Conceptual basis for multi-route intake dose modeling using ... - Nature

3 downloads 31 Views 387KB Size Report
methods, and the location of a person's home, job or school. (McCurdy ... Standard(s); NERL, National Exposure Research Laboratory; OAQPS,. Office of ... follow the first law of thermodynamics in conserving energy ..... Dr. William Nelson of NERL's Human Exposure Assess- ..... Walker B.S., Boyd W.C., and Asimov I., 1957.
Journal of Exposure Analysis and Environmental Epidemiology (2000) 10, 86 ±97 # 2000 Nature America, Inc. All rights reserved 1053-4245/00/$15.00

www.nature.com/jea

Conceptual basis for multi-route intake dose modeling using an energy expenditure approach1 THOMAS McCURDY National Exposure Research Laboratory, US Environmental Protection Agency, Research Triangle Park, North Carolina

This paper provides the conceptual basis for a modeling logic that is currently being developed in the National Exposure Research Laboratory ( NERL ) of the U.S. Environmental Protection Agency ( EPA ) for use in intake dose assessments involving substances that can enter the body via multiple routes of exposure. NERL is simultaneously developing a consolidated human activity database that will provide much of the data needed to parameterize equations used to implement the logic. EPA's Office of Air Quality Planning and Standards ( OAQPS ) is, in part, using the logic in its on - going review of the carbon monoxide national ambient air quality standard. Journal of Exposure Analysis and Environmental Epidemiology (2000) 10, 86 ± 97. Keywords: energy expenditure, exposure, intake dose, multi - route exposures.

Introduction There is a trend toward ``harmonization'' of scientific and analytic practices in human health risk assessment (Barton et al., 1998 ) . This harmonization has many dimensions, but the one of interest here is developing a consistent approach to estimating intake, or uptake, dose rates for multi - route exposure assessments. Most of these assessments to date have used independent rates for inhalation, ingestion, and dermal absorption. They often are derived using standard factors Ð or, more appropriately Ð distributions of these factors, such as those contained in the U.S. Environmental 1. Abbreviations: BM, body mass ( in kg ) ; BMR, basal metabolic rate ( kcal per time ) ; CHAD, Consolidated Human Activity Database; DBM, daily basal metabolism ( kcal day 1 ) ; DIT, dietary - induced thermogenesis; EE, energy expenditure ( kcal ) ; EPA, U.S. Environmental Protection Agency; EPOC, excess post - exercise oxygen consumption; FFM, fat - free mass; LN, log - normal distribution; MET ( S ) , metabolic equivalents of work; N, normal distribution; NAAQS, National Ambient Air Quality Standard ( s ) ; NERL, National Exposure Research Laboratory; OAQPS, Office of Air Quality Planning and Standards; OBLA, onset of blood lactate accumulation; PAI, physical activity index; SMR, sleeping metabolic rate; TDEE, total daily energy expenditure ( kcal day 1 ) ; U, Ç E, ventilation ( breathing ) rate ( l min 1 ) ; Vo2, uniform distribution; V Ç o2, oxygen consumption rate ( l min 1 ) ; VQ, oxygen consumption ( l ) ; V ventilatory equivalent. 2. Address all correspondence to: Thomas McCurdy, 1915 Hillock Place, Durham, NC 27712. Tel.: ( 919 ) 541 - 0782. Received 16 February 1998; accepted 30 April 1999. 1 Disclaimer: The information contained in this paper has been funded in part by the U.S. Environmental Protection Agency. It has been subjected to Agency review and approved for publication. Mention of trade names or commercial products does not constitute endorsement or recommendation for use.

Protection Agency's ( EPA's ) Exposure Factors Handbook (1997) or the Exposure Factors Sourcebook of the American Industrial Health Council (1994 ). However, it must be recognized that multi -route uptake rates are biologically interdependent within an individual. Sampling from independent distributions for the various routes ignores the correlated nature of an individual's metabolic and physiological processes. Doing so leads to a systematic negative bias in exposure/ dose and subsequent health risk assessments. This bias is most pronounced in ``high -end'' susceptible individuals who are at most risk to environmental pollutants. People may be susceptible due to intrinsic considerations, such as: genetic makeup, pre -existing disease, obesity or malnourishment, age, developmental stage of life, altered physiological state ( such as pregnancy ), or even emotional stress. Extrinsic considerations giving rise to risk include lifestyle and activity - related factors, such as: inactivity, hobbies, occupation, commuting methods, and the location of a person's home, job or school (McCurdy, 1997 ). In general, most exposure and intake dose assessments ignore these considerations. Even when considered, the assessments overestimate low -end and underestimate high - end exposures and intake doses (Ott et al., 1988; Law et al., 1997 ). Thus, there is a ``regression toward the mean'' bias in health risk assessments, and this bias is exacerbated for inherently interdependent multi route assessments. The trend toward explicit mathematical consideration of biological processes in health risk assessments is broad and pervasive. Emphasis currently is placed on pathway specific dosimetry, mechanistic mode -of -action modeling, physiologically based pharmacokinetic (and dynamic )

Multi-route intake dose modeling

modeling, and the explicit treatment of temporal considerations (O'Flaherty et al., 1995; Barton et al., 1998 ) . Since many biological processes are non- linear and dependent upon the prior pattern of dose received, a time series approach is required instead of an aggregate, integrated, or cumulative handling of time ( Opdam, 1991; Rappaport, 1993; McDonnell and Smith, 1994; Jarabek, 1995; Georgopoulos et al., 1997 ). This trend is due at least in part from knowledge gained about the etiology of xenobiotic substances and the realization that non- linear clearance, repair, and saturation kinetics largely govern target- organ impacts from a dose received ( Smith, 1991, 1992; Checkoway and Rice, 1992; Clayson and Iverson, 1996 ). Such kinetic mechanisms filter out both low and high intake doses before they have a chance to adversely affect a target organ. The pattern of intake dose received also affects subsequent target- organ impacts due to adaptation responses, which act to ``damp'' or mitigate subsequent higher doses, at least up to some dose level ( Rappaport and Spear, 1988 ); the pattern also affects elimination processes. Simultaneously, a lot of emphasis is being placed on modeling human variability in the biological and physiological processes being analyzed ( Hattis and Silver, 1994; Hattis et al., 1998 ). The inescapable conclusion drawn from these trends is that metabolically consistent, time -series, and correlated route - specific uptake rates are needed as input into biological dose -response models. This puts a heavy demand on model input data, but metabolism can become the systemic organizing principle for simplifying intake / uptake dose modeling. An important measure of metabolism is energy expenditure ( EE ). To understand how EE can be used to model multi -route intake dose, a concept first discussed by Layton ( 1993 ) for inhalation exposures, requires that we briefly explore relevant physiological concepts and terminology.

