ACCESS BOARD RESEARCH

A Review of Technical Requirements for Ramps

Final Report 

January 31, 1996

Jon A. Sanford, Project Director
The Center for Universal Design
North Carolina State University
School of Design
Box 8613
Raleigh, NC 27695-8613

Prepared for:
US. Access Board
Contract Number QA930020

Contents

REVIEW OF CURRENT LITERATURE AND STANDARDS
A STUDY OF THE POPULATION OF PEOPLE WITH MOBILITY IMPAIRMENTS
HUMAN SUBJECTS TESTING
RECOMMENDATIONS
REFERENCES AND SELECTED READINGS

REVIEW OF CURRENT LITERATURE AND STANDARDS

Background

With the advent of various legislative acts such as the Rehabilitation Act of 1973, the Fair Housing Amendments Act (1988), and the Americans with Disabilities Act (1990), it is anticipated that there will be an increased reliance on ramps as a means of entrance and egress. At the same time, the validity of the existing code and standard requirements for ramps has been questioned with increasing frequency. In fact, in a recent study of fatal wheelchair accidents in the United States, 8% of accidents caused by environmental factors occurred on ramps (Calder & Kirby, 1990). With the significant change in demographics over the past two decades and the projected increase in the number of older people and people with disabilities, a reevaluation of the current ADA requirements for the design of ramps for their usability by current and anticipated populations seems appropriate.

Reevaluating Existing Requirements

The initial step in the reevaluation of existing requirements for ramps is an analysis of other codes and standards as well as state-of-the-art literature and research studies. This analysis covers the specific sections of ADAAG which address accessibility in ramps including: Sections 3.5 Definitions and 4.8.1 General; Section 4.8.2 Slope and Rise; Section 4.1.6(3)(a) and (b) Special Technical Provisions for Alterations to Existing Buildings and Facilities; Section 4.1.7 (3)(a) Provisions for Minimum Requirements for Historic Preservation; 4.8.3 Clear Width; 4.84 Landings; 4.8.5 Handrails; 4.8.6 Cross Slope and Surfaces; and 4.8.7 Edge Protection; and 4.8.8 Outdoor Conditions. A final section, Other Considerations, has also been included to accommodate items that are not covered in ADAAG. Finally, the analysis compares the provisions in ADAAG to: the federal standards (UFAS); the four model codes (UBC, NBC, SBC, and BOCA); and several versions of the model standard (ANSI A117.1 1980, 1986, 1992). The provisions in ADAAG are also compared with relevant state codes.

Definitions. The one criterion that comes as close to a nationally accepted standard is the definition of a ramp. Virtually all codes and standards define a ramp as having slope steeper than 1:20 except Washington which defines a ramp as any walking surface with a slope steeper than 1:48. (Although New Jersey specifically defines a ramp as having a slope equal to or greater than 1:20, this interpretation is implicit in all codes and standards.) Clearly, a ramp is walking surface with a slope greater than (and including) 1:20; whereas a walking surface with a slope less than 1:20 is a simply a pathway. In addition, ADAAG defines the term running slope as the slope that is parallel to the direction of travel. This clearly distinguishes between the slope of the ramp and cross slope (slope that is perpendicular to the path of travel).

Slope and Rise. ADAAG, ANSI, and UFAS all permit a maximum rise of 30 inches for each ramp run. The three differ, however, in their treatment of slope, rise, and run as variables. Both ADAAG and UFAS assume that only run is a variable, whereas ANSI suggests that slope and run are both variables, with restrictions in one often compensated by the other. As a result, ADAAG and UFAS allow a maximum run of 40 feet for slopes between 1:16 and 1:20, ANSI permits ramps with a slope of 1:20 to have a maximum run of 50 feet. The proposed Massachusetts' state code is also noteworthy in that it specifically defines slope as "measured between any two points on the ramp." This provision ensures that the slope does not exceed 1:12 for short distances on the ramp run, even though the total slope of the run may be compliant. Despite the general concurrence among the codes related to slope, rise, and run, research suggests that the maximum rise of 30 inches between landings (which would result in a 30 foot ramp length at the recommended 1:12 gradient) may be excessive. However, because prior studies have been conducted with different populations, consensus is lacking regarding the optimum ramp run.

Alterations. The model codes and standards that include provisions for alterations, provide a lower standard for ramps that are constructed in existing facilities, where conditions may prevent the use of 1:12 slope or less. ADAAG, UFAS, and ANSI permit a maximum rise of 6 inches for slopes between 1: 10 -1:12 and a maximum rise of 3 inches for slopes between 1:10 -1:8. However, UFAS and ANSI '80 and '86 limit the rise at the shallower slopes by limiting the total runs for the prescribed slopes at 5 feet and 2 feet, respectively. Only two states, Washington and Wisconsin have provisions for ramps in alterations. The Washington code is less stringent than the federal standards and model codes, permitting a rise of up to 6 inches for slopes between 1:8 and 1:10 and a maximum of 3 inches for slopes steeper than 1:8. In contrast, Wisconsin's code seems somewhat unrealistic, permitting a slope up to 1:8 for a maximum rise of 2 feet.

Historic Preservation. ADAAG and UFAS are the only standards or codes that provide an exception for ramp slope in historic preservation. The exception permits a maximum slope of 1:6 for a run up to 2 feet, although the rise is not explicitly limited. If it is assumed that people can manage a ramp with a steep slopes when the ramp is short (i.e., two feet), then a maximum rise is not necessary. In contrast, the Appendix material in ANSI 1992 suggests a slope up to 1:9 for a maximum rise of 16 inches and a maximum run of 12 feet; and a slope up to 1:6 for a maximum rise of 4 inches and a maximum run of 2 feet.

Clear Width. Although federal standards and model codes are uniform in requiring ramps to be at least 36 inches wide, none specify how the clear width is to be measured. ADAAG, in the definitions section, describes "clear" as being "unobstructed," although unobstructed may vary between people who are ambulatory and those who use wheelchairs. The proposed Massachusetts code, the New Jersey code and the Wisconsin code for interior ramps are the only codes that state how the requirement for clear width is to be determined, by either measuring between the railings, or measuring the width exclusive of edge protection or flared sides.

The 36 inch dimension is consistent with the minimum required width of an accessible route. However, a number of states including North Carolina, Washington, and Wisconsin require access ramps to be between 44 to 48 inches in order to be consistent with egress ramp requirements in the model building codes for buildings with occupancies of 50 (or more) persons. Alternatively, Kentucky requires ramp width to be determined according to the model code requirement for occupant load.

Although the 44 to 48 inch dimension would permit an ambulatory person and a person using a wheelchair to pass each other, prior research by the Center for Accessible Housing indicates that in facilities in which there are a number of people who use wheelchairs (i.e., more than one or two), wider ramps may be appropriate to accommodate two-way wheelchair traffic on the ramp.

Landings. The federal standards and model codes as well as the New Jersey state code require the width of landings on a straight run ramp to be as wide as the ramp run leading to it (a minimum of 36 inches), or as wide as the widest ramp run leading to it. Where a ramp changes direction, most standards and codes require a 60 inch by 60 inch landing. The Texas Code does not require a level landing when changes in direction are less than 45 degrees and are smoothly blended with intersecting surfaces.

The few sources in the literature that covered landing dimensions recommended landings that exceed those required by the any of the codes and standards. Templer, et al. (1982) recommended that a minimum landing distance of 66 inches at the top and bottom of a ramp was necessary to allow adequate stopping distance for 85% of those descending the ramp and for 91% of those ascending the ramp. Travers (1991) recommended that rest platforms at least 71 inches x 71 inches be located at the top of a ramp.

Handrails. All of the codes and standards require handrails on ramps that have a slope greater than 1:20 except Kentucky which only requires handrail on ramps with slopes less than 1:15. The location for handrails is generally along both sides of ramp segments, except in several states (i.e., North Carolina, Wisconsin and Colorado) which consider handrails as a function of ramp slope and the slope or drop off of adjacent terrain in determining railing locations. Travers (1991) recommends that handrails should be provided at least on the outer edge of the ramp, and preferably on both sides, in order to allow ambulant people with mobility limitations to negotiate the ramp.

The height of handrail requirements varies from 34 to 38 inches (ADAAG, ANSI '92, UFAS, UBC), to 30 to 34 inches (ANSI '80 and '86, North Carolina, Wisconsin), 32 to 34 inches (New Jersey), and 32 inches (Colorado). ADAAG, UFAS and ANSI, in their appendices, suggest the use of a second handrail at a lower height for facilities where children are the primary users. However, they do not suggest a height at which this handrail may be placed. The Massachusetts code (proposed and existing) requires handrails to be provided in pairs, at 34 inches and 19 inches height. This lower rail enables use by children, and also serves as guardrail or edge protection where there is no edge curb or wall.

Cross Slope and Surfaces. Maximum cross-slope varies between 1:48 and 1:50, although several codes do not have a provision for cross-slope along a ramp. The cross-slope of 1:50 is utilized by ADAAG, ANSI '80 and '86, Massachusetts and Texas, while ANSI '92, UBC, North Carolina, and New Jersey specify 1/4 inch in 1 foot width (1:48). In terms of usability, the difference between a cross-slope of 1:48 and 1:50 is probably negligible. The only significant difference between the two slopes appears to be that of the scale used. A 1:48 slope is clearly compatible with the English Standard System, being equivalent to a cross slope of 1/4 inch in 1 foot or 1 inch in 4 feet; whereas a 1:50 slope clearly compatible with the Metric System. Several articles addressed the issue of cross slope, although none evaluated cross slopes that were consistent with the code requirements. For example, Brubaker et al. (1986) found that a 2o (1:29) side slope adversely affected a wheelchair user's ability to traverse a ramp. A second source, Travers (1991) noted that the cross fall should not exceed 1:100 in order to allow the wheelchair user to maintain control.

The codes and standards are fairly uniform in their specification for "stable, firm and slip-resistant" surfaces. ADAAG is the most explicit, as it is the only document that includes a static coefficient of friction of 0.8 for ramps. In addition, the federal standards and model codes, as well as some of the state codes include specifications for carpeted ramp surfaces. The findings of several studies that evaluated ramp surfaces, clearly indicate that ramp surfaces can affect performance. Walter (1971) found that the degree of friction between the ramp and the wheelchair tires significantly affected the amount of effort required to ascend the ramps. A second study, Sweeney et al. (1989a) found that appropriate surface finishes are dependent on the type of user. They found that self-propelling wheelchair users preferred a rough surface in order to maximize traction and facilitate safety on the ramp. In contrast, attendants who propelled wheelchairs preferred a smoother surface in order to facilitate wheelchair movement.

Edge protection. With few exceptions (i.e., Massachusetts, North Carolina, Kentucky, Wisconsin and Colorado) all codes and standards require ramps with drop-offs to have curbs, walls, railings or projecting surfaces that prevent people from slipping off the ramp. Although several types of edge protection are possible, only curbs have a specified minimum height. Consideration should be given to specifying maximum and/or minimum heights of wall, railings or projecting surfaces. One possibility is the Massachusetts requirement for a maximum height lower rail of 19 inches, which is intended to be used by children and to provide edge protection. Finally, although the four inch minimum height for curbs specified by ANSI 1992 and UBC is more stringent than the two inch height required by ADAAG, there is no research evidence to verify that four inches would provide greater protection.

Outdoor Conditions. One study that evaluated the use of ramps in adverse conditions found that unprotected ramps became impassable with snow and ice, and that the use of salt helped to contribute to slippery conditions. Although the federal standards, model codes, and several state codes require that approaches to outdoor ramps and ramps themselves be designed so as to prevent water accumulation on their walking surfaces, none of the codes (except New Jersey which permits the use of grates and Minnesota which requires protective structures) suggest potential design solutions. Moreover the code provisions do not necessarily address problems caused by freezing temperatures. If ice and snow are considered to be forms of water, then the current ADAAG provision would also cover these conditions. However, this seems to be stretching the interpretation of the current provision. Minnesota is the only state that addresses this problem, requiring that exterior ramps be protected from external climatic conditions in order to insure that an access ramp would remain usable even in severe winter conditions.

