Head Injury and Aging: The Importance of Bleeding Injuries

Abstract

The current study analyzed 1993–2007 data from NASS/CDS (National Automotive Sampling System / Crashworthiness Data System) to explore the types of serious head injuries sustained by adult motor vehicle crash occupants and how the types of head injuries sustained shifted with age. The purpose was to determine which head injuries are most important for older occupants by identifying specific injuries that become more likely for aging occupants and taking into consideration previous reports on the potential outcome of those injuries for an older population. Results confirmed previous reports that older head injury victims in motor vehicle collisions were more likely to sustain bleeding injuries than younger head injury victims. The current study showed that, in particular, the rate of extra-axial bleeding injury (which includes epidural, subdural, and subarachnoid bleeding) increased with age. The increase in extra-axial bleeding injury rate was especially prominent in relatively low Delta-V crashes. Among the extra-axial bleeding injuries that had increased odds of injury for older occupants, subdural hematoma and subarachnoid hemorrhage were notable, with increased odds of injury for occupants age 50 to 69 as well as for occupants age 70 and older. The importance of subdural hematoma for aging occupants is emphasized by previous studies showing its high mortality rate, while the impact of subarachnoid hemorrhage is linked in previous studies to its aggravating effect on other injuries. The results highlight a need to further explore the injury mechanisms of subdural hematoma and subarachnoid hemorrhage in older occupants in order to define age-adjusted injury tolerance and develop countermeasures.

INTRODUCTION

Head injury has been shown to produce worse outcomes with increasing age [ Hukkelhoven, Steyerberg, Rampen et al., 2003 , Jane and Francel, 1996 , Mosenthal, Lavery, Addis et al., 2002 , Vollmer, Torner, Jane et al., 1991 ]. Epidemiological studies show that, of patients with head injuries, older patients sustain different types of head injuries than younger patients, with an increase in intra-cranial bleeding among older patients [ Coronado, Thomas, Sattin et al., 2005 , Vollmer et al., 1991 ]. Studies of motor vehicle crash occupants have also shown that younger drivers were more likely than older ones to suffer concussions, and older drivers had a significantly higher frequency of intracranial hemorrhages than younger drivers [ Bauza, LaMorte, Burke et al., 2008]. These hospital studies focused only on these broad types of head injuries and did not report the frequency of specific types of intracranial bleeding injuries. The studies also did not account for impact severity or exposure (i.e. the total number of drivers in crashes, including those who were not injured ).

The current study analyzed data from NASS/CDS (National Automotive Sampling System/Crashworthiness Data System) to explore serious head injuries in adult occupants and the effect of increasing age on the incidence of specific head injuries. The study used Delta-V (velocity change) from NASS/CDS cases to represent crash severity and included both injured and non-injured occupants to estimate exposure. This exposure information allowed for calculation of injury rates among all occupants in motor vehicle crashes, rather than among only those with head injuries.

The purpose of this study was to determine which head injuries are most important among older vehicle occupants by identifying types of injuries that become more likely for aging occupants in crashes. In particular, specific intra-cranial bleeding injuries were compared, taking into consideration previous reports on the severity and potential outcome of those injuries for an older population.

METHODS

A hierarchical approach was used to identify which broad injury types were most common for different adult age groups, then to narrow in on types of injuries and then to specific injuries most frequent among older vehicle occupants.

Source Data

The data in this study were drawn from NASS/CDS from 1993 to 2007. Analysis was performed using SAS, Version 9.1 (SAS Institute Inc, Cary NC).

Rollover and ejection cases were eliminated because of the possible confounding differences in crash conditions between young and older occupants in these crashes. Rollovers were excluded by eliminating cases where the primary type of damage was attributed to rollover/overturn. Ejection and partial ejection cases were excluded by the limitation that the EJECTION variable was equal to 0 for all included cases.

Adults aged 20 and older were included and children and teenagers were excluded. Fatal cases were not excluded from the study. Although non-injury cases were retained in order to calculate injury rates, cases were excluded if the total number of injuries was coded as ‘97’, which indicated that the occupant was injured but with details unknown.