Energy expenditure Energy is the capacity for work. Human beings closely follow the first law of thermodynamics in conserving energy, and the energy released during cellular respiration is used to sustain biologic work. This is an exergonic (heatreleasing ) process that convertsÐ in the presence of oxygen Ð chemical ( potential ) energy stored in ingested glucose, fat, and protein molecules into mechanical or chemical work and the transport of fluids within the body. Enzymes and coenzymes aid in this conversion process. Physical activity involves complex cardiopulmonary, metabolic, musculoskeletal, neural, and physiological mechanisms. These relationships are influenced by a host of factors both endogenous and exogenous to the body, including noise, emotional stress, temperature, altitude, Journal of Exposure Analysis and Environmental Epidemiology (2000) 10(1)

McCurdy

physical fitness, genetics (including ethnicity ), disease status, and environmental factors. For a full discussion of these phenomena, and the complex neural, chemical, and gas exchange / transport kinetics of energy in the human body, see McArdle et al. (1991 ) or Whipp and Wasserman (1991) . Bodily chemical and metabolic changes caused by physical activity also affect the rate, pattern, and distribution of xenobiotic substances in the respiratory and circulatory systems. Total daily energy expenditure (TDEE ) in humans usually is disaggregated into the following categories: sleeping; resting (basal ) metabolism; dietary -induced thermogenesis ( DIT ) (heat produced from the digestion of food ); and the thermogenic effect of physical activity and ``recovery'' from strenuous physical activity (Elliot and Goldberg, 1985 ) . A general allocation of TDEE is depicted in Figure 1, which is a synthesis of information from many sources, especially Brozek and Grande (1955 ) and Richard and Rivest ( 1989 ) . Since these attributes are important in consistently modeling intake dose, we briefly discuss each one. For a good review of methods used to estimate human EE, see Roberts et al. ( 1986 ). The most accurate method used to measure EE on a daily basis is the ``doubly labeled water'' ( 2 H18 2 O ) technique, but direct and indirect calorimetry approaches also are used (McNeill et al., 1987 ). Basal Metabolism and the Basal Metabolic Rate ( BMR ) Basal metabolism is also known as ``resting energy expenditure'' or ``resting metabolism.'' ( Some authors differentiate between these terms, but the distinctions often have no practical difference in that the measurement protocols are similar. ) One definition of human basal metabolism is the energy requirement of a fasting body, at physical and mental rest and at 208CÐ a thermoneutral environment. It approximates the unavoidable loss of heat due to cell metabolism and the energy expended in maintaining minimal bodily functions: circulation, respiration, digestion, and involuntary muscle tone (Diem and Lentner, 1970 ) ; most basal energy is expended to keep the brain, liver, and skeletal muscles functioning properly. Basal metabolism is measured in units of kcal or kJ per time ( e.g., kcal h 1, MJ day 1 ) , and often is normalized (divided ) by body mass (BM, in kg) , fat - free mass ( FFM, in kg ), or body surface area (BSA, as m2 ). Scaling BMRÐ especially on a FFM basis Ðsignificantly reduces its variability in the overall population, and is used to compare rates between adults and children and between genders. BMR varies as a power function of BM (Waterlow, 1986 ). Probably, the best comparative BMR metric is per FFM 0.67 ( Winter, 1992; Nevill, 1994; Nevill and Holder, 1995; Rogers et al., 1995 ), but population distributions of FFM are difficult to obtain, so it will not be discussed further here. BMR also may be expressed as an oxygen consumption rate in units of milliliter O2 per time, and in 87

McCurdy

Multi-route intake dose modeling

Figure 1. General allocation of daily energy expenditure in an individual.

fact, BMR often is measured using oxygen consumption / carbon dioxide production relationships ( Short and Sedlock, 1997) . Finally, some authors define BMR to be equivalent to one metabolic equivalent of work ( MET; see below )3.5 ml O2 min 1 kg 10.0167 kcal min 1 kg 1, but this equivalency masks population variability in basal metabolism and will not be used in this paper. The variation in daily basal metabolism (DBM ) for an individual generally is on the order of ‹ 4 ±5% in children and adolescents, and ‹ 7± 8% in adults (Visser et al., 1995 ). Larger daily variations have been measured. The variation among individuals in a population of approximately the same age and gender is ‹ 30± 35% on a per BM basis. Most 88

of this variation is related to fitness and the amount and type of physical activities undertaken. Most of a person's TDEE is spent on basal metabolism. As Figure 1 indicates, 60 ±80% of a sedentary person's TDEE is basal. This proportion drops to 45± 60% in active people. Numerous prediction equations exist in the literature to estimate a person's BMR and/ or TDEE based upon anthropomorphic characteristics; a compilation of these equations is available from the author upon request. Sleeping For purposes of TDEE, sleeping normally is subsumed under basal metabolism. However, energy expended during Journal of Exposure Analysis and Environmental Epidemiology (2000) 10(1)

Multi-route intake dose modeling

McCurdy

sleep generally is less than that measured under BMR conditions. The sleeping metabolic rate (SMR ) is about 85 ±110% of BMR, and usually it is on the lower side of that range: 87% (Shah et al., 1988 ); 95% (Westerterp et al., 1992 ); 82 ±100% (Goldberg et al., 1988 ). SMR may be higher than BMR because of snoring and ``tossing and turning.'' The SMR -to -BMR ratio is highly reproducible for an individual, but varies widely among individuals.

25 times a person's basal rate. Another classification scheme focuses on what type of physical activity is undertaken, and common distinctions that are used are also shown in Figure 1. It should be noted that physical activity includes shivering and fidgeting, which have MET values of 3 ± 5 (Wong et al., 1996 ) . Very strenuous activity results in an ``oxygen debt'' when work continues anaerobically.

DIT: The Thermic Effect of Food DIT is between 2% and 10% of TDEE for active people and 10 ±17% for sedentary people (Tremblay et al., 1983; LeBlanc et al., 1984; Goran and Nagy, 1996; see Figure 1 ). BMR is elevated by up to 25% after a meal, the amount of DIT being non -linearly related to the caloric size of the meal (Hill et al., 1984 ). DIT has been monitored as long as 5 h after a meal (D'Alessio et al., 1988 ), with elevations of 5 ±8% being seen 3 h afterward (Belko et al., 1987 ). Daily variability in DIT is high in some individuals: ‹ 43% has been measured (Tataranni et al., 1995) , although values of ‹ 10% are more common. In general, DIT has low reproducibility and low reliability in EE monitoring studies.