Circular Ramps. Massachusetts has the only state code that prohibits the use of circular ramps as access ramps. Similarly, ANSI 1992 has related appendix information in which circular ramps are permitted if it is determined through engineering analyses that the slope of a circular ramp does not exceed 1:12 anywhere on the ramp surface. ANSI's prohibition of circular ramps is based on the premise that the inability to have all four wheels of a wheelchair simultaneously at rest creates a dangerous situation. This analysis seems reasonable, however, further research is necessary to substantiate such allegations.

Conclusions

The literature review revealed significant and interesting findings about the populations previously tested and methodologies used. In addition, despite the general agreement among the major codes and standards, as well as the state codes, there are a number of areas in which unique provisions in some of the state and model codes as well as findings from research studies that indicate that changes to ADAAG may be warranted.

First, studies of ramp performance were evaluated in terms of the degree to which they have addressed ramp negotiation with a wide variety of mobility impairments over a wide age range. Most studies, however, used fairly homogeneous populations both in terms of age and ability. As a consequence, recommended gradients were found to vary widely, and seem to be associated with several factors related to the population evaluated. These factors include degree of mobility (i.e., nonambulatory individuals who use self-propelled, attendant-propelled, or powered wheelchairs or ambulatory individuals who walk with difficulty and/or use assistive walking devices) and amount of stamina. For example, both Elmer (1957) and Canale, Felici, Marchetti & Ricci (1991) recommended fairly steep ramp slopes (up to 1:6 and 1:6.7, respectively) based on younger samples with good upper body strength who were able to propel their own wheelchairs fairly easily. Steinfeld, Shroeder & Bishop (1979), on the other hand, concluded that ramp slope should vary between 1:16 and 1:20 based on a sample of individuals with more restricted functional abilities and a wider range of ages.

Second, performance variables and procedures used to measure them varied across studies. Studies have included a number of objective performance measures such as time required to traverse a ramp, maximum distance attained, distance between stops, number of strokes made on the wheelchair rim (for independent wheelchair users), and number of steps taken (for attendant-propelled wheelchair users and those using assistive walking devices) as well as subjective measures such as asking participants to stop when they needed to rest (e.g., Templer, et al., 1982). Interestingly, however, most studies that have evaluated energy expenditure of people who use wheelchairs and assistive devices (which could be a critical factor in an individual's ability to traverse a ramp) have done so on level surfaces.

Finally, based on this analysis, a number of changes to ADAAG should be considered. These include: ensuring that slope does not exceed 1:12 over short distances of a ramp; reducing the maximum rise in new construction and alterations; increasing the clear width of ramps to meet provisions for egress, thus permitting ramps to function as an accessible means of egress and by multiple individuals in wheelchairs; increasing the size of landings; clarifying the provision for height of handrails; reducing the maximum permissible cross slope; specifying maximum heights for walls, railing, or projecting surfaces used as guardrails; and protecting exterior ramps in cold climates.

A STUDY OF THE POPULATION OF PEOPLE WITH MOBILITY IMPAIRMENTS

Prevalence of impairment

There is no single data source that directly assesses the prevalence of mobility impairments in the US population. Mobility impairments may have many etiologies. And, with few exceptions (e.g., cervical spinal cord injuries), all individuals represented in a diagnostic group may not have mobility impairments. Fortunately, several sources define disability in terms of activity or functional limitation. By examining these sources carefully, the proportion of activity or functional limitations associated with mobility impairments can be estimated.

One source of data, the National Center for Health Statistics (NCHS), uses the concept of activity limitation to measure disability. NCHS defines activity limitation as any long-term reduction in activity resulting from chronic disease or impairment (Ficke, 1992). In the annual National Health Interview Survey (NHIS) conducted by NCHS, activity limitations are defined in relation to the typical major activities for one's age group: for example, ordinary play for children under age five, attending school for school-aged children, working or keeping house for adults 18-69, and capacity for independent living for persons over the age of 69.

Based on the NHIS, the NCHS estimates that activity limitations are present in approximately 14% of the current US population, including 4.5% with limitations in non-major activities. Estimates from the 1988, 1989, and 1990 NHIS place the number of non-institutionalized people in the US with activity limitations between 33 and 34.5 million. Pope and Tarlov (1991) estimated that 38% or 12.9 million people who report activity limitations on the NHIS due to a mobility limitation.

In contrast to the NCHS, the US Bureau of the Census uses the concept of functional limitation to identify disability in the population of non-institutionalized persons aged 15 years and older. The Bureau of the Census assesses functional limitation in the periodic Survey of Income and Program Participation (SIPP). Functional limitation is defined in terms of the individual's ability to perform nine sensory and physical activities. From the SIPP of US households, the Bureau of the Census estimates that 20.6% of the population over age 15 (37.3 million people) have a functional limitation as defined by reported difficulty in one or more activities. This estimate includes 7.5% of the population with severe functional limitations. Based on the 1986 SIPP, Ficke (1992) estimated that 2.8% of the non-institutionalized population (5.2 million people) aged 15 years and older report "difficulty getting around outside."

A third alternative, and perhaps the most accurate estimate of the prevalence of mobility impairment in the US is the use of assistive devices, derived from the 1990 Assistive Devices Supplement to the National Health Interview Survey. The 1990 Supplement includes data from interviews with a representative sample of 117,042 individuals living in the US. A total of 3,158 respondents (2.7%) report using assistive devices for getting around. This finding suggests that approximately 6.6 million non-institutionalized persons in the US use assistive devices for getting around. These estimates are corroborated by those reported by LaPlante, Hendershot, and Moss (1992) that 2.6% of the non-institutionalized population (or about 6.4 million people) use assistive devices for mobility.

In addition to questions about use of assistive devices, the 1990 Supplement asked respondents if their home was equipped with any special features (e.g., ramp, wide doors, grab bars) to accommodate functional impairments. A total of 4,787 respondents (or 4%) were identified who met one of the following conditions: 1) reported use of a crutch, cane, walker, manual wheelchair, electric wheelchair, scooter, leg or foot braces, an artificial leg or foot, or 2) reported having significant activity limitations, not using an assistive device but having a ramp as an accessibility accommodation in their home. Extrapolating from this analysis, an estimated 9.99 million non-institutionalized persons in the US have significant mobility impairments requiring use of assistive devices or special features in the home. It is also important to note that the majority of individuals who use assistive devices for mobility are older people (73% are people 55 years of age or older). Less than 4% of people reporting mobility impairments are 16 or younger.

Trends impacting prevalence of impairment

The prevalence of disability among the US population appears to be increasing steadily. Zola (1993) and others have concluded that two major trends account for the rise of disability in the US: 1) decreasing mortality rates for a variety of disabling illnesses and injuries, resulting in an increase in the length of time that people live with disabilities and 2) the undeniable aging of the US population.

Mortality and Incidence of Specific Conditions. Several trends are noteworthy with respect to current and projected increases in survival rates for infants, younger people, and older people with disabilities. These trends are influencing the projected incidence rates of congenital disabilities, disabilities that result from traumatic injury, disease functions leading to disability, and comorbidity factors of aging.

Congenital Disabilities. For years infant mortality rates have decreased steadily in large part due to improved living standards, prenatal care, and neonatal medicine. An increasing number of low birth weight (<2500 gm.) and very low birth weight (<1500 gm.) infants are surviving into childhood and beyond, often with manifest chronic conditions, which have far reaching implications for the number of people with mobility impairments.

The most common congenital condition is cerebral palsy (CP), which is most often manifest in hemiplegia (31%), diplegia (22%), and quadriplegia (35%) (Pharoah, Cooke, Cooke, and Rosenbloom, 1990). Most epidemiological studies from industrialized countries report a rise in the prevalence of CP among children in recent decades, largely due to the increasing survival rate among low birth weight (LBW) and very low birth weight (VLBW) infants.

Based on data from the NHIS, several studies have estimated in incidence of CP to be between 0.2% - 0.23% (Alday, 1992; Boyle, Decoufle, and Yeargin-Allsopp, 1994). In a review of recent secular trends in the prevalence of CP in industrialized countries that have population-based CP registries, Bhushan, Paneth, and Kiely (1993) concluded that the only demographic determinant of CP prevalence was increased survival rate of LBW and VLBW infants. As a result, Sola and Piecuch (1994) predicted that the incidence of CP among newborns would be expected to increase about 5% between 1986 and 2010, resulting in an estimated incidence rate of 2.4/1000 births by the year 2010.

A number of other studies (Escobar, Littenberg and Petitti , 1991; Grogaard, Lindstrom, Parker, Culley, and Stahlman, 1991; Nicholson and Alberman, 12992; Pharoah, Cooke, Cooke, and Rosenbloom, 1990; Rosen and Dickinson,1992), have documented the significant increase in prevalence of CP among LBW and VLBW infants. Moreover, in a meta-analysis of VLBW infants, Escobar, Littenberg and Petitti (1991) not only discovered a median incidence rate of CP of 7.7%, but found that VLBW infants had a median incidence rate of 25% for any type of chronic disability. Thus, despite the improved survival rate of VLBW infants, poor outcomes among survivors are common.

Disability from Trauma. Trauma remains one of the leading causes of death and disability in the US, accounting for over 142,000 fatalities annually. Although the mortality rate of serious injury is on the decline, decreasing from 45.4 to 35.8 deaths per 100,000 injuries from 1975 to 1988 (National Safety Council, 1989), 2.3 million people in the US suffer traumatic injuries annually.

Spinal cord injury. Few injuries result in more profound and long-term disability than traumatic spinal cord injury (SCI). Each year an estimated 10-20,000 people in the US sustain a SCI. Kraus (1985) reports a prevalence rate of approximately 200,000 people in any given year. As recently as the 1950s, death was likely in the very early stages of SCI or soon after because of respiratory and other complications. In World War I, for example, only 400 men with wounds that paralyzed them from the waist down survived at all, and 90% of them died before they reached home. In World War II, 2,000 men with paraplegia lived and 1,700 were still alive in the late 1960s (President's Committee on Employment of the Handicapped, 1967). Each decade since has seen a rapid decline in the mortality rate and corresponding increase in long-term survival of persons with paraplegia and, more recently, persons with quadriplegia. Yet, despite an increase in the number of persons who have survived SCI with increasingly severe disabilities, SCI has had a marked impact on their general functioning and longevity.

For example, Gerhart, Bergstrom, Charlifue, Menter, and Whiteneck (1993) found, in an assessment of functional changes of 279 individuals who had sustained spinal cord injuries 20 to 47 years prior to the study, that the functional abilities of a substantial number of people with SCI declines significantly over time. This included additional help with transfers, mobility, dressing, and toileting. In another study of 9,135 persons with SCI, DeVivo, Black, and Stover (1993) found that mortality rates for people with SCI due to unrelated diseases were significantly higher than those for the general population. Moreover, they found that spinal cord injured persons were 82.2 times more likely to die of septicemia, 46.9 times more likely to die of pulmonary emboli, and 37.1 times more likely to dies of pneumonia than the general population. The researchers concluded that although some cause-specific mortality rates for SCI persons have declined dramatically, many remain substantially above normal, thus the life expectancy of people with SCI is still less than the population as a whole.

Traumatic brain injury. Traumatic brain injuries (TBI) represent another significant consequence of injury in the US. Published incidence rates range from 180 to 367 cases per 100,000. The US Committee on the National Agenda for the Prevention of Disabilities (Pope and Tarlov, 1991) estimates that 1.3 million adults sustain head injuries annually, with 70-90,000 individuals sustaining moderate to severe TBI. In addition, Moscato, Trevison, and Willer (1994) found that among adults experiencing TBI, 84% reported co-occurring disabilities, the most prevalent being limited mobility and agility (present in 69% of all cases).

TBI is also the most common cause of acquired disability in childhood, accounting for approximately one-half of all traumatic injuries among children. Approximately 5 million cases occur annually and 200,000 children are hospitalized with TBI each year. At least 15,000 children annually require prolonged hospitalization (di Scala, Osberg, Gans, Chin, and Grant, 1991). Among children, Michaud, Duhaime and Batshaw (1993) reported an average incidence rate of 200/100,000 children per year in the US, including approximately 20% that suffer from long-term disability.