Occupants in all seat positions were included, regardless of restraint use. Cases with unknown Delta-V, impact direction, or vehicle model year were only excluded from analyses that included those variables as specified in the Data Analysis section.

Head Injury Classification

All serious head injuries (REGION90=1) were included based on NASS/CDS AIS (Abbreviated Injury Scale) severity codes 3 to 6. Multiple injuries were included for occupants with more than one serious injury coded. The AIS codes used in NASS/CDS were based on the Association for the Advancement of Automotive Medicine’s AIS-90 from 1993 to 1999, and on AIS-90/98 Update from 2000 to 2007. Head injuries were classified by seven-digit AIS code into broad categories as follows.

• Intra-axial bleeding (within the brain)

  Cerebellum or cerebrum contusion/hematoma/hemorrhage, intra-ventricular hemorrhage, cerebrum or cerebellum laceration or penetrating injury, vessel injury deep to the meninges

• Extra-axial bleeding (outside the brain)

  Epidural hemorrhage, subdural hematoma, subarachnoid hemorrhage and subpial hemorrhage

• Bleeding, Location unknown

  Cerebrum/cerebellum hematoma or hemorrhage, location not further specified

• Fracture

  Crush, fracture

• Closed head injury

  Loss of consciousness or concussion, diffuse axonal injury, or closed head injury/blunt head injury/ traumatic brain injury not further specified

• Swelling and other sequelae

  Swelling, ischemia, pneumocephalus

• Brain stem injuries

  Any injury to the brain stem

• Other

  Any other injury to the AIS head region not included in the categories above, such as pituitary injury or serious scalp injury.

For additional analysis of extra-axial bleeding injuries, this category was further classified into sub-categories: subdural hematoma, epidural hematoma, and subarachnoid hemorrhage.

Data Analysis

The first step of the hierarchical analysis was analysis of relative injury frequency to determine which broad injury types were most common for different adult age groups by decade. Bleeding injuries, identified in the relative frequency analysis as common among older occupants, were the focus of the second step of analysis: injury rate and the effect of speed. Step three of the analysis, odds ratio analysis, focused in still further on extra-axial bleeding injuries by estimating the increased odds of specific extra-axial injuries for older occupants.

Relative Injury Frequency by Age

Relative frequency of serious head injuries of different broad categories was compared for age groups by decade. The SAS SURVEYFREQ procedure was used to calculate the proportion of head injuries by category for each age decade. SAS survey analysis procedures account for the stratified probability sampling procedure used in NASS/CDS and produce estimates based on weighted NASS/CDS data.

For each age group, the weighted proportions of serious injuries in each injury category type were plotted. Analysis was performed on counts of all serious injuries, not on the number of seriously injured occupants. Standard errors for each proportional estimate were tabulated.

Injury Rate and the Effect of Speed for Bleeding Injuries

Injury rates were estimated for bleeding injuries, which were identified in the relative injury frequency analysis as more frequent among older occupants. The injury rate represents the total number of occupants in a selected dataset with at least one head injury of a given type divided by the total number of occupants in the same dataset. Three adult age groups were selected (20–49, 50–69, 70+) based on the results of the relative injury frequency analysis that suggested shifts in head injury patterns at these approximate age thresholds.

Three speed-change categories were used, based on the NASS/CDS Delta-V variable DVTOTAL: 0–19 km/h, 20–39 km/h, and 40+ km/h. The SAS SURVEYFREQ procedure was used to estimate injury rate for each of the three age groups at different levels of crash Delta-V. Injury rate and crash frequency at different levels of Delta-V were compared, and presented along with the total proportion of bleeding injuries occurring to occupants in different age groups. Occupants in crashes with unknown Delta-V were excluded from the analysis of the effect of speed.

Standard error was calculated for the rate estimates and shown on the bar plot for each rate.