During rest the energy requirement of the body is met by the breakdown of ATP [ adenosine triphosphate: an energy- rich compound that forms and conserves potential energy from food and transforms it into chemical energy for use in biological work ], and the supply of ATP is continuously restored by energy derived from the oxidation of foodstuffs (aerobic metabolism ) . During exercise the energy requirement may be increased many times. . .[ and this ] is reflected in an increased rate of oxygen consumption. . .If the maximal rate of oxygen consumption does not provide sufficient energy to restore ATP as rapidly as it is broken down, the balance of energy needed is derived from anaerobic metabolism, in which the pyruvic acid from in glycolysis is reduced to lactic acid instead of being oxidized to carbon dioxide and water. In exercise of this type, the rate of oxygen consumption remains elevated after the exercise is ended and returns gradually to normal during the recovery period. ( Morehouse and Miller, 1976; p. 121 )

Physical Activity and Recovery from Physical Activity Physical activity is also known as work -induced thermogenesis. It constitutes the largest source of variability in TDEE in an individual on a daily basis and within a population for any time period ( Hill et al., 1995 ). As depicted in Figure 1, physical activity (PA ) can vary from 5% to 50% of TDEE on a relative basis and between 50 and 1520 kcal day 1 on an absolute basis ( Schulz et al., 1989; Goran and Poehlman, 1992; Poehlman et al., 1992; Rising et al., 1994) . Difference of that magnitude result in TDEEs that can vary by a factor of two within the same age/gender / BM /height cohort ( Livingstone et al., 1991 ), and differences of that magnitude certainly affect intake dose received from inhaling air, drinking fluids, and eating food. In general, older people expend less energy in physical activity and females less than males at all ages ( Boot et al., 1997 ). Females and the elderly also spend less time per day in strenuous activities (Roberts et al., 1991, 1992; Sawaya et al., 1996 ) . In today's mechanized world, leisure time, or discretionary physical activity, burns more calories for most people than work- related activity. Physical activity often is disaggregated into categories. One classification system involves using relative intensity of work, such as light, medium, etc. (see Figure 1) . A more rigorous system classifies work into METS units. In the form used in this paper, a MET is the unitless ratio of an activity -specific metabolic rate to a person's basal rate. Usually METS ``cutpoints'' are used to distinguish among the classes, but there is little agreement among researchers on the exact values to use. Very strenuous or heavy exercise can have METS of 20 ± 25 (Wong et al., 1996 ) ; i.e., 20- to Journal of Exposure Analysis and Environmental Epidemiology (2000) 10(1)

This elevated oxygen consumption is now known as ``excess post - exercise oxygen consumption,'' or EPOC (Dawson et al., 1996 ). It was first measured in 1910. EPOC is needed to repay the oxygen debt incurred during exercise, which can be as high as 15± 20 l of oxygen (70 kcal ) (Stegemann, 1981 ) . As much as 20 ±30% of the energy requirement of an activity can occur after the activity ceases, so EPOC can be a significant component of the ventilatory demand of exercise (Bahr and Maehlum, 1986 ) . Not accounting for it will systematically underestimate the amount of intake dose received in activities that follow a strenuous activity event. There are fast and slow components of EPOC. The fast components involve oxidation of lactate, reduction of body temperature, mitochondrial respiration, and resynthesis of creatine phosphate. The slow components involve resynthesis of fatty acids, substrate cooling, and calcium homeostasis (Gaesser and Brooks, 1984 ) . The net impact of these interacting components causes EPOC to decline exponentially over time, and its effect has been measured 89

McCurdy

up to 48 h after a very strenuous bout of exercise (Ravussin and Bogardus, 1992 ). Even for less demanding activity, EPOC has been measured 12 ±16 h after exercise (Herring et al., 1992 ). EPOC increases greatly with exercise intensity. Finally, it is lower Ð and of shorter duration Ð for the same work load in trained ( active ) individuals than in ``normals'' (Short and Sedlock, 1997 ) . Retained Energy: Growth and Weight Gain If a person's daily EE is less than their energy intake, he or she experiences ``retained energy.'' Retained energy is used to support growth in children and adolescents, to build muscle in highly active persons, or to gain BM Ð mostly in the form of fat Ðin the rest of us. ( Weight loss is negative retained energy, but since hardly anyone loses weight after 30 it is ignored here! ) As Figure 1 indicates, the percentage of TDEE experienced as retained energy generally is small on a daily basis, particularly for adults. In young children, the energy cost of growth is 2 ±15%, and can be as high as 30% in babies ( Davies et al., 1994) . Unless weight change is explicitly considered in a modeling exercise, most intake dose assessments for adults should assume that retained energy is zero ( i.e., intake =expenditure ) . Otherwise, the modeler will have to address the complex dynamics of metabolism of a weight -changing individual vis - a- vis the metabolism of processing /assimilating xenobiotic substances. The non -steady state nature of these dynamics will affect the rate of storage, distribution, and elimination of these substances. Modeling this would be very resource intensive. Modeling intake dose using EE considerations Exercise physiologists usually calculate daily EE as the time spent in an activity multiplied by its metabolic cost, as METS or in terms of kcal kg 1 min 1; this is called the factorial method of calculating TDEE ( Spurr et al., 1997 ). The method proposed here to model multi -route intake dose utilizes this overall approach, and is depicted in Figure 2 ( discussed below ) . Successful implementation of the method requires that specific data be available to estimate an individual's basal metabolism, the METS associated with particular activities, and intake rate -to -EE metrics for each exposure route. Such data now exist or can readily be estimated for the most part. The remainder of this paper explains how such an approach can be implemented. The first topic is a brief discussion of the CHAD program being developed by EPA's National Exposure Research Laboratory (NERL ). NERL's CHAD This paper introduces CHAD to the modeling community, but is not intended to fully describe it or explain how it was 90