Like SCI, the survival rate among those sustaining severe TBI has increased steadily over the past 15 years. This increase is attributable, in large part, to improved availability of trauma care and emergency medical services throughout the US (Pope and Tarlov, 1991). Despite the decrease in mortality, there has been a concomitant increase in the number of people who suffer from disability.

Causes of trauma. Motor vehicle accidents are the leading cause of all nonfatal TBI and SCI, followed by falls, assaults, and sports or recreational injuries (Pope and Tarlov, 1991). Despite the continuous reduction in fatality risk since the 1930s, crash injury is still a leading cause of death and disability (Viano, 1992; CDC, 1993). Automobile collisions account for 48% of SCI (Price, Makintubee, Herndon, and Istre, 1994) and one-third to one-half of all TBI (Mucci, Eriksen, Crist, Bernath and Chaudhuri, 1991). Numerous interventions have been shown to be effective in reducing the incidence and severity of traumatic injuries. Two primary strategies associated with reducing injury and disability are injury prevention and accident avoidance.

The most important factor in preventing injury is the vehicle. A structure with controlled crush of the front-end and protective cage around the occupants is a key part of the safety system that minimizes injury risks. Use of occupant restraints is the second feature in preventing injury. As parts of the restraint system, lap-shoulder belt use provides a 42% reduction in fatality risk (Evans, 1988; 1990), while a driver airbag provides an 18% reduction.

Recent legislation mandating the use of restraint devices for front seat occupants in automobiles has decreased the incidence and severity of injury following automobile crashes. Between 1977 and 1985 all 50 states enacted legislation requiring the use of child safety seats or safety belts for children. Although these laws can be credited with reducing the number of crash-related injuries by an estimated 8% - 59% (depending on the state), motor-vehicle crash-related injuries remain a major cause of disability and death among US. children (CDC, 1993), primarily due to the continued lack of seat belt use. In a study conducted in West Virginia, Sokolosky, Prescott, Collins, and Timberlake (1993) found that among accident victims who required hospitalization, unbelted patients had a 34% higher injury severity score, a 97% increase in the need for extended care after discharge and a 186% increase in hospital charges than belted patients.

Although more can be done to improve vehicle crash worthiness, the cost/benefits in safety are leveling off. As a result, in the future, more emphasis will be put on vehicle technologies to avoid accidents, such as anti lock brakes, obstacle detection, and radar warning (Viano, 1992). But despite the increasing availability of safety features such as anti-lock brake systems, early indications suggest these features have had little negligible impact on accident avoidance (National Public Radio, 1994).

Disease Functions Leading to Disability. Changes in the mortality or incidence rate of several diseases associated with disability have been documented in recent years. These diseases include multiple sclerosis, Hodgkin disease, and terminal illnesses such as leukemia and cystic fibrosis (Zola, 1994). The most significant of these trends associated with mobility impairment and not otherwise associated with aging, is the increased prevalence of multiple sclerosis (MS). Kurland (1994) estimates that approximately 10,000 people develop MS in the US each year and the annual prevalence rate is estimated at 250,000 cases. Kurtzke (1991) presents evidence from a number of epidemiological studies to conclude that mortality rates have declined and incidence rates have increased sporadically and in certain high-incidence, geographically-specific locations. This data led Kurtzke to conclude that recent evidence suggests MS is a disease much more in flux than might be expected and, therefore, of uncertain future prevalence.

Age-Related Disability. All population data affirm that the fastest growing segment of the US population is made up of those over the age of 65. Throughout most of history, only one in ten people lived past 65; now nearly 80% do (Zola, 1990). However, dramatic declines in age-specific mortality, particularly due to stroke and heart disease, has resulted in large increases in life expectancy. Between 1950 and 1986 mortality rates for those 65 to 74 years old declined 31%, and the decline for those over 85 years old was almost 24%. Within this period of time (1960-1985) the population aged 65 and older increased by 71%, almost three times faster than the rate of increase for those under 65. When only those over 85 years are considered, the 199% increase in population is almost 6 times the rate of growth for the population as a whole (NCHS, 1988).

While the contribution of declining mortality to the unprecedented growth of the 65+ and 85+ populations is easily characterized, three issues that impact the levels of health and functioning of those who survive are of interest here: 1) general health and functional abilities; 2) portion of the population that will be residing in the community; and 3) impact of specific disorders associated with aging.

General Health and Functioning. Many studies have demonstrated that longer life spent with medically managed chronic conditions implies more years of disability for individuals (Chirikos, 1986; Colvez and Blanchet, 1981; Rice and LaPlante, 1988; Roger, Rogers, and Belanger, 1989; Verbrugge, 1984). One study by Mahoney, Estes, and Heumann (1986) demonstrated that the incidence of chronic conditions and the need for assistance rises with age. By age 75 and over, 16% had three or more chronic conditions causing activity limitations. In 1981, 46% had arthritis, 38% had hypertension, and 28% had heart conditions. Many elders also suffered from various impairments including hearing (28%), visual (14%), and orthopedic (13%). In addition, at age 65-74 only 5% need help with basic physical activities such as walking, bathing, etc. By age 85+, 35% need assistance. A second study by Guralnick (1985) examined the ability of variables measured in 1965 to predict physical functioning in 1984 in members of the Alameda County Study who were then 65 to 89 years old. He concluded that "trouble dressing, feeding, or moving," and "trouble climbing stairs and getting outdoors" were strongly associated with chronic health conditions.

Other studies that have analyzed data from national data sets have found contradictory evidence suggesting that older people are healthier than ever before (e.g., Manton, Stallard, and Woodbury, 1991; Miller, Prohaska, Mermelstein, and Van Nostrand, 1993). One study of National Long Term Care Survey data between 1982 and 1989 (Manton, Stallard, and Woodbury, 1991), found that the elderly population had improved health and functioning based on data that the elderly impaired population grew at a slower pace (10%) than the elderly population in general (15%). Moreover, mortality declined for older people without disabilities and increased for those with severe disabilities, thus resulting in an older population with fewer disabilities.

However, despite a decline in the rate of people with impairment, the precipitous growth of the elderly population has still resulted in large increases in the numbers of disabled elderly. Moreover, even with an improvement in functional ability for the older population as a whole, there is a large absolute growth in the number of people who are highly impaired among those in the oldest old (85+) category.

Community Dwelling Elders. Prior to the early 1980s, care for older, chronically ill, disabled individuals occurred in nursing homes. Age-related decrements were viewed as natural concomitants of aging (Brody and Schneider, 1986) and little effort was directed at preventing chronic disease or disability. However, trends in controlling risk factors associated with age-related infirmities have resulted in an increasingly older, increasingly frail population. Moreover, the rapid acceleration in numbers of older individuals has far outpaced the numbers of beds available in long term care institutions. As a result, there are growing numbers of older individuals with chronic conditions living in the community.

Impact of Specific Disorders. With increased age comes increased likelihood of age-related disabling conditions. Most prominent among these are arthritis and other musculoskeletal disorders, osteoporosis and hip fracture, stroke, and diabetes mellitus resulting in lower extremity amputation.

Arthritis. Thirty-seven million people in the United States suffer from arthritis (Abyad and Boyer, 1992). It is the leading cause of work-related disability and the leading cause of disability among persons 65+ years in the United States, although it rarely causes death (Verbrugge, 1987; Verbrugge and Ascione, 1987; CDC, 1994). Arthritis encompasses over 100 specific diseases which primarily impacts the joints (Arthritis Foundation, 1988). High prevalence combined with negligible mortality results in a high incidence of physical disability among older adults, primarily impacting their ability to walk.

Based on data from the Supplement on Aging (SOA) from the 1990 NHIS, the National Center for Health Statistics (1993) reported that for elderly persons with arthritis, walking was the most common ADL difficulty. For males, 17.9% ages 65-74 years, 25.0% 75-84 years, and 37.0% 85+ years reported difficulty. For females, 20.4% ages 65-74 years, 32.8% 75-84 years, and 48.7% 85+ years reported difficulty.

In a comparative study of adults 55+ with and without arthritis, Verbrugge, Gates and Ike (1991) and Verbrugge, Leplowski, and Konkol (1991) used three items from the SOA to demonstrate that people with arthritis are two to three times more likely to be disabled than those without arthritis. Among people with arthritis, 22% had difficulty walking and 68% had functional limitations related to strength and stamina, compared to only 8% and 33%, respectively, for those without arthritis. However, disability was more likely as the numbers of comorbidities rose with age, especially at ages 85+. Thus, although arthritis has a pronounced effect on physical functioning, it does not, by itself, account for dysfunction. Rather, disability is usually the result of arthritis and other concurrent conditions.

It is estimated that as many as 480,000 new cases of osteoarthritis, a degenerative joint disease that primarily affects the knees and hips, arise in the US. white population each year (Abyad and Boyer, 1992; Cushnagan and Dieppe, 1991; Wilson, Michet, Ilstrum, and Melton, 1990). Although osteoarthritis is more prevalent, the impact of rheumatoid arthritis is generally more severe. Verbrugge and her colleagues (1991) found that people with rheumatoid arthritis were more likely to experience physical disability than persons with other forms of the disease. For example, 37.7% of individuals with the rheumatoid type experienced difficulty walking compared to only 21.9% and 25.1% for osteoarthritis and axial arthritis, respectively. Moreover, the odds ratios for difficulty walking increased from 1.31 (ages 55-64) to 4.92 (85+) for rheumatoid arthritis, whereas it only increased from 1.21 to 3.46 for osteoarthritis.

Hip fracture. The estimated 200,000 hip fractures that occur in the United States each year constitute a major cause of morbidity and mortality in persons 65+ years of age. Further, evidence suggests that the incidence of hip fractures is increasing in frequency (Melton, O'Fallon, and Riggs, 1987; Aviol, 1991) and doubles every five years after age 50 (Mehta, Nastasi, 1993). Hip fractures, most often associated with osteoporosis, are most common among women over 50 years old, an estimated 59% of whom will sustain an osteo-related fracture (Chrischilles, Butler, David, and Wallace, 1991).

Stroke. Stroke is the third leading cause of death in the United States and an important cause of disability and morbidity among elderly persons. Using data from the National Vital Statistics System, Feinleib, Ingster, Rosenberg, Maurer, Singh, and Kochanek (1993) found that cerebrovascular disease accounted for a total of 145,551 deaths or 6.8% of the deaths from all causes. However, stroke mortality rates, which started to decline about 1910 (Whelton, 1982), decreased significantly in the 1970s (Klag, Whelton, and Seidler, 1989; Whelton and Klag, 1987), falling from 66.3 deaths per 100,000 in 1970, to 40.8 in 1980, to 28.0 in 1989 (NCHS, 1991). This decline has continued by about 3% per year and shows no signs of decelerating (Feinleib et al., 1993).

Higgins and Thomas (1993) reported that a number of favorable trends in risk factors have contributed to the overall decline in stroke mortality. These include: improvements in economic and living conditions; effective treatment of hypertension, heart disease, and diabetes mellitus; reductions in cigarette smoking and alcohol consumption; and dietary changes such as a decline in consumption of whole milk, lower intake of saturated fat and salt , and rise in consumption of poultry and fresh vegetables.

However, the decline in mortality has not been matched by a coincident decrease in stroke incidence. Using data from the National Hospital Discharge Survey (NHDS), Kovar, Pokras, and Collins (1993) found that the proportion of the population hospitalized for stroke remained fairly constant over the 20 years between 1970 to 1990, while the proportion leaving the hospital alive, more than doubled. These findings were corroborated by Barker and Mullooly (1992), who used World Health Organization criteria in a health maintenance organization population and by Wolf and D'Agostino (1993) who studied 4298 men and women, aged 55 to 64 years, over three successive decades.

The decline in stroke mortality, however, is not without consequences. While the decline continues to be one of the major contributors to improvement in the life expectancy of Americans (Feinleib et al., 1993) increasing longevity, rising prevalence of coronary heart disease, particularly among the elderly, and increased survival following heart attacks and other heart diseases could lead to an increase in the number of stroke victims and those with disabilities in the future.