Odds Ratio Analysis for Extra-axial Bleeding Injuries

For extra-axial bleeding head injuries, identified in the relative frequency analysis and injury rate analysis as frequent among older occupants, an odds ratio analysis was performed. Increased odds of injury for specific extra-axial injuries among the older age groups relative to the youngest age group were estimated.

The SAS SURVEYLOGISTIC procedure was used to estimate odds ratios for sub-categories of extra-axial bleeding injuries: subdural hematoma, epidural hematoma, and subarachnoid hemorrhage. A model was developed for each specific type of bleeding injury, where the dependent variable was a dichotomous categorical variable equal to one for occupants with a given injury and zero otherwise. Categorical variables for age group (age 50–69 and age 70+) were compared to a baseline 20–49 year-old group. For example, the SAS logistic model for extra-axial bleeding injury was EAXIAL(EVENT=’1’) = AGEGROUP. Ninety-five percent confidence limits for the odds ratios were based on variance calculated by the Taylor series. In order to correct for simultaneous estimation of six confidence intervals for odds ratios (two age groups for three different injuries), Bonferroni correction was applied by dividing the significance level of 0.05 by 6 so that alpha was equal to 0.0083.

To account for possible confounding differences between crash and vehicle conditions for younger and older vehicle occupants, an adjusted logistic regression model was developed with additional independent variables: Delta-V, vehicle age, impact direction, and vehicle body type. The dependent variable remained a categorical variable for the presence of injury, and age group remained a categorical variable to compare two age groups of older occupants to the baseline 20 to 49 age group. Delta-V and vehicle age in years at the time of the crash were continuous variables and vehicle body type and impact direction were modeled as categorical variables. Vehicle body type was a categorical variable set to zero for passenger cars, and one for all trucks, light trucks, and vans using the BODYTYPE variable. Impact direction was set to zero for frontal impacts and one for side impacts, where frontal impacts were defined by a direction of force 11 to 1 o’clock or 10/2 o’clock with general area of damage to the front. Side impacts were defined with direction of force from 2–4 o’clock and 8 to 10 o’clock. Rear impacts were excluded from the adjusted model in the odds ratio analysis, which accounts for impact direction, because of a limited number of rear impact injury cases. Cases where vehicle body type, impact direction, or Delta-V was unknown were also excluded for the adjusted odds ratio analysis.

RESULTS

Relative Injury Frequency by Age

After exclusion of 10,562 NASS/CDS occupants documented as injured but with injury details unknown, there were 131,885 NASS/CDS case occupants remaining in the dataset. Of these, 5093 NASS/CDS case occupants had at least one serious head injury, with a total of 10,242 serious head injuries. These cases represented 345,000 weighted occupants with 574,000 serious head injuries or an annual average of 23,000 occupants with serious head injuries.

Distribution of the serious head injuries in the dataset is shown by broad injury type ( Table 1 ) and by age group ( Table 2 ). Among the youngest adults included (age 20–29) there were an estimated weighted total of 179,000 serious head injuries ( Table 2 ). The total number of head injuries estimated for each age decade was progressively lower up to the age group of 60 to 69 where there was an estimated 42,000 serious head injuries. Those in the 70 to 79 age group had an estimated 56,000 serious head injuries, and those aged 80 and older had an estimated 35,000 serious head injuries.

Table 1 .

Number of raw and weighted serious head injuries (AIS 3+), by injury category

	INJURY CATEGORIES
	Intra-axial bleeding	Extra-axial bleeding	Bleeding, Location unknown	Fracture	Closed Head Injury	Swelling and other sequelae	Brain Stem Injury	Other
n (raw)	2462	3578	147	1438	1186	738	674	19
n (weighted) 1000’s	155	190	12	74	80	37	24	2

Table 2 .

Number of raw and weighted serious head injuries (AIS 3+), by age group

	AGE GROUP (YEARS)
	20–29	30–39	40–49	50–59	60–69	70–79	80+
n (raw)	3719	1995	1617	1023	754	678	456
n (weighted) 1000’s	179	101	96	66	42	56	35

Relative frequency of serious head injuries of different types was compared for age groups by decade in Figure 1 , which shows, for example, that 49% of serious injuries in the oldest age group are extra-axial bleeding and 30% are intra-axial bleeding. Percentages and standard errors are tabulated in Appendix Table A1 .