Multi-route intake dose modeling

developed; that is the subject of another paper ( McCurdy et al., in preparation ). CHAD is comprised of over 16,900 person - days of 24 h activity data. These data are available from eight large and two small human activity pattern surveys (Glen et al., 1997 ) . All ages and both genders are included in CHAD. The sequential pattern of activities undertaken by each subject is included, as is the location Ð microenvironmental classification Ðof each activity. Currently, 140 activities and 114 locations are coded in CHAD. Activities can vary in length from 1 min to 1 h; activities taking longer than 1 h are broken into segments that do not cross a clock hour. This is done to facilitate the calculation of hourly averaged intake dose estimates. If time periods longer than 1 h are modeled, the 1 -h segments simply are combined for the time interval of interest. It is possible, of course, to obtain 1- min dose estimates if needed, by assuming that a person's EE does not vary within a multi minute time block. Generally, however, a 1 -h dose estimate increment is sufficient. A distribution of METS values is associated with each activity (a ) in CHAD. The form of these distributions is specified from data gleaned during an extensive search of exercise and nutrition literature. A user of CHAD makes alternate passes through the data to obtain a distribution of METS estimates for each activity (METSa ) via Monte Carlo sampling. Thus, CHAD is organized explicitly to accommodate and foster uncertainty analyses. The prime source of METS estimates is Ainsworth et al. (1993) , a compendium of data developed to ``facilitate the coding of physical activities and to promote comparability of coding across studies'' ( p. 71 ). Hundreds of activities are included in the compendium. EPOC considerations are included in CHAD. Thus, METSa estimates for activities subsequent to a strenuous exercise event are modified upward to account for excess oxygen consumption and energy utilization. Also included in CHAD is an estimate of the BM of each individual ( i); delimited as BMi. Sometimes, these estimates are contained in the original surveys that comprise CHAD, but others are obtained from distributions contained in Burmaster and Crouch (1997 ). An example of one person -day of activity and the concomitant METS associated with each activity appears in Figure 3. Modeling Activities Using CHAD Estimating Basal Metabolism The proposed multi -route intake dose modeling procedure follows ( see Figure 2 ). For each day of interest, a particular individual is sampled from CHAD to represent a specific age / gender cohort, and his or her activities are used to represent what that individual did for the day. BMR for the person ( BMRi ) would be estimated using the BMi estimate in CHAD and a BMR regression equation that is found in the literature; for Journal of Exposure Analysis and Environmental Epidemiology (2000) 10(1)

Multi-route intake dose modeling

McCurdy

Figure 2. Using EE to estimate multi - route intake doses. Note: Elements available from Consolidated Human Activity Database ( CHAD ) are shaded.

example, see Schofield ( 1985 ). ( The author will supply a list of these for an interested reader.) Population variability in this estimate would be addressed by sampling from an error term ( e) distribution (xÅ = 0, = SD of the regression equation) and adding it to the predicted mean estimate of Journal of Exposure Analysis and Environmental Epidemiology (2000) 10(1)

BMR (BMR ) . The unit of BMRi needed is kcal min 1; if it is in kcal kg 1 min 1, it must be converted as shown below. Even though there is a small circadian rhythm in BMR, it probably can be ignored for most modeling efforts. 91

McCurdy

 BMRi ˆ BMR ‡ e  BMi

Multi-route intake dose modeling

 kcal min

1



…1†

Estimating EE for an Activity Activity -specific METSa are converted to activity -specific EE estimates for an individual (EEai ) as follows. The time spent in each activity is delineated as tai and also is part of CHAD. See Figure 3 for an example. EEai ˆ METSa  BMRi  tai

‰kcalŠ

…2†

Converting EE to Oxygen Consumption Individual activity - specific oxygen consumption (Vo2 ai ) estimated using a distribution of the appropriate conversion factor: 1

kcal = 200 ±210 ml O2. Although there is a gender bias to this factor ( males are on the high end and females on the low ), for the present a uniform distribution (U ) for the factor will be assumed for both genders. Vo2 ai ˆ …U : 200

210†  EEai

‰L O2Š

_ 2 ai from Vo2 ai, simply divide by tai. To obtain Vo    _ L O2 min 1 Vo2 ai ˆ Vo2 ai tai

…3†

…4†

Converting Oxygen Consumption Rate to Ventilation Rate Ç E ) Breathing rate (V Ç E) , also called minute ventilation, is (V the metric of interest for estimating intake rate of airborne

Figure 3. Example activity day information ( partial ) contained in CHAD for a particular 6 - year - old boy. 92

Journal of Exposure Analysis and Environmental Epidemiology (2000) 10(1)

Multi-route intake dose modeling

McCurdy

Ç E is contained in EPA's Exposure substances. Data on V Factors Handbook (National Center for Environmental Assessment, 1997 ), but an integrated dose approach needs _ 2 ai. Ç E estimates obtained directly from Vo V Exercise physiologists have been measuring the ratio _ 2 for years, and call it the ``ventilatory Ç E - to - Vo of V _ 2 have units of l Ç E and Vo equivalent'' ( VQ ). Since both V 1 1 min or l kg 1 min , VQ is mathematically unitless. In adults, the mean VQ is about 25 for exercise levels below which those that cause lactic acid to accumulate in the body Ðcalled the ``onset of blood lactate accumulation, '' or OBLA ( McArdle et al., 1991 ) . Above that level, VQ increases to 35 ±40 at the person's maximum level of _ 2 max ). The relationship oxygen consumption (called Vo between the two measures of ventilation follows a lognormal (LN ) distribution for an individual, even above the OBLA. However, there is a fairly wide distribution of VQ among individuals: the coefficient of variation ( COV ) of VQ is about 30%. Thus, a modeler should sample from a LN distribution of VQ in order to address population variability in this metric. Alternatively, the modeler could predict VQ from a regression equation _ 2. Ç E to Vo relating V VQ in children generally is slightly higher than in adults. Past National Ambient Air Quality Standard (NAAQS ) modeling efforts in EPA's Office of Air Quality Planning and Standards (OAQPS ) have used VQ in estimating inhalation intake dose, and the reader is referred to those documents for illustrative examples of an actual application ( Johnson et al., 1996a,b ). Ç Ei for each activity ( V Ç Eai ) is obtained thusly: V   _ Eai ˆ …LN : VQ ‡ e†  Vo _ 2 ai V L min 1 …5† Estimating Daily EE (Energy Intake) Daily total EE for an individual is obtained simply by summing all of the time weighted activity - specific estimates for the day. DTEEi ˆ  EEai

‰kcalŠ

…6†

for all activities in the day ( a= 1!n) . If retained energy is ignored, DTEEi provides a direct estimate of the amount of food needed to be consumed in order to sustain a person's daily activities. Thus, DTEEi is the estimated amount of energy intake for ingestion modeling purposes, and corresponds to clinical nutritionist's use of EE instead of intake because EE is a more reliable metric of actual consumption (Black et al., 1991, 1993 ). People systematically underestimate their energy intake, especially the obese ( Tarasuk and Beaton, 1991; Sawaya et al., 1996 ) . Estimating Food Ingestion The most reasonable time basis for dietary intake dose modeling probably is a day. Thus, the appropriate energy metric to use would be DTEEi. DTEEi could be divided into the main nutritional categories using Journal of Exposure Analysis and Environmental Epidemiology (2000) 10(1)

standard factors found in the literature. For example, Poehlman et al. ( 1990 ) state that energy intake of individuals is comprised as follows: Young