Diabetes. Diabetes mellitus is the most important cause of lower extremity amputation and a major cause of disability in the US. Data based on the NHIS, the NHDS, and Medicare claims indicate that, in 1989, approximately 6.7 million persons in the United States reported having diabetes and a similar number probably had this disease without being aware of it (Geiss, Herman, Goldschmid, DeStefano, Eberhardt, Ford, German, Newman, Olson, and Sepe, 1993). There also continues to be an increase in the rate of diabetes with 900,000 new cases between 1980-1989. Although the incidence of the disease reached a plateau, increasing only 4% in the 1980s, the incidence among black males increased 28% in the decade, with an average of 648,000 cases per year (Geiss et al., 1993).

The prevalence of disability from lower extremity amputation (LEA) due to diabetes is strongly associated with age, ranging from 2.5% (under 44 years) to 7.5% (65-74 years) to 18% (75+). Thus, despite stability in the rate of LEAs in the 1980s (Geiss et al., 1993), the absolute number of persons with disabilities due to amputation is expected to increase due to the precipitous increase in the older cohorts.

Conclusions: Changing Population Distribution

Data indicate that the "Baby Boom" shifted the modal age from the 10-14 age group in 1970, to 30-34 age group in 1990 (Statistical Abstracts of the United States: 1982-83 (103d edition) and 1992 (112th edition). In addition, the percent of the US population 65+ years increased from 9.2% (20.1 million) in 1970 to 12.6% (28.9 million) in 1990 and will grow to 14% (34.9 million) in 2000, and 20% (39.3 million) of the population in 2020 (Abrams and Berkow, 1990).

In addition, by the first half of the 21st Century, age will become the major risk factor in the epidemiology of disability (Kunkel and Applebaum, 1992). Using 1986 baseline disability data, these researchers projected that there would be a threefold increase (from 5.1 to 14.8-22 million) in the number of Americans with long-term disabilities by the year 2040 when baby boomers reach age 85+. There will also be a higher incidence of disability among women due to an increased life expectancy and higher incidence of osteoporosis and hip fractures.

Almost three-quarters of all assistive devices used for mobility are used by people over 55 years (Assistive Devices Supplement to the 1990 NHIS). LaPlante et al. (1992) found age-related patterns of use that are likely due to changing activity patterns and functional capabilities. They reported that the use of "anatomical technology" (braces, prosthetic limbs) declines significantly and regularly with age and the use of mobility technologies (e.g., canes, walkers, wheelchairs) increases with age. Given the predominance of assistive mobility device use among older people, it may be expected that the aging US population will result in a corresponding increase in the use of assistive devices for mobility.

The impact of aging on disability is expected to peak when the baby boom cohorts reach age 85+ between 2032 and 2048. Based on data from the 1984 National Long Term Care Survey (NLTCS) and the 1985 National Nursing Home Survey (NNHS), Manton (1989) projected that the chronically disabled, community resident elderly population will increase from 5.6 to 15.4 million between 1985 and 2060. The comparable population 85+ is projected to increase from 1.1 to 5.6 million. As a result, by the mid-21st Century, a significant portion of the community dwelling US. population is expected to have impaired functional abilities.

HUMAN SUBJECTS TESTING

Sampling Frame

Based on the literature review undertaken in Task 1 and the population trends identified in Task 2, two sampling frames were developed. One that used the 1990 Assistive Devices Supplement to accurately reflect the population profile of people with mobility impairments in the US today (Table 1); and one that accounted for an anticipated 2.5% increase in the over 55 population to derive a representative sample in 2010 (Table 2). In both cases the total sample of 192 subjects was distributed by age and gender among the seven categories of assistive devices used for mobility and an eighth category of individuals with mobility impairments who do not use assistive mobility aids.

Although the distribution of subjects represented in Table 2 more accurately reflects the US population in 2010, the sampling frame presented in Table 1 was more practical for purposes of this study. First, the older population was already over-represented in the current population (i.e., 73% of participants aged 55 years and older). Therefore, an increase in the proportion of these subjects would not affect the validity of the results. Second, an increase in older population to reflect percentages in 2010 would result in a concomitant decrease in the number of subjects in the three youngest age categories. Whereas the current population had sufficient numbers of subjects in a combined under 16 and the 17-34 age categories to perform statistical analyses, similar analyses would not be possible in the projected distribution in Table 2. Finally, from a logistical standpoint, the general frailty and decreased mobility of the older population would increase the likelihood that a larger sample of older subjects would result in more individuals who were unable to complete all of the test trials.

Methodology

Subject Recruitment. Several methods were used to solicit subjects to take part in the study. First, individuals listed in the Center for Accessible Housing's Design Advisory Network (DAN) database who matched the sampling frame were identified and contacted by telephone. Second, telephone inquiries were made to local disability support groups and advocacy groups as well as rehabilitation facilities and agencies, to obtain their assistance in identifying subjects. Flyers were delivered to local libraries, churches, group homes and care facilities, senior citizens' centers, and suppliers of assistive devices (e.g., orthotics shops). Ads were run in the Raleigh newspaper, and because a large number of older adults were needed, in a local publication targeted to seniors. Center staff made presentations at several meetings of senior citizens and other local disability support groups.

Table 1. Distribution of Subject Sample to Reflect Current Population

Table 2. Distribution of Subject Sample to Reflect Population in 2010

Test Apparatus. The 30 ft. aluminum ramp consisted of three 3 ft. wide by 10 ft. long sections with a triple row of tubular aluminum side railings (at 8, 21.5, and 34 in. above the ramp surface). A 5 ft. ramp created a transition from the floor to the 3.5 in. height of a 4 ft. square level entry platform located at the starting point on the ramp. The top end of the ramp was connected to a 5 ft. by 8 ft. aluminum terminal landing fitted with a triple row of tubular aluminum side railings. The ramp was adjustable to seven calibrated slopes: level, 1:20, 1:16, 1:14, 1:12, 1:10 and 1:8 (Figure 1). Two manually operated winches adjusted the height of the ramp. The terminal landing height was controlled by two additional manually-operated winches. The ramp surface was marked with 1/2 in. vinyl tape across the entire width of the ramp at 6 in. intervals and labeled in one ft. increments, from "0" to "30" along either side of the ramp in order to facilitate documentation of the locations of events that occurred on the ramp. The entire ramp apparatus was designed to be assembled and disassembled with relative ease so it could be transported to different sites, as needed, to access participants.

Test Site. The test site was the lobby space from a former hotel that had been purchased by the University. This space was ideal as a minimum floor space of 20 ft. by 50 ft. and a minimum vertical clearance of 12 ft. was necessary to accommodate the test ramp. In addition, the site was accessible and had accessible restroom facilities, if needed. Finally, an indoor site was required to be able to maintain a comfortable temperature and safe conditions for the subjects throughout the testing period as well as adequate lighting for videotaping the trials.

Test Procedures. In order to minimize variation in testing conditions and possible biasing of experimental data the standardized test procedure was strictly adhered to. At a minimum, three staff were present for each trial. One was responsible for videotaping the session, one served as a spotter for the subject and monitored physical exertion, and one recorded data. All staff members involved in testing were certified by the American Red Cross in basic first aid and CPR. Prior to initiating the test trials a signed consent form was obtained, test procedures, equipment, and safety issues were explained, surface electrodes were positioned on the subject's forehead, and the subject was escorted to the bottom landing.

The two baseline trials (forward and return) were always conducted first, with the ramp in the level position. Following the two baseline trials, subjects were asked to ascend and descend 6 additional slopes (1:8, 1:10, 1:12, 1:14, 1:16, and 1:20). Thus, each subject undertook a total of 14 trials. In order to minimize the effects of fatigue on subjects' ability to use the ramp, the 6 inclines were presented in a random order determined by rolling a die prior to the session.

Prior to each trial, subject's were asked if they would attempt the incline alone. Participants were then instructed to traverse the ramp at a comfortable pace, pausing as needed and simulating the way they would typically use a ramp. Pulse and oxygen saturation levels were monitored continuously throughout the trial to alert the spotter should the subject become overexerted. The participant then rated the ramp difficulty. The spotter(s) remained with the subject on the top or bottom landing, as necessary, until pulse and oxygen levels stabilized to within 3% of resting levels. Upon completion of all trials, the participant was asked to complete a participant evaluation form to provide feedback on their impressions of the testing procedures. Participants were asked to judge how closely the test procedures matched their real-life situations with ramps in the everyday environment. Trial data were recorded on preprogrammed Newton Message Pad 110s. At the end of each day of testing, data were uploaded from the Newtons to a Macintosh computer and onto a database application.

Figure 1. Comparison of Ramp SlopesTested

Data Collection

Data for seven dependent variables were collected for each ramp trial.

1. Total Distance was a measure of the farthest distance traveled on the ramp. Minimum and maximum distances were 0 and 30 feet, respectively. Total distance was defined as the last 6-inch interval crossed by the back of a subject's trailing foot or the rear-most wheel of the wheelchair or scooter. Total distance was recorded when either the entire ramp had been traversed or forward motion on the ramp had ceased due to: 1) a subject's inability to continue, 2) the spotter terminating the trial (e.g., because of escalated pulse rate or reduced oxygen saturation level), or 3) a subject ceasing forward motion for more than 5 seconds for the third time during the trial.

2. Total Time was a measure of the elapsed time from the time the starting point was crossed to the stopping point. Timing began when the back of the subject's trailing foot or the rear-most wheel of a wheelchair/ scooter crossed the starting line. Timing was discontinued when the subject either traversed the entire ramp or had ceased forward motion (when neither foot or rear-most wheel was moving in the direction of travel for at least five seconds).

3. Location and Duration of Rest Stops were recorded each time a subject paused on the ramp for at least 5 seconds. The location was determined as the last six-inch interval crossed by the subject's trailing foot or the rear-most wheel of his/her wheelchair or scooter. The duration was defined as the elapsed time from the instant forward motion ceased until forward motion was resumed. Subjects were allowed to take up to two, 15 minute rest stops during a trial.

4. Pulse Rate and Oxygen Saturation Level were recorded electronically using a Nellcor N20 portable pulse oximeter that was attached to a surface electrode on subjects' foreheads and held in place with a cotton elastic headband. Before each trial, the subject's resting pulse and blood oxygen saturation level were recorded. Changes in the subject's pulse and blood oxygen saturation at the end of each trial were recorded to document exertion due to the trial.

5. Difficulty Levels and Problems Encountered including ability to control speed of ascent and descent as well as railing usage (one or both sides) were noted. In addition, types and locations of problems including stumbling, slipping, tipping of wheelchairs, collisions, faltering due to too little speed on ascents, hesitant descents, unusual directions of travel, and meandering were recorded.

6. Type of Assistance given to a subject (e.g., physical support or preventing loss of control) as well as location on the ramp was recorded.

7. Subjective Ratings of Trial Difficulty were recorded on a 10-point graphic scale after each trial. The scale was anchored by five faces above the scale, registering expressions ranging from satisfaction to frustration. Subjects were asked to choose the face or number that most accurately represented how difficult they perceived that trial.

Results

Inter-rater Reliability. Two observers simultaneously observed and independently recorded performance in all trials for 68% of the subjects. If the total time taken to travel the ramp or the length of a rest stop for any trial differed by more than 5% between the two observers, or if the maximum distance traveled or the location of any rest stop differed by more than six inches (0.5 feet), the videotape of that trial was reviewed by the two observers in order to reconcile the error. This procedure resulted in a 100% inter-rater reliability.

Sample. The sample was selected to match the general population of users of mobility aids. A total of 171 subjects (Table 3) representing 43 of the 51 cells in the sampling frame participated. The ratio of males (41.5%) to females (58.5%) remained true to the sampling frame (41.7% to 58.3%). Seven of the eight cells for whom no participants were found were single-representative cells, the eighth was a two-person cell. Older participants over 75 years of age accounted for the largest number (10) of missing subjects. However, because this age group was the largest in the sampling frame, 85.7% of the targeted number of subjects participated. By mobility category, subjects who used canes or crutches had the largest shortfall (14) from the sampling frame. However, this subgroup accounted for the largest single cell in the sampling frame (99). As a result, 85.9% of the targeted sample was tested.

Table 3. Test Sample by Age and Mobility Aid

Table 4. Frequency of Test Trials by Subject

* Because gait of crutch and cane users was radically different, the two groups were analyzed separately.

** Other includes participants who used: electric wheelchair (1), scooter (1), crutches (7), prosthesis (3), and no aid but had an activity limitation (5).