Figure 1 .

Proportion of serious head injuries of each type, for each age group. ( Not all age groups add to 100%, because “Other” injuries are not shown. ).

Injuries that accounted for a lower proportion of head injuries in the older occupants than in younger occupants included swelling and other sequelae, closed head injury and fracture ( Figure 1 ). Closed head injuries made up 17 to 18% of serious head injuries among adults under age 40, but only 3 to 9% of head injuries for those age 60 and older.

Seventeen percent of head injuries in occupants aged 20–29 were fractures. This percentage declined for every decade up to age 80 or greater, where the percentage of serious head injuries attributed to fractures was only 5%.

Bleeding injuries were more frequent, relative to other injuries, among older occupants than among younger occupants. The proportion of extra-axial bleeding among serious head injuries began increasing as early as the 30’s decade and exceeded 40% by the 50’s decade. Extra-axial bleeding injuries accounted for 49% of the serious head injuries in the oldest age group. The increase in the proportion of intra-axial bleeding with age appeared to start at an older age and was less steady, vacillating between 24% and 29% for occupants younger than 70.

When the percentages of serious head injuries in each age group that involved bleeding in any location were combined, the majority of serious head injuries were bleeding injuries in all age groups ( Figure 2 ). Bleeding injuries accounted for approximately 52% of head injuries for adults in their 20s, and a gradually increasing proportion for older occupants. In age groups age 70 or older, more than 80% of serious head injuries were bleeding injuries. While there was a steady increase in the proportion of injuries that were bleeding injuries, the steepest increases in the combined percentage of the two types of bleeding injuries occurred for occupants at approximately the 70’s decade and, to a lesser extent, in the 50’s decade, where there was also a steep increase in the percentage of extra-axial injuries. The shifts at these age thresholds were used for binning age groups in subsequent analyses in this study.

Figure 2 .

Bleeding injuries as a percentage of all serious head injuries in each age group.

No more than 1.2% of injuries in any age group were in the “other” category, indicating that the broad categories used in this analysis captured most serious injuries in the data.

Injury Rate and the Effect of Speed

Bleeding injuries, identified in relative frequency analysis as increasingly frequent with age, were further analyzed. Injury rates for bleeding injuries were calculated by age group and by Delta-V, using three age groups: adults aged 20–49, 50–69, and older adults (age 70+).

Injury rate for crashes at all speeds combined is shown in Figure 3 . For intra-axial bleeding injury, the oldest group (70+) had an injury rate more than three times that of the youngest group (20–49). The difference in injury rate by age group was even greater for extra-axial injury: the 50–69 age group had twice the injury rate of the youngest group and the oldest group had more than 5 times the injury rate of the youngest group.

Figure 3 .

Injury rate for types of bleeding injuries by age group. Standard error (SE) shown by error bars.

The total number of bleeding injuries sustained by occupants in different age groups was affected by the number of crashes that occurred in each speed range, as well as by the rate of injury in each speed range. Figure 4 shows the distribution of Delta-V for all crashes with known Delta-V in each age group. These data represent exposure to crashes and include all occupants, with or without injury. As shown in Figure 4 , the majority of weighted NASS/CDS crashes are below a Delta-V of 30 km/h for all age groups.

Figure 4 .

Distribution of Delta-V for all crashes, by age group (each age group sums to 100%).

In order to understand the effect of age at different crash severity levels, the rates of intra-axial bleeding injury and extra-axial bleeding injury by age group were sub-divided by Delta-V range as shown in Figure 5 and Figure 6 . For intra-axial bleeding injury, the most notable difference in injury rate with age was in crashes with Delta-V over 40 km/h where the oldest age group (age 70 and older) showed a rate of injury more than three times the rate for either of the two younger groups.

