement jumps to a warm-up resulted in a moderate increase in CHS (³⁸). Therefore, including a form of postactivation potentiation within a warm-up may be beneficial before golf participation.

Beyond dynamic stretching and resistance exercise, several other studies reported significant performance improvements. Two studies used within-subject, repeated-measures designs to assess golf performance after a self-selected warm-up and then again after the protocol of interest (^3,20). These studies reported improvements in performance after the addition of whole-body vibration stretches (³) and a 500-m walk (²⁰). Unfortunately, the lack of counterbalancing makes it difficult to determine whether the improvements were due to the protocol being superior to the standard warm-up, due to an order effect, or simply due to more thorough warming before testing.

A final area of the research worth commenting on is aerobic warm-ups. Previous research has highlighted that aerobic activity during warm-ups may increase muscle temperature, resulting in improvements in physical performance (²⁹). Increased muscle temperature is associated with increased muscle fiber conduction velocity and rate of force development in type II fibers (²⁹). As CHS is associated with rate of force development and impulse (^46,47), raising muscle temperature through aerobic activity during a warm-up could, in theory, be beneficial for a golfer. The descriptive data suggest that very few golfers perform any sort of aerobic warm-ups (^11,12,15). Several of the experimental studies included aerobic activity as part of the warm-up protocols (^14,33,43), but only one assessed aerobic activity in isolation (²⁰). They found that driving distance was greater after the addition of a 500-m walk in a more competitive group of golfers, but not in a group that played infrequently. As noted above, the lack of counterbalancing makes interpretation difficult, but it is possible that aerobic activity could be of benefit as part of a warm-up.

Future research should continue to compare different warm-up protocols on measures of performance with the addition of a control group or condition. Although aerobic activity and whole-body vibration showed significant improvements (^3,20), they should be assessed using randomized or counterbalanced designs. Likewise, it will be important for researchers to determine how long the benefits of warm-up protocols persist. Gergley (¹⁸) and Moran et al. (³³) assessed results over various time points, but more studies are needed. A round of golf generally lasts 2.5–6 hours (⁴²), and short-lasting improvements in CHS, ball speed, accuracy, etc., may offer only minimal overall benefits over an entire round. If effects are short-lived, studies may want to assess “rewarming” protocols, which could be used at various time points during a round to enhance overall performance. Finally, Fradkin et al. (¹⁴) reported the potential of a conditioning effect from regularly performing their three-part warm-up protocol; therefore, future research should explore whether regularly warming-up has chronic effects on performance. The authors should consider reporting whether subjects currently warm-up in a similar fashion. It is plausible that acute or chronic responses to warm-up protocols could be greater if it provides a novel stimulus that the golfer has not experienced previously. In addition, as golfers may be more likely to perform warm-up protocols that are relatively short in duration, the authors should consider reporting data on how long the warm-ups took to complete.

Practical Applications

The results of this review suggest that many recreational golfers do not warm-up before golf play or practice. Of those who do, warm-ups tend to be short in duration and usually involve air swings or stretching of some form. Many golfers believe that warming-up has value for injury risk or performance enhancement, yet most seem unsure of how to warm-up properly to achieve these benefits. Golfers who report knowing how to properly warm-up seem more likely to follow through and warm-up regularly. Given this information, educational sessions or seminars could be used to improve the warm-up behaviors of recreational golfers moving forward, although this should be evaluated in future research.

Given the mixed results with warm-up behaviors and injury, it is difficult to make practical suggestions until more data are available. That said, the injury prevention potential of warming-up is supported by research done outside of golf (⁴⁸), so if a golfer has to choose between warming-up or not, opting to warm-up seems like the prudent choice. Still, much more work is needed to clarify the optimal types of warm-ups for golfers. In contrast to the lack of research on warm-ups and injury, there is decent evidence supporting the use of various warm-up protocols for performance enhancement. Specifically, warm-ups should include dynamic stretching or activity, and possibly some form of light resistance exercise to enhance several measures of golf performance, whereas static stretching should be avoided. From a practical standpoint, most golfers will not have access to a full gym before playing golf, so warm-ups that require minimal equipment are preferable. Many dynamic exercises can be performed with just body mass or with the assistance of a golf club on the range. Resistance bands could be a useful method of adding resistance exercise to a warm-up, as they are relatively inexpensive, easy to transport, and can be used at the practice facility of a golf course before play. Future research should aim to determine optimal warm-up protocols, examine the duration of performance benefits, explore the potential of “rewarming” protocols throughout the course of a golf round, and assess the role of various warm-up behaviors on risk of golf-related injury. In addition, efforts should be made to improve the warm-up behaviors of recreational golfers moving forward.

Acknowledgments

No sources of funding were used to assist the preparation of this systematic review.

Alex Ehlert and Patrick Wilson declare that they have no conflict of interests relevant to the content of this systematic review.

The authors thank Jaison Wynne for assisting with the literature search and article screening process.

References