Old

Protein Fat

14 ‹ 4% 30 ‹ 6%

14 ‹ 2% 33 ‹ 6%

Carbohydrates

54 ‹ 9%

48 ‹ 6%

(The daily proportions do not total to 100% because they were independently derived.) There are age, gender, and ethnic differences in these proportions. Once the appropriate composition is obtained, it could be used for modeling purposes to provide an overall allocation to the general type and amount of food that a person might consume. In this manner, an individual's food ingestion estimate would be consistent with his or her EE and inhalation estimates. Finally, a dietary exposure model ( Berry, 1997; Tomerlin et al., 1997 ) and related information (e.g., Freeman et al., 1997 ) would be used to estimate specific foods ingested and, with some additional data, the amount of chemical residues consumed. Estimating Fluid Ingestion The usual approach to estimating water consumption in intake dose assessments is to base it on age and /or anthropometric factors. Distributional data on water consumption are available in Roseberry and Burmaster ( 1992 ) and National Center for Environmental Assessment (1997 ). Using this approach in isolation, however, does not maintain the correlated nature of water consumption and human activity that affects the rate of fluid consumed for any particular individual of interest. Water intake is in de minimus balance with EE and oxygen consumption ( Roberts et al., 1986 ). This means that there is a certain minimal level of water consumption that is required for metabolic processes, but more water can be consumed: a person's urinary output just increases. Water intake comes from food and fluids, in roughly equal proportions (43 ±62% fluids, the remainder via food ). Fluids include water, milk, alcohol, and flavored beverages. In addition to ingestion, some water needed to maintain bodily processes is produced within the body itself via metabolism of energy substrates; this is called ``metabolic water'' (Walker et al., 1957 ) . It usually comprises between 10% and 15% of total water needed to maintain life. The normal person is in a state of water equilibrium over a day, and water consumption varies with energy utilized. In fact, the preferred technique for estimating respiration and EE, the ``doubly labeled'' water method, is based upon the relationship measured between water metabolism and respiration (Ravussin and Bogardus, 1992 ) . However, this water balance state cannot be accurately established by just comparing fluids ingested and urine and sweat excreted. 93

McCurdy

Spit produced and evaporative losses via respiration and through the skin (both sweating and ``insensible water loss'') also contribute to water loss in the body. About 17 ± 24% of bodily water loss is through the respiratory system of an individual in equilibrium; it is about 0.13 ml H2O kcal 1 expended. About 21± 32% of water loss is through the skin, and the majority of it is in the form of sweat; it amounts to about 0.30 ml H2O kcal 1. Sweat rates can be as high as 2 l h 1 and 10± 12 l day 1 ( Askew, 1996 ). Fecal elimination uses up another 6 ± 10% of water intake, and urination constitutes the remainder (between 34% and 55% ). Similarly to the oxygen recovery concept noted above, any water debt incurred during activity or exercise must be restored afterward ( Maughan and Shirreffs, 1997 ). Water expenditure for adults is between 1.0 and 1.5 ml H2O kcal 1 and is between 1.0 and 2.9 ml H2O per kcal for babies (Wardlaw and Insel, 1993; Whitney et al., 1998 ). It is much higher for pregnant and lactating women. Hypohydration and voluntary dehydration during exercise are two considerations that compromise the water- to energy relationship (Armstrong et al., 1997 ). Nevertheless, relating water and fluid ingestion to energy consumption allows a modeler to correlate the two exposure routes in a multi - route assessment. The operational procedure for accomplishing this is similar to the one outlined above for food ingestion. The overall amount of water /fluid ingested is estimated based on EE, and additional information is required to allocate it to the types and amounts of beverages consumed. EPA's Exposure Factors Handbook contains such an allocation. Then, a dietary model would be used Ðalong with concentration data Ð to estimate the amounts of xenobiotic substances ingested via fluids. Dermal exposure/uptake The other major route of multi -route exposure /uptake dose is dermal absorption. This route is not amenable to being modeled using an EE approach, even those predicated upon unsteady- state mass flux principles ( Roy et al., 1996 ) . Data gathered related to dermal absorption generally are not directly placed into an activity context, but rely on passive absorption of a substance onto an ``indicator'' material or are gathered via mechanical sampling methods. For an individual in a microenvironment, dermal absorption probably is monotonically related to the person's physical activity in that location ( Cherrie and Robertson, 1995) . If nothing else, the amount of body surface area available for contact will increase with movement in a location, all other things being equal. Thus, dermal absorption is roughly proportional to EE, but there are no data on this relationship to model it with any confidence. 94

Multi-route intake dose modeling

Classifying individuals by EE metrics There is one additional use for EE metrics in intake dose assessment. It is to classify individuals into activity categories to focus in on groups that may be at particular risk because of their exercise patterns. For instance, OAQPS recently set a new ozone NAAQS to protect two active population sub -groups against daily ozone exposures. These groups were defined to be ``outdoor workers'' and ``outdoor children'' ( children who spend more than 2 h day 1 outside ). Previous analyses had shown these groups to be potentially more active Ð i.e., expend more energyÐ than most other people. OAQPS thus focused their modeling efforts on them (Johnson et al., 1996a ) . However, the classification scheme that was used in their analyses really focused on where a person was instead of what he or she did. An approach has been developed over the years by clinical nutritionists that allows a modeler to classify individuals by activity categories, and it has been incorporated into CHAD. Conceptually, it is straightforward: DTEEi is divided by total DBM (DBMi ) to develop a PAIi for each individual. PAIi ˆ DTEEi  DBMi

‰UnitlessŠ

…7†

where DTEEi is calculated as in Equation 6; DBMi = BMRi*1440 (the number of minutes day 1 ). The following schema, modified from clinical nutritional articles (Hoyt et al., 1991; Prentice, 1992; Cordain et al., 1998; Schoeller et al., 1997 ), can be used to classify the activity of individuals for modeling purposes:

Activity category

PAI

Sedentary

< 1.55

Low

1.55 ± 1.75

Moderate

1.75 ± 2.00

Active

> 2.00

Using these PAI cutpoints, the modeler can efficiently focus on a particular subgroup of interest from an activity viewpoint and just choose people from that group. Summary This paper explained how EE concepts can be used as a unifying theme for multi -route exposure /intake dose assessment. The concepts described here expand upon the seminal work by Layton ( 1993 ) and apply them to all routes of exposure, although the dermal absorption route may not be compatible with an EE approach. For the other routes, however, using EE allows a modeler to assess total exposure/ intake dose in a metabolically consistent manner. Although not discussed in this paper, special consideration in applying EE concepts must be given to obese Journal of Exposure Analysis and Environmental Epidemiology (2000) 10(1)

Multi-route intake dose modeling

individuals and to pregnant or lactating women. Energy- to intake relationships are dissimilar for these individuals, and different relationships should be utilized. This subject is being investigated further.