Figure 5 .

Injury rate for intra-axial bleeding injury by Delta-V and age group. Standard error (SE) shown by error bars.

Figure 6 .

Injury rate for extra-axial bleeding injury by Delta-V and age group. Standard error (SE) shown by error bars.

In contrast, injury rate variation by age for extra-axial bleeding injury was greatest for lower Delta-V crashes. In crashes at Delta-V 20 to 29 km/h, the extra-axial bleeding injury rate for occupants age 70 and older was estimated to be 21 times higher than the rate for occupants younger than age 50.

The combination of an elevated rate of extra-axial bleeding injury for older occupants in relatively low Delta-V crashes along with the high frequency of crashes at these lower severity levels leads to a disproportionately high number of extra-axial injuries in older occupants, as illustrated in Figure 7 and Figure 8 . Figure 7 shows age distribution of occupants in crashes of known Delta-V, compared to the age distribution of occupants with intra-axial and extra-axial bleeding injuries. Occupants in the oldest age group (70+) make up 24% of occupants with intra-axial bleeding injuries and 30% of occupants with extra-axial bleeding injuries in the included crashes, in spite of accounting for only 7% of the total number of occupants in these crashes. When narrowed to include only crashes with Delta-V less than 30 km/h ( Figure 8 ), older occupants are especially over-represented in the group of occupants who sustained extra-axial bleeding injury in low- to mid-speed crashes. Occupants in the 70+ age group accounted for 47% of the extra-axial bleeding injuries reported in these relatively low-speed crashes, while again only accounting for 7% of the crashes in this Delta-V range.

Figure 7 .

Proportion of occupants in crashes(known Delta-V) by age, compared to proportion of intra-axial and extra-axial bleeding injuries by age group

Figure 8 .

Proportion of occupants in crashes at Delta-V less than 30 km/h by age group, compared to proportion of intra-axial and extra-axial bleeding injuries by age group for same Delta-V range.

Odds Ratio Analysis

Odds ratios were calculated to compare the odds of sustaining specific extra-axial bleeding injuries for older age groups compared to the 20–49 age group. Subpial hemorrhage was excluded from odds ratio analysis because there were only three cases in the included dataset. Table 3 shows odds ratios calculated with a model that accounts only for the effects of age on injury odds, and Table 4 shows odds ratios calculated with an adjusted model that accounts for potentially confounding variables Delta-V, vehicle age, impact direction, and vehicle body type, as well as for age group.

Table 3 .

Odds ratios for extra-axial bleeding head injuries, without adjustment for possible confounding variables (Significant point estimates in bold font. Baseline young adult group age 20 to 49.)

	Point Estimate of Odds Ratio	95% Confidence Limits
Subdural hematoma
Age 50–69 compared to young adults	2.93	1.16	7.39
Age 70+ compared to young adults	6.71	4.26	10.59
Epidural hematoma
Age 50–69	0.79	0.31	1.99
Age 70+	0.40	0.11	1.48
Subarachnoid hemorrhage
Age 50–69	1.70	1.40	2.08
Age 70+	4.15	2.75	6.25

Table 4 .

Odds ratios for extra-axial bleeding head injuries, accounting for Delta-V, vehicle age, body type, and impact direction in model (Significant point estimates in bold font. Baseline young adult group age 20 to 49.)

	Point Estimate of Odds Ratio	95% Confidence Limits
Subdural hematoma
Age 50–69	3.73	1.11	12.49
Age 70+	9.38	3.74	23.48
Epidural hematoma
Age 50–69	0.81	0.25	2.66
Age 70+	0.50	0.12	2.05
Subarachnoid hemorrhage
Age 50–69	2.12	1.46	3.08
Age 70+	5.18	2.60	10.32

In this comparison of injury odds for older occupants compared to young adults, injuries that had a point estimate odds ratio greater than 1.0 and that did not include 1.0 in the Bonferroni-corrected 95% confidence interval (alpha= 0.05/6=.0083) are shown in bold font indicating significantly increased odds of injury for the older age groups. The resulting adjusted odds ratios for subdural hematoma, for example, estimates that the odds of injury for an individual in the oldest age group were 9.38 times the odds of injury for a young adult in a crash at the same Delta-V and impact direction and in the same age and body type of vehicle ( Table 4 ).