Distance Traveled. The sample as a whole had little trouble completing each of the trials in ascent, and had even less difficulty in descent. In fact, so few trials (only 3 out of 1178) were not completed in descent, that the data are not significantly different from chance. Moreover, the incomplete trials occurred at three different slopes, including one subject who did not walk the return 30 feet at zero (0) slope.

The sample as a whole also completed the vast majority of trials (1162 out of 1179) in ascent. However, despite the 98.6% completion rate for all trials, the distribution of the incomplete trials fell almost exclusively within the manual wheelchair subsample -- 15 of the 17 (88.2%) were by subjects who used manual wheelchairs and the remaining two (11.8%) were by those who used canes. However, these figures only represented 10.8% and 0.4% of all trials involving people who used wheelchairs and those who used canes, respectively.

When ascent is analyzed by slope and mobility category, the two trials in which a cane user was unable to reach the top of the ramp occurred on the last two trials at the two shallowest slopes after the subjects were able to walk all 30 feet at each of the other slopes. As a result, the failures were probably attributable to overall fatigue due to testing rather than the subjects' inability to traverse a given slope.

In contrast, for subjects who used wheelchairs, the three steepest slopes accounted for the 80% of the incomplete trials. Moreover, the same five subjects were responsible for all of the incomplete trials (Table 5). One subject failed to complete all six slopes, and two additional subjects failed to complete the three slopes steeper than 1:14. Therefore, of the 20 subjects in manual wheelchairs, 95% were able to complete slopes between 1:20 and 1:14. Only 85% were able to traverse 30 feet at a slope of 1:12, 80% were able to traverse 30 feet at 1:10, and 75% were able to traverse 30 feet at 1:8. Interestingly, the five the subjects who failed to ascend all 30 feet were all women, four of whom were over 65 years of age. However, age by itself, was neither a significant factor in determining success nor mean distance traveled (Table 6).

Table 5. Subjects in M

anual Wheelchairs who did not Ascend all 30' by Slope

Table 6. Mean Distance Traveled (in feet) in Ascent by Age Group (all mobility groups)

Rest Stops. Of the 1179 ascent trials only 24 (2.0%) involved the participant stopping to rest. Subjects who did not complete a trial also stopped on the ramp more often, and at shorter distances than those who completed the trials. This is evidenced by the greater mean number of rest stops for those who did not traverse all 30 feet (1.24 per trial), than those who were able to traverse all 30 feet (.01). The mean distance to rest stops for the incomplete group was also less (7.44 feet for rest stop 1 and 12.83 feet for rest stop 2) than for subjects who traversed all 30 feet (16.98 feet and 25.75 feet, respectively). However, when number, distance, and duration of stops were analyzed by slope, there was no significant effect of slope on these measures. Although the number of subjects who stopped once increased between slopes of 1:20 and 1:8, this increase was small (1.2% to 5.2%) and represented a difference of only 7 people. Moreover, of the nine that stopped on the 1:8 slope, 7 (77.8%) continued to the end of the ramp.

Speed. The overall trend,with few exceptions was that speed of ascent decreased as slope increased from 1:20 to 1:8 (Figure 2). Although the decrease in speed within each mobility aid subgroup was generally not significant, there was a downward trend. The data also reveal that subjects with different mobility aids traveled at different rates of speed, regardless of slope. Although increases in ramp gradient had a proportionately similar effect on speed for all ambulatory and "other" subjects, it had a dramatic effect on speed of wheelchair users. Interestingly, however, this effect was fairly equally evident across all slopes, not just the steeper gradients. The only significant difference noted was the 50% reduction in speed between the level (0 slope) and a 1:8 slope within the manual wheelchair group.

Figure 2. Mean Rate of Travel in Ascent

Pulse. Across mobility categories, there were no significant differences at zero slope indicating that pulse rates, as would be expected were similar for all subgroups on a level surface. However, subjects who used wheelchairs had significantly higher changes in pulse rate at each gradient, indicating that slope had a greater effect on people who used wheelchairs. When pulse was examined within mobility aid category, rate remained fairly stable for the two ambulatory subgroups that used mobility aids (canes, and walkers). In contrast, there were significantly higher and increasing pulse rates among the wheelchair and brace subgroups. For subjects who used braces, there were significant increases in pulse rates between zero and the two steepest slopes, although there were no differences between any other gradients. Among subjects who used manual wheelchairs, there was a significant increase between 0 and 1:14 slope. At 1:14 the change in pulse rate reached a plateau with no significant differences among slopes of 1:14, 1:12, and 1:10. At 1:8, pulse rates increased again.

Oxygen Saturation. Mean levels of oxygen saturation ranged from 98.24% to 99.77%. Not only were there no statistical differences in these measures, but the mean levels were so high (exertion is only evident when saturation begins to fall below 90%) that exertion was not evident in any of the mobility subgroups. This suggests that 30 feet at any of the gradients tested may not have been a strenuous task for any of the mobility subsamples.

Figure 3. Mean Change in Pulse in Ascent

Subjective Responses. Although mean difficulty ratings in ascent increased significantly within each mobility aid subgroup as slope increased, only two mobility subgroups -- manual wheelchairs (5.84 on 1:10 and 6.73 1:8) and walkers (5.16 on 1:8) had mean difficulty ratings above 5 on a 10-point scale on any slope. It is also interesting to note that these two subgroups rated each level more difficult than the other mobility aid subgroups. In contrast, in descent, the differences among subgroups were less evident as no mean rating exceeded 4.0 for any slope, except on the steepest slope by those who used walkers (mean = 5.58).

When subjects were asked if they would try to ascend the ramp alone, 5 (2 with braces, 1 cane user, 2 with walkers) responded that they would not try a slope of 1:8 and 1 (with a walker) would not try a slope of 1:10. In descent, 9 subjects reported that they would not try slopes ranging from 1:14 to 1:8, although 7 of the 9 involved the two steepest slopes. These subjects included individuals who used walkers (1 at 1:12, 2 at 1:10, and 3 at 1:8), canes (1 at 1:8), and braces (one at 1:8). One person in a manual wheelchair would not have attempted a slope of 1:14.

Conclusions

Although not particularly surprising, the data indicate that none of the factors of interest in this study - slope, mobility aid, gender, and age -- had much of a bearing on subjects' ability to travel or amount of effort expended in descending a 30 foot long ramp. However, a number of factors, most notably the interaction between ramp slope and manual wheelchair users had an impact in ascent.

Distance traveled is perhaps the most notable evidence suggesting that slope impacts performance by wheelchair users. Of the 151 ambulatory and "other" subjects, only two cases (both involving cane users at shallow gradients), resulted in subjects terminating prior to reaching the top of the ramp. Similarly, other studies (Steinfeld, et al., 1979; Van der Voordt, 1981; and Walter, 1971), have reported very high percentages of ambulatory participants using mobility aids who were able to traverse ramps of varying slopes and distances.

In contrast, data from this research indicates that only 85% of the participants who used manual wheelchairs were able to traverse 30 feet at a slope of 1:12, 80% were able to traverse 30 feet at 1:10, and 75% were able to traverse 30 feet at 1:8. Although these findings are consistent with some previous research, the varying slopes and lengths of ramps used in previous research makes direct comparisons difficult. For example, the Templer, Wineman, and Zimring (1982) also found that 80% of manual wheelchair users could traverse a ramp with a slope of 1:10.1. However, these researchers concluded that this population could traverse a 99 foot-long ramp at a gradient of 1:10.1. More recently, Sweeney, Clarke, Harrison, and Bulstrode (1989) reported that 88% of the subjects who participated in their study were able to traverse short gradients between 1:12 and 1:7. In contrast, other studies have reported greater numbers of self-propelled manual wheelchair users having difficulty on gradients of 1:12 and steeper. For example, Steinfeld, et al. (1979) found that almost 50% of their subjects were unable to ascend a 40 foot ramp with a 1:12 slope and only 30% of the manual wheelchair users could complete a distance of 5 feet at this gradient.

Thus, a 1:12 slope at distances of 30 feet may be too difficult for certain segments of the wheelchair population, most notably those who are older and female. This is further evidenced by the increased number of rest stops required by manual wheelchair users within the 30 foot distance in comparison to the rest of the sample. However, subject recruitment efforts for this study suggest that whereas the population of older women who use manual wheelchairs is relatively large, the number who independently travel outdoors may be relatively small. These individuals tended to use manual wheelchairs to get around the home when they tired, but relied on someone else to push them when they were outside their homes. This seems to be the result of their frailty and inability to propel themselves over the distances required, rather than the slope of the surfaces. As a result, the actual impact of ramp slope on independent mobility of this population may be minimal. Further research should be conducted to determine the extent to which older women are likely to engage in independent outdoor mobility such that the impact of ramp slope can be assessed.

The data also indicate, that only people who use manual wheelchairs experienced significant decreases in speed due to slope and only between the baseline and a 1:14 slope and between 1:14 and 1:8. These subjects, unlike the rest of the sample, also experienced a significant increase in pulse rate, although it was only between the baseline and a slope of 1:14. In contrast, the oxygen saturation data, like that from previous studies (Findley and Agre, 1988), indicate that wheelchair users expend only slightly more energy than ambulatory subjects in traversing an incline, although none of the subgroups exhibited significant cardiovascular exertion. It is likely that other factors, such as muscle fatigue due to distance or speed may have a greater impact on exertion than does slope.

In summary, the data suggest that 30 foot-long ramps, particularly those at slopes of 1:12 and higher, present some difficulty for manual wheelchair users, but not for people who use other types of mobility aids. However, with the exception of older females, whose tendency to engage in independent outdoor mobility is questionable, a surprisingly high percentage of manual wheelchair users were able to traverse fairly steep ramps (75% at 1:8), and did so without expending energy at risky levels. This suggests that perhaps even older individuals could traverse ramps steeper than 1:12 if there were shorter distances between level landings. Such an assertion is supported by others (Canale, Felici, Marchetti, & Ricci, 1991; Cappozzo, Felici, Figura, Marchetti, & Ricci, 1991; Sweeney, et al., 1989; Sweeney & Clarke, 1991) who have suggested that people who use wheelchairs may be more successful at ascending short ramps of steeper gradients than long ramps of slopes 1:12 or less. However, despite acceptable performance in ascending steeper inclines, data from this as well as previous studies (van der Voordt, 1981; Walter, 1971) suggest that people who use wheelchairs fear tipping over backwards on slopes greater than 1:12. As a result, steeper gradients may be less desirable for this population.

RECOMMENDATIONS

Based on the performance of the entire sample it could be argued that the maximum allowable slope of ramps could be increased, even if for distances shorter than 30 feet. Alternatively, it could be argued that the maximum distance allowable for 1:12 ramps could be increased beyond 30 feet. In fact, these conclusions are bolstered by anecdotal evidence that suggests that those manual wheelchair users, who were most impacted by ramp slope -- women over 65 -- may be too frail to routinely travel outdoors independently, even on level surfaces. However, a number of other factors suggest that the technical requirements for ramps should not be changed at this time.

Most importantly, there are no data, other than anecdotal evidence, to discount older women as potential users of ramps. As a result, the 20 - 25% of people who use manual wheelchairs who had difficulty traversing slopes greater than 1:12, but only 15% at 1:12, are a major consideration in keeping the technical requirement for ramp slope at 1:12. Nonetheless, further research with elderly users of wheelchairs might suggest the need for ramps with shallower slopes at facilities that have higher proportions of older people, such as senior centers, nursing homes, assistive living facilities, and independent living communities.