Acknowledgments This paper has benefitted from discussions on the topic with Dr. William Nelson of NERL's Human Exposure Assessment Branch. The anonymous peer reviewer provided excellent comments that hopefully resulted in an improved paper. References Ainsworth B.E., Haskell W.L., Leon A.S., Jacobs D.R., Jr., Montoye H.J., Sallis J.F., and Paffenbarger R.S., Jr. Compendium of physical activities: classification of energy costs of human physical activities. Med. Sci. Sports Exercise 1993: 25: 71 ± 80. American Industrial Health Council, 1994. Exposure Factors Sourcebook. AIHC, Washington, DC. Armstrong L.E., Maresh C.M., Gabaree C.V., Hoffman J.R., Kavouras S.A., Kenefick R.W., Castellani J.W., and Ahlquist L.E. Thermal and circulatory responses during exercise: effects of hypohydration, dehydration, and water intake. J. Appl. Physiol. 1997: 82: 2028 ± 2035. Askew E.W., 1996. Water. In: Ziegler E.E., Filer L.J., Jr. ( Eds. ) , Present Knowledge in Nutrition. ILSI Press, Washington, DC, pp. 98 ± 107. Bahr R., and Maehlum S. Excess post - exercise oxygen consumption: a short review. Acta Physiol. Scand. 1986: 128 ( suppl 556 ) : 99 ± 104. Barton H.A., Andersen M.E., and Clewell H.J., III. Harmonization: developing consistent guidelines form applying mode of action and dosimetry information to cancer and noncancer risk assessment. Hum. Ecol. Risk Assess. 1998: 4: 75 ± 115. Belko A.Z., Van Loan M., Barbieri T.F., and Mayclin P. Diet, exercise, weight loss, and energy expenditure in moderately overweight women. Int. J. Obes. 1987: 11: 93 ± 104. Berry R.B. Advances in dietary exposure research at the United States Environmental Protection Agency Ð National Exposure Research Laboratory. J. Exp. Anal. Environ. Epidemiol. 1997: 7: 3 ± 16. Black A.E., Goldberg G.R., Jebb S.A., Livingstone M.B.E., Cole T.J., and Prentice A.M. Critical evaluation of energy intake data using fundamental principles of energy physiology: 2. Evaluating the results of published surveys. Eur. J. Clin. Nutr. 1991: 45: 583 ± 599. Black A.E., Prentice A.W., Goldberg G.R., Jebb S.A., Bingham S.A., Livingstone M.B.E., and Coward W.A. Measurements of total energy expenditure provide insights into the validity of dietary measurements of energy intake. J. Am. Diet. Assoc. 1993: 93: 572 ± 579. Boot A.M., Bouquet J., De Ridder M.A.J., Krenning E.P., and Keizer Schrama S. Determinants of body composition measured by dual energy X - ray absorptiometry in Dutch children and adolescents. Am. J. Clin. Nutr. 1997: 66: 232 ± 238. Brozek J., and Grande F. Body composition and basal metabolism in man: correlation analysis versus physiological approach. Hum. Biol. 1955: 27: 22 ± 31. Burmaster D.E., and Crouch E.A.C. Lognormal distributions for body weight as a function of age for males and females in the United States, 1976 ± 1980. Risk Anal. 1997: 17: 499 ± 505.

Journal of Exposure Analysis and Environmental Epidemiology (2000) 10(1)

McCurdy

Checkoway H., and Rice C.H. Time - weighted averages, peaks, and other indices of exposure in occupational epidemiology. Am. J. Ind. Med. 1992: 21: 25 ± 33. Cherrie J.W., and Robertson A. Biologically relevant assessment of dermal exposure. Ann. Occup. Hyg. 1995: 39: 387 ± 392. Clayson D.B., and Iverson F. Cancer risk assessment at the crossroads: the need to turn to a biological approach. Regul. Toxicol. Pharmacol. 1996: 24: 45 ± 59. Cordain L., Gotshall R.W., Eaton S.B., and Eaton S.B., III. Physical activity, energy expenditure and fitness: an evolutionary perspective. Int. J. Sports Med. 1998: 19: 328 ± 335. D'Alessio D.A., Kavle E.C., Mozzoli M.A., Smalley K.J., Polansky M., Kendrick Z.V., Owen L.R., Bushman M.C., Boden G., and Owen O.E. Thermic effect of food in lean and obese men. J. Clin. Invest. 1988: 81: 1781 ± 1789. Davies P.S.W., Coward W.A., Gregory J., White A., and Mills A. Total energy expenditure and energy intake in the pre - school child: a comparison. Br. J. Nutr. 1994: 72: 13 ± 20. Dawson B., Straton S., and Randall N. Oxygen consumption during recovery from prolonged submaximal cycling below the anaerobic threshold. J. Sports Med. Phys. Fitness 1996: 36: 77 ± 84. Diem K., and Lentner C. ( Eds. ) , 1970. Scientific Tables. Geigy Pharmaceuticals, Ardsley, NY. Elliot D.L., and Goldberg L. Nutrition and exercise. Med. Clin. North Am. 1985: 69: 71 ± 82. Freeman N.C.G., Ettinger A., Berry M.R., and Rhoads G. Hygiene - and food - related behaviors associated with blood lead levels of young children from lead - contaminated homes. J. Exp. Anal. Environ. Epidemiol. 1997: 7: 103 ± 118. Gaesser G.A., and Brooks G.A. Metabolic bases of excess post - exercise oxygen consumption: a review. Med. Sci. Sports Exercise 1984: 16: 29 ± 43. Georgopoulos P.G., Walia A., Roy A., and Lioy P.J. Integrated exposure and dose modeling and analysis system: 1. Formulation and testing of microenvironmental and pharmacokinetic components. Environ. Sci. Technol. 1997: 31: 17 ± 27. Glen G., Lakkadi Y., Tippett J.A., and Del Valle - Torres M., 1997. Development of NERL / CHAD: The National Exposure Research Laboratory Consolidated Human Activity Database. ManTech Environmental Technology, Research Triangle Park, NC. Goldberg G.R., Prentice A.M., Davies H.L., and Murgatroyd P.R. Overnight and basal metabolic rates in men and women. Eur. J. Clin. Nutr. 1988: 42: 137 ± 144. Goran M.I., and Nagy T.R. Effect of the pre - testing environment on measurement of metabolic rate in children. Int. J. Obes. 1996: 20: 83 ± 87. Goran M.I., and Poehlman E.T. Total energy expenditure and energy requirements in healthy elderly persons. Metabolism 1992: 41: 744 ± 753. Hattis D., and Silver K. Human interindividual variability Ð a major source of uncertainty in assessing risk for noncancer health effects. Risk Anal. 1994: 14: 421 ± 431. Hattis D., Banati P., Goble R., and Burmaster D.E. Human interindividual variability in parameters related to health risks. Risk Anal. 1998: in press. Herring J.L., Mole P.A., Meredith C.N., and Stern J.S. Effect of suspending exercise training on resting metabolic rate in women. Med. Sci. Sports Exercise 1992: 24: 59 ± 65. Hill J.O., Heymsfield S.B., Mcmannus C., III, and Digorolamo M. Meal size and thermic response to food in male subjects as a function of maximum aerobic capacity. Metabolism 1984: 33: 743 ± 749. Hill J.O., Melby C., Johnson S.L., and Peters J.C. Physical activity and energy requirements. Am. J. Clin. Nutr. 1995: 62 ( suppl ) : 1059S ± 1066S. 95