With or without adjustment for possible confounding variables, both the 50 to 69 and the 70 and older age groups showed increased odds of subdural hematoma and of subarachnoid hemorrhage. Neither age group showed increased odds of epidural hematoma over younger occupants, either with or without accounting for possible confounding variables.

DISCUSSION

Bleeding injuries accounted for at least half of serious head injuries reported among adult occupants in all age groups, and made up an increasing proportion of head injuries among older occupants. More than 80% of serious head injuries among those aged 70 and over were bleeding injuries. While the oldest occupants (age 70 and older) have the highest rates of intra-axial and extra-axial bleeding injuries compared to younger occupants, the age-related increase in extra-axial bleeding injuries was substantial even for occupants as young as those in their 50s and 60s.

The effect of age on rate of extra-axial injury was greatest in lower Delta-V crashes ( Figure 6 ), while the effect of age on rate of intra-axial injury was predominant in high Delta-V crashes ( Figure 5 ). Given that less than 15% of crashes in any age group occurred at over 30 km/h ( Figure 4 ), the elevated intra-axial injury rate in high Delta-V crashes would not be expected to result in a substantial increase in injuries to older occupants. However, the elevated extra-axial injury rate for older occupants in lower Delta-V crashes is especially important because this speed range is where most crashes occur ( Figure 4 ).

The combination of an elevated rate of extra-axial bleeding injury for older occupants in relatively low Delta-V crashes along with the high frequency of crashes at these lower severity levels led to a disproportionately high number of extra-axial injuries to older occupants. Occupants in the oldest age group (70+) make up 47% of those with extra-axial bleeding injuries reported in crashes with Delta-V less than 30 km/h, while accounting for only 7% of the occupants in crashes in this Delta-V range.

With or without adjustment for possible confounding variables, both the 50 to 69 age group and the 70 and older group showed increased odds of subdural hematoma and subarachnoid hemorrhage relative to the youngest adult age group. Conversely, the odds of epidural hematoma showed no increase for the older age groups, indicating that the age-related increases in extra-axial bleeding injuries identified in this study were driven largely by subdural and subarachnoid bleeding.

To assess the relative importance of subdural hematoma and subarachnoid hemorrhage, these injuries were evaluated in the context of outcome, in particular for older patients.

Although the subdural mortality rate has been reported to be as high as 50% or higher in comatose patients or surgery cases [ Gennarelli et al., 1982 , Haselsberger et al., 1988 , Marshall et al., 1991 , Seelig et al., 1981 , Servadei, 1997 ], Perel reported an overall subdural hematoma mortality rate of 33% [ Perel et al., 2009 ], which is a more appropriate estimate to compare to the NASS/CDS cases in the current study which includes non-surgical and non-comatose cases. After adjusting for the contribution of other injuries, Perel reported the odds of mortality from large subdural hematoma was greater than for other space-occupying bleeding injuries, which is consistent with reports that subdural hematoma is especially lethal among bleeding head injuries [ Marshall et al., 1991 , Seelig et al., 1981 ] and responsible for as many as 43.5% of head injury deaths [ Gennarelli et al., 1982 ]. Subdural hematomas are particularly lethal for older patients [ Hanif et al., 2009 ]. Age 54 has been identified as the age threshold for increased mortality for subdural hematoma [ Stitzel et al., 2008 ]. Combining the increased incidence of subdural hematoma for occupants aged 50 and older from the current study with the previously-reported increased mortality rate for subdural hematoma over age 54 presents a double risk for older occupants.