In addition, a number of factors related to the research design that were necessary to ensure subject safety, may have created conditions that enhanced performances that would otherwise not have occurred in a real world setting. First, subjects who participated may not have been truly representative of individuals in the mobility categories. Because subjects were self-selecting, it is likely that only individuals who thought that they could complete the test trials volunteered. As a result, participants were probably healthier and had better strength and stamina than others in the general population who use the same mobility aids. Second, for medical reasons, only subjects who were in good physical condition were permitted to participate. Third, because the ramp was located indoors in a climate controlled setting it was not subject to weather conditions such as rain or ice, which could adversely affect use. Fourth, in retrospect, the ramp itself may have been perceived to be safer and easier to use than ramps typically found in public settings. The test ramp was designed for use by one individual at a time. It was three feet wide and had handrails on both sides such that an individual could grab both rails at the same time. Although subjects were instructed to only use one rail, they had the option of using either one (in contrast to standard US convention of always traveling to the right side), or both, if necessary. In addition, rails were provided at three heights, providing greater convenience than typically found in the real world. Therefore, it was not surprising that many subjects commented that they could not have traversed the ramp had it not been for their ability to use the handrails. However, in real world situations, ramps other than ones located at private residences will be more than three feet in width in order to accommodate multiple users. Therefore, unless the effective width of a ramp is limited to 3 feet (i.e., a three-foot maximum distance between handrails), which will require that all ramps be constructed in increments of three-foot widths with intermediate handrails (e.g., a six-foot wide ramp with handrail down the middle), many individuals who were able to traverse the entire length of the test ramp, may not be able to do so in a real world situation. Unfortunately, requiring ramps to have handrails no more than three feet apart seems to be costly and quite impractical. Although it could be considered as equivalent facilitation if a ramp steeper than 1:12 is desired. Finally, the presence of the spotter on the ramp during the test trials probably biased the results. Although the spotter was not expected to, and did not intervene, unless necessary, several subjects stated that they would not have attempted to traverse the steeper gradients by themselves, and only did so because the spotter was present. As such, even though individuals were able to traverse ramps steeper than 1:12, they probably would not attempt to do so in real life situations.

Therefore, based on the considerations outlined above, changes to the technical requirements for ramp slope and length cannot be recommended at this time. However, this issue should be considered again following further research directed at understanding the functional limitations of older wheelchair users as well as the impact of effective ramp width on ramp use.

REFERENCES AND SELECTED READINGS

Current Literature and Standards

Annesley, A.L., Almada-Norfleet, M., Arnall, D.A., & Cornwall, M.W. (1990). Energy expenditure of ambulation using the Sure-Gait crutch and the standard auxiliary crutch. Journal of Physical Therapy, 70(1), 18-23.

Baitch, S.P., Blake, R.L., Fineagan, P.L., & Senatora, J. (1991). Journal of American Podiatry Medical Association, 81(12), 647-652.

Banta, J.V., Bell, K.J., Muik, E.A., & Fezio, J. (1991). Parawalker: energy cost of walking. European Journal of Pediatric Surgery, 7-10.

Berg, K.O., Maki, B.E., Williams, J.I., Holliday, P.J., & Wood-Dauphinee, S.L. (1992). Clinical and laboratory measures of postural balance in an elderly population. Architectural-Physiological-Medical Rehabilitation, 73(11), 1073-1080.

Bhambhani, Y.N., Holland, L.J., & Steadward, R.D. (1993). Anaerobic threshold in wheelchair athletes with cerebral palsy: validity and reliability. Architecture-Physiology-Medical Rehabilitation, 74(3) 305-311.

Black E.R. (1980). Use of special aids. United States, 1977 Vital and Health Statistics. National Center for Health Statistics.

Bogey, R.A., Barnes, L.A., & Perry J. (1992). Computer algorithms to characterize individual subject EMG profiles during gait. Architecture-Physiological-Medical Rehabilitation, 73(9), 835-841.

Bowker, P., Messenger, N., Ogilvie, C., & Rowley, D.I. (1992). Energetics of paraplegic walking. Journal of Biomedical Engineering, 14(4), 344-350.

Brodke, D.S., Skinner, S.R., Lamoreux, L.W., Johanson, M.E., St. Helen, R., Moran, S.A., & Ashley, R.K. (1989). Effects of ankle-foot orthoses on the gait of children. Journal Pediatric Orthopedics, 9(6), 702-708.

Brubaker, C.E., McLaurin, C.A., & McClay, I.S. (1986). Effects of slide slope on wheelchair performance. Journal of Rehabilitation Research and Development, 23(2), 55-58.

Butler, P.B., Thompson, N., Major, & R.E. (1992). Improvement in walking performance of children with cerebral palsy: Preliminary results. Developmental-Medical-Child Neurological, 34(7), 567-576.

Calder, C. J, & Kirby, R. L. (1990). Fatal wheelchair-related accidents in the United States. American Journal of Physical Medicine and Rehabilitation, 69(4), 184-190.

Canale, I., Felici, F., Marchetti, M., & Ricci, B. (1991). Ramp length/grade prescriptions for wheelchair dependent individuals. Paraplegia, 29, 479-485.

Cappozzo, A., Felici, F., Figura, F., Marchetti, M., & Ricci, B. (1991). Prediction of ramp traversability for wheelchair dependent individuals. Paraplegia, 29(7), 470-480.

Cardus, D., McTaggart, W.G., Ribas-Cardus, F., & Donovan, W.H. (1989). Energy requirements of gamefield exercises designed for wheelchair-bound persons. Architecture-Physical Medicine-Rehabilitation, 70(2), 124-127.

Center for Accessible Housing. (1992). Recommendations for Accessibility Standards for Children's Environments. U.S. Architectural and Transportation Barriers Compliance Board, Technical Report #QA90003001.

Cerf-Beare, A. (1975). Environmental design and architectural barriers: An ergonomic approach. Man-Environment-Systems, 5(1), 49-57.

Chapanis, A. (1974). Human engineering environments for the aged. Applied Ergonomics, 5(2), 72-80.

Corlett, E. N, & Hutchinson (1972). Ramps or stairs--the choice using physiological and biomechanical criteria. Applied Ergonomics, 24, 195-201.

Couch, R.H. (1992). Ramps not steps: a study of accessibility preferences. Journal of Rehabilitation, 58(1), 65-69.

Cowan, R. (1987). When small steps are giant leaps. Architects Journal, 185 (17), 15.

Crouse, S.F., Lessard, C.S., Rhodes, J., & Lowe, R.C. (1990). Oxygen consumption and cardiac response of short-leg and long-leg prosthetic ambulation in a patient with bilateral above-knee amputation: Comparisons with able-bodied men. Architecture-Physical Medicine-Rehabilitation, 71(5), 313-317.

Davis, G.M., Shepard, R.J., & Jackson, R.W. (1981). Cardio-respiratory fitness and muscular strength in the lower-limb disabled. Canadian Journal of Applied Sport Sciences, 6(4), 159-165.

Dubow, L. L., Witt, P. L., Kadaba, M. P., Reyes, R., & Cochran, G. V. (1983). Oxygen consumption of elderly persons with bilateral below knee amputation: Amputation vs. wheelchair propulsion. Archives of Physical Medicine and Rehabilitation (64), 255-259.

Dudley, N.J. (1993). Wheelchair safety. Lancet, 341(8839), 246.

Ehara, Y., Beppu, M., Nomura, S., Kunimi, Y., & Takahashi, S. (1993). Energy storing property of so called energy-storing prosthetic feet. Architecture-Physiology-Medical Rehabilitation, 74(3), 68-72.

Elmer, C. D. (1957). A study to determine the specifications of wheelchair ramps. Unpublished masters thesis. University of Iowa.

Engelke, K.A., Shea, J.D., Doerr, D.F., & Convertino, V.A. (1992). Enhanced carotid-cardiac baroreflex response and elimination of orthostatic hypotension 24 hours after acute exercise in paraplegics. Paraplegia, 30(12), 872-879.

Farris, D.A., Urquizo, G.C., & Beattie, D.K. (1993). A simplified accelerometer system for analysis of human gait. Experimental Techniques, 17, 33-36.

Francis, G.A, & Norstrom, D. (1992). Guidelines specifications for passive lifts, active lifts, wheelchair ramps, and securement devices. U.S. Federal Transit Administration.

Feeney, R.J, & Galer, M.D. (1981). Ergonomics research and the disabled. Ergonomics, 24(1), 821-830.

Ferguson, A. (1986). Practice standards: Draft codes on disabled escape. Architects Journal, 183(8), p73.

Findley, T.W, & Agre, J.C. (1988). Ambulation in the adolescent with spina bifida. Architecture-Physiology-Medical Rehabilitation, 69(10), 855-861.

Fisher, S.V, & Patterson, R.P. (1981). Energy cost of ambulation with crutches. Archives of Physical Medicine and Rehabilitation (62), 250-256.

Franks, C.A., Palisano, R.J., & Darbee, J.C. (1991). The effect of walking with an assistive device and using a wheelchair on school performance in student with myelomeningocele. Journal of Physical Therapy, 71(8), 570-579.

Frieden, R.A., Ryniker, D., Kenan, S., & Lewis, M.M. (1993). Assessment of patient function after limb sparing surgery. Architecture-Physiology-Medical Rehabilitation, 74(1) 38-43.

Gartshore, P. J, & Sime, J. D. (1987). Assisted escape-Some guidelines for designers, building managers and the mobility impaired. Design for Special Needs, 42 (1), 6-9.

Guidera, K.J., Smith, S., Raney, E., Frost, J., Pugh, L., Griner, D., & Ogden, J.A. (1993). Use of reciprocating gait orthosis in myelodysplasia. Journal of Pediatric Orthopedics, 13(3), 341-348.

Hartung, G.H., Lally, D.A., & Blancq, R.J. (1993). Comparison of treadmill exercise testing protocols for wheelchair users. European Journal of Applied Physiology, 66(4), 362-365.

Haskvitz, E.M. (1992). Energy cost of assistive ambulation in order adults.Unpublished doctoral dissertation, University of Virginia.

Hinton, C. A, & Cullen, K. E. (1982). Energy expenditure during ambulation with Ortho crutches and axillary crutches. Physical Therapy (62), 813-819.

Hirokawa, S., Grimm, M., Le, T., Solomonow, M., Baratta, R.V., Shoji, H., & D'Ambrosia, R.D. (1990). Energy consumption in paraplegic ambulation using the reciprocating gait orthosis and electric stimulation of the thigh muscles. Architecture-Physical Medicine-Rehabilitation, 71(9), 687-694.

Huang, C. T., Jackson, J. R., Moore, N. B., et. al. (1979). Amputation: energy cost of amputation. Archives of Physical Medicine and Rehabilitation (60), 18-23.

Hunter, J. (1987). Energy costs of wheelchair propulsion by elderly and disabled people. International Journal of Rehabilitation Research, 10(4), 50-54.

Hutzler, Y. (1993). Physical performance of elite wheelchair basketball players in armcranking ergometry and in selected wheeling tasks. Paraplegia, 31(4), 255-261.

Jaegers, S.M., Vos, L.D., Rispers, P., & Hof, A.L. (1993). The relationship between comfortable and most metabolically efficient walking speed in persons with unilateral above-knee amputation. Architectural-Psychological-Medical Rehabilitation, 74(5), 521-525.

Joyce, B.M, & Kirby, R.L. (1991). Canes, crutches and walkers. American Family Physician, 43(2), 535-542.

Ju, Ming-Shaung & Mansour, J.M. (1988). Simulation of the double limb support phase of human gait. Journal of Biomechanical Engineering, 110, 223-229.

Kirby, R.L., Atkinson, S.M., & MacKay, E.A. (1989). Static and dynamic forward stability of occupied wheelchairs: Influence of elevated footrests and forward stabilizers. Architecture-Physiology-Medical Rehabilitation, 70(9), 681-686.

Kirby, R.L, & McLean, A.D. (1990). Preventing occupied wheelchairs from falling down stairs. Journal of Rehabilitation Research and Development, 27(1), 27-32.

Kronick, M.K. (1987). Children and canes, an adaptive approach. Journal of Visual Impairment and Blindness, 81(2), 61-62.

Lees, A, & Arthur, S. An investigation into anaerobic performance of wheelchair athletics. Ergonomics, 31(11), 1529-1537.

Lehmann, J.F. (1993). Push-off and propulsion of the body in normal and abnormal gait. Correction by ankle-foot orthoses. Clinical Orthopedics, 288, 97-108.

Lehmann, J.F., Warren, C.G., Halar, E., Stonebridge, J.B., & DeLateur, B.J. (1974). Wheelchair propulsion in the quadriplegic patient. Archives of Physical Medicine and Rehabilitation, 55, 183-186.