McCurdy

Hoyt R.W., Jones T.E., Stein T.P., McAinch G.W., Lieberman H.R., Askew E.W., and Cymerman A. Doubly labeled water measurement of human energy expenditure during strenuous exercise. J. Appl. Physiol. 1991: 71: 16 ± 22. Jarabek A.M. Consideration of temporal toxicity challenges current default assumptions. Inhalation Toxicol. 1995: 7: 927 ± 946. Johnson T., Capel J., Mozier J.W., and McCoy M., 1996a. Estimation of Ozone Exposures Experienced by Outdoor Children in Nine Urban Areas Using a Probabilistic Version of NEM. International Technology, Durham, NC. Johnson T., Capel J., and McCoy M., 1996b. Estimation of Ozone Exposures Experienced by Urban Residents Using a Probabilistic Version of NEM and 1990 Population Data. IT Corporation, Durham, NC. Law P.L., Lioy P.J., Zelenka M.P., Huber A.H., and McCurdy T.R. Evaluation of a probabilistic exposure model applied to carbon monoxide ( pNEM / CO ) using Denver personal exposure monitoring data. J. Air Waste Manage. Assoc. 1997: 47: 491 ± 500. Layton D.W. Metabolically consistent breathing rates for use in dose assessments. Health Phys. 1993: 64: 23 ± 36. LeBlanc J., Mercier P., and Samson P. Diet - induced thermogenesis with relation to training state in female subjects. Can. J. Physiol. Pharmacol. 1984: 62: 334 ± 337. Livingstone M.B.E., Strain J.J., Prentice A.M., Coward W.A., Nevin G.B., Barker M.E., Hickey R.J., McKenna P.G., and Whitehead R.G. Potential contribution of leisure activity to the energy expenditure patterns of sedentary populations. Br. J. Nutr. 1991: 65: 145 ± 155. Maughan R.J., and Shirreffs S.M. Recovery from prolonged exercise: restoration of water and electrolyte balance. J. Sports Sci. 1997: 15: 297 ± 303. McArdle W.D., Katch F.I., and Katch V.L., 1991. Exercise Physiology, 3rd edn. Lea & Febiger, Philadelphia. McCurdy T. Modeling the dose profile in human exposure assessments. Rev. Toxicol.: In Vivo Toxicol. Risk Assess. 1997: 1: 3 ± 23. McDonnell W.F., and Smith M.V. Description of acute ozone response as a function of exposure rate and total inhaled dose. J. Appl. Physiol. 1994: 76: 2776 ± 2784. McNeill G., Cox M.D., and Rivers J.P.W. The Oxylog consumption meter: a portable device for measurement of energy expenditure. Am. J. Clin. Nutr. 1987: 45: 1415 ± 1419. Morehouse L.E., and Miller A.T., 1976. Physiology of Exercise, 7th edn. C.V. Mosby Company, Saint Louis. National Center for Environmental Assessment, 1997. Exposure Factors Handbook: I. General Factors. U.S. Environmental Protection Agency, EPA / 600 / P - 95 / 002Ba, Washington, DC. Nevill A.M. The need to scale for differences in body size and mass: an explanation of Kleiber's 0.75 mass exponent. J. Appl. Physiol. 1994: 77: 2870 ± 2673. Nevill A.M., and Holder R.L. Scaling, normalizing and per ratio standards: an allometric modeling approach. J. Appl. Physiol. 1995: 79: 1027 ± 1031. O'Flaherty E.J., Polak J., and Andriot M.D. Incorporation of temporal factors into physiologically based kinetic models for risk assessment. Inhalation Toxicol. 1995: 7: 917 ± 925. Opdam J.J.G. Linear systems dynamics in toxicokinetic studies. Am. Occup. Health 1991: 35: 633 ± 649. Ott W.R., Thomas J., Mage D., and Wallace L. Validation of the simulation of human activity and pollutant exposure ( SHAPE ) model using paired days from the Denver, CO carbon monoxide field study. Atmos. Environ. 1988: 22: 2101 ± 2113. Poehlman E.T., McAuliffe T.L., Van Houten D.R., and Danforth E., Jr. Influence of age and endurance training on metabolic rate and hormones in healthy men. Am. J. Physiol. 1990: 259: E66 ± E72. 96