Subarachnoid hemorrhage, also identified in the current study as more frequent among older occupants, has been linked to increased risk of poor outcome after head injury [ Chieregato, Fainardi, Morselli-Labate et al., 2005 , Greene, Marciano, Johnson et al., 1995 , Mattioli, Beretta, Gerevini et al., 2003 , Servadei, Murray, Teasdale et al., 2002 , Vollmer et al., 1991 ], but this increased risk is not related to subarachnoid hemorrhage as a primary injury. Instead, the relationship between subarachnoid hemorrhage and poor outcome has been attributed to subarachnoid hemorrhage as an indicator of more severe underlying mechanical injury or on the negative effect subarachnoid hemorrhage can have on the resolution of primary head injuries. While subarachnoid hemorrhage may not be as important as subdural hematoma as a primary cause of mortality, it may be aggravating injury outcomes.

Further exploration of age-related changes in tolerance to subdural hematoma and subarachnoid hemorrhage is needed to improve our understanding of the mechanisms for these injuries. In turn, understanding why older occupants have reduced tolerance to these injuries can be used to define age-adjusted injury thresholds for head injury and, ultimately, to develop countermeasures for older occupants.

CONCLUSION

This study supports previous observations that older head injury victims in motor vehicle collisions are more likely to sustain bleeding injuries than younger head injury victims.

In particular, the rate of extra-axial injury appeared to increase with age, and was substantial for occupants as young as those in their 50’s. The increase in extra-axial injury rate with age was especially prominent in relatively low Delta-V crashes, which represent the majority of all crashes.

Subdural hematoma is of primary importance among the extra-axial bleeding injuries that had increased odds of injury for older occupants because of the combination of increased incidence and increased mortality for those as young as their 50s.

Subarachnoid hemorrhage, which can potentially contribute to poorer injury outcomes, also showed significantly increased odds of injury for older occupants.

Exploration of the mechanisms of subdural hematoma and subarachnoid hemorrhage in the older occupant is needed to define age-adjusted injury tolerance and to develop head injury countermeasures for older occupants.

APPENDIX.

Table A1 .

Proportion of serious head injuries (AIS 3+) in each injury category, by age group. Standard error (SE) reported in brackets.

	AGE GROUP (YEARS)
	20–29	30–39	40–49	50–59	60–69	70–79	80+
n (raw)	3719	1995	1617	1023	754	678	456
n (weighted) 1000’s	179	101	96	66	42	56	35
Brain stem injuries	4.0 (1.0)	5.3 (0.9)	3.9 (0.9)	5.2 (0.8)	4.0 (1.6)	2.1 (0.3)	2.7 (1.4)
Closed head injury	18.1 (2.7)	17.1 (4.3)	13.7 (1.6)	14.7 (4.3)	8.6 (1.8)	3.4 (0.5)	6.3 (4.1)
Fractures	17.2 (2.9)	14.6 (3.1)	13.6 (1.1)	9.5 (1.6)	8.9 (2.6)	5.8 (2.6)	5.3 (2.7)
Extra-axial bleeding	25.3 (1.2)	27.1 (2.8)	33.5 (1.6)	41.4 (1.1)	40.3 (3.1)	43.4 (7.6)	49.3 (2.8)
Intra-axial bleeding	23.7 (1.0)	29.0 (6.4)	24.9 (2.7)	24.3 (1.0)	27.7 (2.9)	38.8 (9.3)	30.4 (3.8)
Bleeding, Location unknown	3.3 (1.5)	1.3 (0.2)	1.7 (0.5)	1.1 (0.3)	3.7 (1.7)	1.1 (0.5)	2.2 (1.4)
Swelling and other sequelae	8.2 (0.6)	5.4 (0.8)	8.6 (1.8)	3.5 (1.3)	6.9 (1.9)	4.3 (1.2)	3.8 (1.6)
Other	0.2 (0.1)	0.1 (0.1)	0.2 (0.1)	0.3 (0.2)	0	1.2 (1.2)	0
Total *	100.0%	99.9%	100.1%	100.0%	100.1%	100.1%	100.0%

^* Not all age groups sum to 100.0% due to rounding