Levangie, P.K., Brouwer, J., McKeen, S.H., Parker, K.L., & Shelby, K.A. (1989). The effects of the standard rolling walker and two posterior rolling walkers on gait variables of normal children. Physical and Occupational Therapy in Pediatrics, 9(4), 19-31.

Levangie, P.K., Chimera, M., Johnston, M., Robinson, F., & Wobeskya, L. (1989). The effects of posterior rolling walkers versus the standard rolling walker on gait characteristics of children with spastic cerebral palsy. Physical and Occupational Therapy in Pediatrics, 9(4), 1-17.

Lin, K.H., Lai, J.S., Kao, M.J., & Lien, I.N. (1993). Anaerobic threshold and maximal oxygen consumption during arm cranking exercise in paraplegia. Architecture-Physiology-Medical Rehabilitation, 74(3) 515-520.

Luna-Reyes, O.B., Reyes, T.M., So, F.Y., Matti, B.M., & Lardizabal, A.A. (1988). Energy cost of ambulation in healthy and disabled Filipino children. Architecture-Physiology-Medical Rehabilitation, 69(11), 946-949.

Mattison, P.G., Hunter, J., & Spence, S. (1989). Development of a realistic method to assess wheelchair propulsion by disabled people. International Journal Rehabilitation Research, 12(2), 137-145.

Miller, B.C. (1990). Handicapped ramps: an unpaid debt. Professional Safety, 35, 20-25.

Mossberg, K.A., Linton, K.A., & Friske, K. (1990). Ankle-foot orthoses: effect on energy expenditure of gait in spastic diplegic children. Architecture-Physical Medicine-Rehabilitation, 71(7), 490-494.

Nava, L.C, & Laura, P.A. (1985). Design and use of improved walking aids. Journal of Biomedical Engineering, 7(4), 329-333.

Nene, A.V, & Jennings, S.J. (1992). Physiological cost index of paraplegic locomotion using the ORLAU ParaWalker. Paraplegia, 30(4), 246-252.

Nielson, D.H., Harris, J.M., Minton, Y.M., Motley, N.S., & Rowley, J.L. (1990). Energy cost, exercise intensity and gait efficiency of standard versus rocker-bottom axillary crutch walking. Journal of Physical Therapy, 70(8), 487-493.

Ogilvie, C., Bowker, P., & Rowley, D.I. (1993). The physiological benefits of paraplegic orthotically aided walking. Paraplegia, 31(2) 111-115.

Oneto, G. (1988). Stairways, Ramps and Grades. Ville Giardini,228,XIX-XXIV.

Opila, K.A., Nicol, A.C., & Paul, J.P. (1987). Upper limb loading of gait with crutches. Journal of Biomechanical, 109, 285-290.

Parette, H.P, & van Biervliet, A. (1991). Rehabilitation assistive technology issues for infants and young children with disabilities: a preliminary examination. Journal of Rehabilitation, 57(3), 27-36.

Parziale, J.R. (1991). Standard versus lightweight wheelchair propulsion in spinal cord injured patients. American Journal of Physical Medical Rehabilitation, 70(2), 76-80.

Perry, J., Mulroy, S.J., & Renwick, S.E. (1993). The relationship of lower extremity strength and gait parameters in patients with post-polio syndrome. Architecture-Psychology-Medical Rehabilitation, 74(2), 165-169.

Phillips, C.A. (1989). Functional electrical stimulation and lower extremity bracing for ambulation exercise of the spinal cord injured individual: a medically prescribed system. Physical Therapy, 69(10), 842-849.

Pinzur, M.S., Asselmeier, M., & Smith, D. (1991). Dynamic electromyography in active and limited walking below-knee amputees. Orthopedics, 14(5), 535-538.

Reid, M., Lawrie, A.T., Hunter, J., & Warren, P.M. (1990). The effect of steering on the physiological energy cost of wheelchair propulsion. Scandinavian Journal of Rehabilitative Medicine, 22(3), 139-143.

Robinson, J. (1986). Means of escape for disabled people: defining a realistic approach. Design for Special Needs, 39, 14-16.

Rose, J., Medeiros, J.M., & Parker, P. (1985). Energy cost index as an estimate of energy expenditure of cerebral-palsied children during assisted ambulation. Developmental Medicine of Child Neurology, 27(4), 485-490.

Sedlock, D.A., Knowlton, R.G., & Fitzgerald, P.I. (1990). Circulatory and metabolic response of women to arm crank and wheelchair ergonometry. Architecture-Physical Medicine-Rehabilitation, 71(2), 97-100.

Simard, C., Noreau L., Pare, G., & Pomerleau, P. (1993). Maximal physiological response during exertion in quadriplegic subjects. Journal of Applied Physiology, 18(2), 163-174.

Skinner, H.B. (1993). Pathokinesiology and total joint arthroplasty. Clinical Orthopedics, 288, 78-86.

Smith, A.C. (1985). Poor building designs hinder safe emergency escape by handicapped. Occupational Health Safety, 14(4), 303-312.

Smoyak, S.A. (1991). Barriers and ramps for the mentally ill to community living. Journal of Psychosocial Nursing and Mental Health Services, 29(4), p46.

Steinfeld, E. (1979). Designing entrances and internal circulation to meet barrier free goals. Architectural Record, 166(1), 65-67.

Steinfeld, E. (1979). Designing the site to meet barrier free goals. Architectural Record, 165(5), 69-71.

Steinfeld, E., Schroeder, S., & Bishop, M. (1979). Accessible Buildings for People with Walking and Reaching Limitations. Washington, DC: U.S. Department of Housing and Urban Development, US Government Printing Office.

Summers, B.N., McClelland, M.R., & el-Masri, W.S. (1988). A clinical review of the adult hip guidance orthosis (ParaWalker) in traumatic paraplegics. Paraplegia, 26(1), 19-26.

Sweeney, G.M, & Clark, A.K. (1991). Gradients of portable ramps. BMJ, 302(6787), 1277.

Sweeney, G.M., Clark, A.K., Harrison, R.A., & Bulstrode, S.J. (1989). An evaluation of portable ramps. British Journal of Occupational Therapy, 52(12), 473-475.

Sweeney, G.M., Harrison, R.A., & Clark, A.K. (1989). Portable ramps for wheelchair users--an appraisal. International Disability Studies, 11(2), 68-70.

Templer, J. (1979). Provisions for Elderly and Handicapped Pedestrians: Vol. 1 Executive Summary. Washington, DC: Department of Transportation, Federal Highway Administration, Report # FHWA-RD-79-1.

Templer, J., Lewis, D., & Sanford, J. (1983). Ground and floor surface treatments: technical paper with comments. U.S. Architectural and Transportation Barriers Compliance Board.

Templer, J.A., Wineman, J., & Zimring, C.M. (1984). Guidelines for accommodating the handicapped on crossings: final report. Washington, DC: Department of Transportation, Federal Highway Administration.

Tornburn, L., Perry, J., Ayyappa, E., & Shanfield, S.L. (1990). Below-knee amputee gait with dynamic elastic response prosthetic feet: a pilot study. Journal of Rehabilitation Research and Development, 27(4), 369-384.

Tranter, R.T., Slater, R., & Vaughn, N. (1991). Barriers to mobility: physically-disabled and frail elderly people in their local outdoor environment. International Journal of Rehabilitation Research, 14(4), 303-312.

Travers, A.F. (1991) Ramps and rails. BMJ, 302(6782), 951-954.

Torburn, L., Perry, J., Ayyappa, E., & Shanfield, S.L. (1992). Below-knee amputee gait with dynamic elastic response prosthetic feet: A pilot study. Journal of Rehabilitation Research Development, 27(4), 369-384.

Turner, G, & Collins B. L. (1981). Pedestrian movement characteristics on building ramps. Report NBSIR 81-2310, Washington, DC: National Bureau of Standards.

van der Voordt, D.M. (1979). Ramps tested for usefulness. Architect (The Hague) 10, (12) 78-83.

van der Voordt, D.M. (1981). Accessibility by means of ramps--some research data from the Netherlands. Delft: University of Technology.

van der Woude, L.H.V., Hendrich, K.M.M., Veeger, H.E.J., van Ingen Schenau, G.J., Rozendal, R.H., de Groot, G., & Hollander, A.P. (1988). Manual wheelchair propulsion: Effects of power output on physiology and technique. Medicine and Science in Sports and Exercise, 20, 70-78.

Voigt, E.D, & Bahn, D. (1969). Metabolism and pulse rate in physically handicapped when propelling a wheel chair up an incline. Scandinavian Journal of Rehabilitation Medicine, 1, 101-106.

Walter, F. (1971). Four architectural movement studies for the wheelchair and ambulant disabled. Part 3--ramp gradients. London: Disabled Living Foundation.

Waters, R., Torbum, L., & Mulroy, S. (1992). Energy expenditure in elderly patients using assistive devices. Topics in Geriatric Rehabilitation, 8(2), 12-19.

Waters, R.L, & Lunsford, B.R. (1985). Energy cost of paraplegic locomotion. Journal of Bone-Joint Surgery, 67(8), 1245-1250.

Waters, R.L., Perry, J., Conaty, P., Lunsford, B., & O'Meara, P. (1987). The energy cost of walking with arthritis of the hip and knee. Clinical Orthopedics, 214, 278-284.

Waters, R.L., Yakura, J.S., & Askins, R.H. (1993). Gait performance after spinal cord injury. Clinical Orthopedics, 87-96.

Wirta, R.W., Mason, R., Calvo, K., & Golbranson, F.L. (1991). Effect on gait using various prosthetic ankle-foot devices. Journal of Rehabilitation Research and Development, 28(2), 13-24.

Wolfe, G., Waters, R., & Hislop, H. (1977). Influence of floor surface on the energy cost of wheelchair propulsion. Physical Therapy, 57(9), 1022-1027.

People with Mobility Impairments

Abrams, W.B. & Berkow, R., Editors. (1990). The Merck Manual of Geriatrics. Rahway, NJ: Merck and Co.

Abyad, A. & Boyer, J.T. (1992, April). Arthritis and Aging. Current Opinion in Rheumatology. 4 (2). 153-159.

Aviol, L.V. (1991). Significance of osteoporosis a growing international health care problem. Calcification Tissue International, 49:55-57.

Barker, W. & Mullooly, J. (1992). Declining stroke mortality among older persons in recent years. Presentation at the 44th Annual Scientific Meeting of the Gerontological Society of America , Washington, DC, November 20.

Bhushan, V., Paneth, N., & Kiely, J.L. (1993). Impact of improved survival of very low birth weight infants on recent secular trends in the prevalence of cerebral palsy. Pediatrics, 91(6). 1094-1100. June.

Boyle, C.A., Decoufle, P., & Yeargin-Allsopp, M. (1994, March). Prevalence and health impact of developmental disabilities in US children. Pediatrics, 93(3). 399-403.

Brody, J.A. & Schneider, E. (1986). Diseases and Disorders of Aging: An Hypothesis. Journal of Chronic Diseases, 39:871-6.

Bureau of the Census. (1983). Statistical Abstract of the United States, 1982-83, 103rd Edition. Economics and Statistics Administration, US Department of Commerce.

Bureau of the Census. (1992). Statistical Abstract of the United States, 1992, 112th Edition. Economics and Statistics Administration, US Department of Commerce.

Centers for Disease Control. (1990). Comorbidity of chronic conditions and disability among older persons -- United States, 1984. JAMA, 263 (2): 209-210, January 12.

Centers for Disease Control. (1993). Impact of adult safety-belt use on restraint use among children 11 years of age -- selected states, 1988 and 1989. Morbidity and Mortality Weekly, 42(14): 275-288, April 16.

Centers for Disease Control and Prevention. (1993). HIV/AIDS Surveillance Report. 5. Atlanta, GA: Author. September.

Centers for Disease Control. (1994). Arthritis prevalence and activity limitations -- United States, 1990. Morbidity and Mortality Weekly, 43 (24): 433-8, June 24.

Chirikos, T.N. (1986). Accounting for the historical rise in work disability prevalence. Milbank Memorial Fund Quarterly Healthy and Society, 64: 271-301.

Chrischilles, E.A., Butler, C.D., David, C.S., & Wallace, R.B. (1991). A model of lifetime osteoporosis: Impact. Archives Internal Medicine, 151: 2026-32.