Multi-route intake dose modeling

Poehlman E.T., Berke E.M., Joseph J.R., Gardnero A.W., Katzman - Rooks S.M., and Goran M.I. Influence of aerobic capacity, body composition, and thyroid hormones on the age - related decline in resting metabolic rate. Metabolism 1992: 41: 915 ± 921. Prentice A.M. Energy expenditure in the elderly. Eur. J. Clin. Nutr. 1992: 46 ( suppl 3 ) : S21 ± S28. Rappaport S.M. Biological considerations in assessing exposures to genotoxic and carcinogenic agents. Int. Arch. Occup. Environ. Health 1993: 65: S29 ± S35. Rappaport S.M., and Spear R.C. Physiological damping of exposure variability during brief periods. Ann. Occup. Hyg. 1988: 32: 21 ± 33. Ravussin E., and Bogardus C. A brief overview of human energy metabolism and its relationship to essential obesity. Am. J. Clin. Nutr. 1992: 55: 242S ± 245S. Richard R.D., and Rivest S. The role of thermogenesis and energy balance. Can. J. Physiol. Pharmacol. 1989: 67: 402 ± 409. Rising R., Harper I.T., Fontvielle A.M., Ferraro R.T., Spraul M., and Ravussin E. Determinants of total daily energy expenditure: variability in physical activity. Am. J. Clin. Nutr. 1994: 59: 800 ± 804. Roberts S.B., Coward W.A., Schlingenseipen K. - H., Nohria V., and Lucas A. Composition of the doubly labeled water ( 2H218O ) method with indirect calorimetry and a nutrient - balance study for simultaneous determination of energy expenditure, water intake, and metabolizable energy intake in preterm infants. Am. J. Clin. Nutr. 1986: 44: 315 ± 322. Roberts S.B., Heyman M.B., Evans W.J., Fuss P., Tsay R., and Young V.R. Dietary energy requirements of young adult men, determined by using the doubly labeled water method. Am. J. Clin. Nutr. 1991: 54: 499 ± 505. Roberts S.B., Young V.R., Fuss P., Heyman M.B., Fiatarone M., Dallal G.E., Cortiella J., and Evans W.J. What are the dietary needs of elderly adults? Int. J. Obes. 1992: 16: 969 ± 976. Rogers D.M., Olson B.L., and Wilmore J.H. Scaling for the VO2 - to - body size relationship among children and adults. J. Appl. Physiol. 1995: 79: 958 ± 967. Roseberry A.M., and Burmaster D.E. Lognormal distributions for water intake by children and adults. Risk Anal. 1992: 12: 99 ± 104. Roy A., Weisel C.P., Lioy P.J., and Georgopoulos P.G. A distributed parameter physiologically - based pharmacokinetic model for dermal and inhalation exposure to volatile organic compounds. Risk Anal. 1996: 16: 147 ± 160. Sawaya A.L., Tucker K., Tsay R., Willett W., Saltzman E., Dallal G.E., and Roberts S.B. Evaluation of four methods for determining energy intake in young and older women: comparison with doubly labeled water measurements of total energy expenditure. Am. J. Clin. Nutr. 1996: 63: 491 ± 499. Schoeller D.A., Shay K., and Kushner R.F. How much physical activity is needed to minimize weight gain in previously obese women? Am. J. Clin. Nutr. 1997: 66: 551 ± 556. Schofield W.N. Predicting basal metabolic rate, new standards, and review of previous work. Hum. Nutr.: Clin. Nutr. 1985: 39C ( suppl 1 ) : 5 ± 41. Schulz S., Westerterp K.R., and Bruck K. Comparison of energy expenditure by the doubly labeled water technique with energy intake, heart rate, and activity recording in man. Am. J. Clin. Nutr. 1989: 49: 1146 ± 1154. Shah M., Miller D.S., and Geissler C.A. Lower metabolic rates of post obese versus lean women: thermogenesis basal metabolic rate and genetics. Eur. J. Clin. Nutr. 1988: 42: 741 ± 752. Short K.R., and Sedlock D.A. Excess postexercise oxygen consumption and recovery rate in trained and untrained subjects. J. Appl. Physiol. 1997: 83: 153 ± 159. Smith T.J. Pharmacokinetic models in the development of exposure indicators in epidemiology. Am. Occup. Hyg. 1991: 35: 543 ± 560. Smith T.J. Occupational exposure and dose over time: limitations of cumulative exposure. Am. J. Ind. Med. 1992: 21: 35 ± 51. Journal of Exposure Analysis and Environmental Epidemiology (2000) 10(1)

Multi-route intake dose modeling

Spurr G.B., Dufour D.L., Reina J.C., and Haught T.A. Daily energy expenditure of women by factorial and heart rate methods. Med. Sci. Sports Exercise 1997: 29: 1255 ± 1262. Stegemann J., 1981. Exercise Physiology. Year Book Medical Publishers, Inc., Chicago. Tarasuk V., and Beaton G.H. The nature and individuality of within subject variation in energy intake. Am. J. Clin. Nutr. 1991: 54: 464 ± 470. Tataranni P.A., Larson D.E., Snitker S., and Ravussin E. Thermic effect of food on humans: methods and results from use of a respiratory chamber. Am. J. Clin. Nutr. 1995: 61: 1013 ± 1019. Tomerlin J.R., Berry M.R., Tran N.L., Chew S. - B., Petersen B.J., Tucker K.D., and Fleming K.H. Development of a dietary exposure potential model for evaluating dietary exposure to chemical residues in food. J. Exp. Anal. Environ. Epidemiol. 1997: 7: 81 ± 101. Tremblay A., Cote J., and LeBlanc J. Diminished dietary thermogenesis in exercise - trained human subjects. Eur. J. Appl. Physiol. 1983: 52: 1 ± 4. Visser M., Deurenberg P., Van Staveren W.A., and Hautvast J.G.A.J. Resting metabolic rate and diet - induced thermogenesis in young and elderly subjects: relationship with body composition, fat

Journal of Exposure Analysis and Environmental Epidemiology (2000) 10(1)

McCurdy

distribution, and physical activity level. Am. J. Clin. Nutr. 1995: 61: 772 ± 778. Walker B.S., Boyd W.C., and Asimov I., 1957. Biochemistry and Human Metabolism. The Williams & Wilkins Co., Baltimore. Wardlaw G.M., and Insel P.M., 1993. Perspectives in Nutrition. Mosby Publications, St. Louis. Waterlow J.C. Notes on the new international estimates of energy requirements. Proc. Nutr. Soc. 1986: 45: 351 ± 360. Westerterp K.R., Meuer G.A.L., Janssen E.M.E., Saris W.H.M., and Ten Hoor F. Long - term effect of physical activity on energy balance and body composition. Br. J. Nutr. 1992: 68: 21 ± 30. Whipp B.J., and Wasserman K. ( Eds. ) , 1991. Exercise. Marcel Dekker, New York. Whitney E.N., Cataldo C.B., and Rolfes S.R., 1998. Understanding Normal and Clinical Nutrition. West / Wadsworth, Blemont, CA. Winter E.M. Scaling: partitioning out differences in size. Pediatr. Exercise Sci. 1992: 4: 296 ± 301. Wong W.W., Butte N.F., Hergenroeder A.C., Hill R.B., Stuff J.E., and Smith E.O'B. Are basal metabolic rate prediction equations appropriate for female children and adolescents? J. Appl. Physiol. 1996: 81: 2407 ± 2414.

97