Collins, J.G. (1991). Impairments due to injuries: United States, 1985-87. Vital & Health Statistics--Series 10: Data from the National Health Survey. (177). 1-55. August.

Colvez, A. & Blanchet, M. (1981). Disability trends in the United States population 1966-76: Analysis of reported causes. American Journal of Public Health, 71: 464-471.

Cushnagan, D.L. & Dieppe, J.T. (1991). Study of 500 patients with limb and joint osteoarthritis: Analysis by age, sex and distribution of symptomatic joint sites. American Rheumatology Disease, 50:8-13.

DeVivo, M.J., Black, K.J., & Stover, S.L. (1993). Causes of death during the first 12 years after spinal cord injury. Archives of Physical Medicine & Rehabilitation, 73(3), 248-254. March

DeVivo, M.J., Rutt, R.D., Black, K.J., Go, B.K., & Stover, S.L. (1992). Trends in spinal cord injury demographics and treatment outcomes between 1973 and 1986. Archives of Physical Medicine & Rehabilitation, 73: 424-430.

di Scala, C.D., Osberg, J.S., Gans, B.M., Chin, L.J., & Grant, C.C. (1991). Children with traumatic head injury: morbidity and postacute treatment. Archives of Physical Medicine and Rehabilitation, 72. 662-666. August.

Escobar, G.J., Littenberg, B., & Petitti, D.B. (1991). Outcome among surviving very low birthweight infants: A meta-analysis. Archives of Disease in Childhood. 66(2). 204-211. February.

Evans, L. (1988). Occupant protection device effectiveness in preventing fatalities. In Proceedings of the 11th International Technical Conference on Experimental Safety in Vehicles. Washington, DC: US Department of Transportation, National Traffic Safety Administration, DOT HS 80-7 233: 220-227. May.

Evans, L. (1990). Restraint effectiveness occupant ejection from cars, and fatality reductions. Accident Analysis and Prevention, 22:167.

Feinleib, M., Ingster, L., Rosenberg, H., Maurer, J., Singh, G., & Kochanek, K. (1993). Patterns in stroke mortality--United States. Annals of Epidemiology, 3(5), 458-465.

Ficke, R.C. (1992). Digest of Data on Persons with Disabilities 1992. Washington, DC: National Institute on Disability and Rehabilitation Research. January.

Geiss, L.S., Herman, W.H., Goldschmid, M.G., DeStefano, F., Eberhardt, M.S., Ford, E.S., German, R., Newman, J.M., Olson, D.R., Sepe, S.J., et al. (1993). Surveillance for diabetes mellitus--United States, 1980-1989. MMWR CDC Surveillance Summaries. 42(55.2) 1-20. June.

Gerhart, K.A., Bergstrom, E., Charlifue, S.W., Menter, R.R., & Whiteneck, G.G. (1993). The John Stanley Coulter Lecture. Health care reform and disability: affirming our commitment to community. Archives of Physical Medicine & Rehabilitation, 74(10). 1017-1024. October.

Grogaard, J.B., Lindstrom, D.P., Parker, R.A., Culley, B., & Stahlman, M.T. (1990). Increased survival rate in very low birth weight infants (1500 grams or less): no association with increased incidence of handicaps. Journal of Pediatrics, 117(1 Pt 1). 139-146. July.

Guralnick, J.M. (1985). Determinants of functional health status in the elderly. Ph.D. dissertation. Berkeley, CA: Dept. of Epidemiology, University of California, Berkeley.

Higgins, M. & Thomas, T. (1993). Trends in stroke risk factors. Annals of Epidemiology, 3(5):550-554, September.

Kaplan, G.A. (1992). Maintenance of functioning in the elderly. Annals of Epidemiology, 2(6), 823-34, November.

Klag, M.J. & Whelton, P.K. (1993). The decline in stroke mortality: An epidemiologic perspective. Annals of Epidemiology, 3(5), 571-575.

Klag, M.J., Whelton, P.K., & Seidler, A.J. (1989). Decline in U.S. stroke mortality: Demographic trends and antihypertensive treatment. Stroke, 20: 14-21.

Kovar, M.G., Pokras, R., & Collins, J.G. (1993). Trends in medical care and survival from stroke. Annals of Epidemiology. 3(5). 466-470. September.

Kraus, J.F. (1985). Epidemiological aspects of acute spinal cord injury: A review of incidence, prevalence, causes, and outcome. In D.P. Becker and J.T. Povlishock (Eds.), Central Nervous System Trauma Status Report. Bethesda, MD: National Institute of Neurological and Communicative Disorders and Stroke, National Institutes of Health.

Kunkel, S.R. & Applebaum, R.A. (1992). Estimating the prevalence of long-term disability for an aging society. Journal of Gerontology, 47(5): S253-60, September.

Kurland, L.T. (1994). Trauma and multiple sclerosis. Annals of Neurology, 36(Suppl). S33-S37.

Kurtzke, J.F. (1991). Multiple sclerosis: Changing times. Neuroepidemiology. 10(1).1-8.

LaPlante, M.P., Hendershot, G.E., and Moss, A.J. (1992). Assistive technology devices and home accessibility features: Prevalence, Payment, Need, and Trends. Advance Data, No. 217. National Center for Health Statistics, Centers for Disease Control, Public Health Service, US Department of Health and Human Services.

Mahoney, C.W., Estes, C.L., & Heumann, J.E. (1986). Toward a Unified Agenda: Proceedings of a National Conference on Disability and Aging. Wingspread Conference Center, Racine, WI.

Manton, K.G. (1991). The dynamics of population aging: demography and policy analysis. Milbank Quarterly. 62(2). 309-338.

Manton, K.G., Stallard, E., & Woodbury, M.A. (1991). A Multivariate Event History Model Based upon Fuzzy States: Estimation from Longitudinal Surveys with Informative Nonresponse. Journal of Official Statistics. Stockholm, Sweden.

Mehta, A.J. & Nastasi, A.E. (1993). Rehabilitation of fractures in the elderly. Clinics of Geriatric Medicine, 9(4), 717-30, November.

Melton, L.J., O'Fallon, W.M., & Riggs, B.L. (1987). Secular trends in incidence of hip fractures. Calcification Tissue International, 41:57-64.

Michaud, L.J., Duhaime, A.C., & Batshaw, M.L. (1993). Traumatic brain injury in children. Pediatric Clinics of North America. 40(3), 553-565. June.

Miles, T.P. & Bernard, M.A. (1992). Morbidity, disability, and health status of black American elderly: a new look at the oldest-old. Journal of the American Geriatrics Society, 40(10), 1047-54. October.

Miller, B., Prohaska, T., Mermelstein, R., & van Nostrand, J.F. (1993). Health data on older Americans: United States, 1992. Changes in functional status and risk of institutionalization and death. Vital and Health Statistics- Series 3, Analytical and Epidemiological Studies, 27: 41-75. January.

Moscato, B.S., Trevisan, M., & Willer, B.S. (1994). The prevalence of traumatic brain injury and co-occurring disabilities in a national household survey of adults. Journal of Neuropsychiatry & Clinical Neurosciences. 6(2). 134-142. Spring.

Mucci, S.J., Eriksen, L.D., Crist, K.A.; Bernath, L.A., & Chaudhuri, P.K. (1991). The pattern of injury to rear seat passengers involved in automobile collisions. Journal of Trauma. 31(10), 1329-1331. October.

National Center for Health Statistics. (1989). Health United States, 1988. Washington, DC: US Government Printing Office; March. DHHS Publication, no. 89-1232.

National Center for Health Statistics. (1990). Assistive Devices Supplement of the National Health Interview Survey, Series 10, No. 4. Center for Disease Control, Public Health Service, US Dept. of Health and Human Services.

National Center for Health Statistics. (1992). Health United States, 1991. Hyattsville, MD: Public Health Service.

National Head Injury Foundation. (1991). National Directory of Head Injury Rehabilitation Services. Washington, DC: Author.

Nicholson, A. & Alberman, E. (1992). Cerebral palsy--an increasing contributor to severe mental retardation? Archives of Disease in Childhood. 67(8), 1050-1055. August.

Osberg, J.S. & di Scala, C. (1992). Morbidity among pediatric motor vehicle crash victims: the effectiveness of seat belts. American Journal of Public Health. 82(3), 422-450. March.

Pharoah, P.O., Cooke, T., Cooke, R.W., & Rosenbloom, L. (1990). Birthweight specific trends in cerebral palsy. Archives of Disease in Childhood, 65(6), 602-606. June.

Price, C., Makintubee, S., Herndon, W., & Istre, G.R. (1994). Epidemiology of traumatic spinal cord injury and acute hospitalization and rehabilitation charges for spinal cord injuries in Oklahoma, 1988-1990. American Journal of Epidemiology, 139(1): 37-47, January.

Pope, A. M. & Tarlov, A. R. (Eds.) (1991). Disability in America: Toward a national agenda for prevention. Washington, DC: National Academy Press.

Rice, D.P. & LaPlante, M.P. (1992). Medical expenditures for disability and disabling comorbidity. American Journal of Public Health. 82(5), 739-741. May.

Rogers, R.G., Rogers, A., & Belanger, A. (1989). Active life among the elderly in the United States: Multistate life-table estimates and population projections. Milbank Memorial Fund Quarterly Health and Society, 67: 370-411.

Rosen, M.G. & Dickinson, J.C. (1992). The incidence of cerebral palsy. American Journal of Obstetrics & Gynecology. 167(2), 417-423. August.

Rutledge, R. & Stutts, J. (1993). The association of helmet use with the outcome of motorcycle crash injury when controlling for crash/injury severity. Accident Analysis & Prevention, 25(3), 347-353. June.

Sokolosky, M.C., Prescott, J.E., Collins, S.L., & Timberlake, G.A. (1993). Safety belt use and hospital charge differences among motor vehicle crash victims. West Virginia Medical Journal. 89(8), 328-330. August.

Sola, A. & Piecuch, R.E. (1994). Prevalence of cerebral palsy: estimations, calculations and neonatal care. Pediatrics. 93(1), 152-153. January.

U.S. Department of Health and Human Services Public Health Service. (1990). Vital and Health Statistics: Types of injuries by selected characteristics: United States, 1985-1987. Series 10: Data from the National Health Survey No. 175. Hyattsville, MD: Centers for Disease Control.

Verbrugge, L.M. (1987). From sneezes to adieux: Stages of health for American men and women. In Ward R.A. & Tobin S.S. (eds). Health in Aging: Sociological Issues and Policy Directions. New York: Springer, 17-57.

Verbrugge, L.M. & Ascione, F.J. (1988). Exploring the iceberg: Common symptoms and how people care for them. Medical Care, 25: 1020-1023.

Verbrugge, L.M., Gates, D.M., & Ike, R.W. (1991). Risk factors for disability among U.S. adults with arthritis. Journal of Clinical Epidemiology, 44(2): 167-82.

Verbrugge, L.M., Leplowski, J.M., & Konkol, L.L. (1991). Levels of disability among U.S. adults with arthritis. Journal of Gerontology, 46(2): S71-83. March.

Viano, D.C. (1992). Causes and control of spinal cord injury in automotive crashes. World Journal of Surgery, 16(3), 410-419. May-June.

Whelton, P.K. (1982). Declining mortality from hypertension and stroke. South Medical Journal, 75:33-8.

Whelton, P.K. & Klag, M.J. (1087). Recent trends in the epidemiology of stroke: What accounts for the stroke decline in western nations? Current Opinion in Cardiology, 2: 241-7.

Wilson, M.G., Michet, C.J., Ilstrum, D.M., & Melton, L.J. (1990). Idiopathic symptomatic osteoarthritis of the hip and knees: a population based incidence study. Mayo Clinic Proceedings, 65: 1214-1221.

Wolf, P.A. & D'Agostino, R.B. (1993). Secular trends in stroke in the Framingham Study. Annals of Epidemiology, 3(5), 471-475.

Zola, I.K. (1990). aging, disability and the home-care revolution. Archives of Physical Medicine and Rehabilitation, 71: 93-96.

Zola, I.K, (1993). Disability statistics: What we count and what it tells us. Journal of Disability Policy Studies, 4(2), 9-39.