| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
National Cancer Institute (G.B., B.H.P., P.R.T.), Bethesda;
USDA Agricultural Research Center, Beltsville (A.R.M., O.A.L.), Maryland;
Our Lady of Mercy Medical Center, Bronx, New York (E.P.N.)
Address reprint requests to: Gladys Block, PhD, 426 Warren Hall, University of California, Berkeley, CA 94720.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Methods: After one month of stabilization on 60 mg vitamin C/day, subjects underwent two complete depletion-repletion cycles (one cycle=one month of vitamin C depletion with nine mg/day, followed by one month of repletion with 117 mg per day). Subjects (68 men, ages 30 to 59 years) did not smoke or drink alcohol during the study. All food was provided by the study.
Results: There was extreme individual variability in the plasma AA level achieved on an identical repletion dose: after four weeks at 117 mg/day of vitamin C, AA ranged from 26.8 µmol/L to 85.8 µmol/L. Body weight was inversely associated with plasma AA attained (p<0.0001). Regression analysis indicated that, compared to a 130-lb man, a 200-lb man reached 10 µmol/L lower AA after the first repletion and 18 µmol/L lower AA after the second repletion. One-third of the subjects did not reach a plasma plateau after the first repletion. Prior depletion and apparent repletion also had a major impact, and only 10% of subjects reached the same plasma AA on the second repletion as on the first repletion.
Conclusions: Plasma AA attained on a given dose depends on body weight (or dose per kg of body weight) and on whether or not any prior depletions had been repleted adequately. The results have implications for nutrition recommendations and research design.
Key words: ascorbic acid, vitamin C, nutrient requirements, obesity
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
A number of investigators have examined vitamin C requirements by feeding subjects ascorbic acid-deficient diets and then feeding varying amounts of vitamin C. Observations in these studies pertained to the point at which deficiency symptoms were seen and the amount of vitamin C that caused the regression of those symptoms [3–9] or to the plasma plateau achieved on a given dose [10]. The largest of these studies included only 11 subjects. One study [8] included a partial second depletion-repletion cycle, and in that study the second repletion was for only one week, with nine subjects on two different doses. All of these studies found substantial variability in plasma level attained on any given dose, but had an insufficient number of subjects to investigate the extent or sources of individual variability in dose response. The only previous depletion-repletion study to include a substantial number of subjects was that of Blanchard et al. [11], who depleted 29 young and 29 elderly subjects and then repleted them with 500 mg/day.
The study reported here, with 68 subjects, provides more data on and quantification of sources of individual variability in plasma levels attained on nonpharmacologic doses of vitamin C. In addition, inclusion of a second complete depletion-repletion cycle provides the opportunity to examine whether conclusions about dose response that might be reached after a single depletion and repletion would be different after a second depletion-repletion cycle.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
Rational for Dosage Amounts.
The purpose of the study was to study the effects of physical factors and of different food sources on blood levels attained with dietary levels of vitamin C. Average vitamin C intake in the U.S. is typically in the 100 to 120 mg/day range. Since one of the factors we wanted to investigate was whether vitamin C obtained from a supplement produced the same blood level as the same amount obtained from fruit or vegetables, we first sought a commercial vitamin C supplement that provided 100 mg. After assaying several commercially available tablets of nominally 100 mg, we found one that was close to that amount and provided 108 mg (others were higher). This 108 mg, added to the depletion diet providing nine mg/day, resulted in a target daily amount of 117 mg/day for the Supplement Group. Diets in the Fruit and Vegetables Groups thus were also designed to provide a total daily dose of 117 mg/day during repletion.
Subjects.
Participation was limited to healthy men 30 to 59 years of age, who had not smoked for at least six months. Subjects were recruited from the greater Beltsville, Maryland, area through advertisements in newspapers and local places of employment. Seventy-one volunteers enrolled, and 68 completed the entire study.
Diets.
Subjects were fed a 14-day rotating menu composed of foods common in the U.S. diet, providing approximately 50% of energy from carbohydrate and 36% from fat [1]. Subjects were weighed weekly, and energy intake was adjusted to prevent weight change during the study. On weekdays, subjects consumed breakfasts and dinners at the study facility. Lunches and snacks were provided as bag lunches; weekend meals were provided, frozen and packed in coolers, and consisted of the same types of hot meals which were provided on weekdays. This type of controlled-diet study is identical to the approach used by the Dietary Approaches to Stop Hypertension (DASH) study sponsored by the National Heart, Lung and Blood Institute [12].
During the depletion periods, the menus were nutritionally adequate except for vitamin C. During repletion, subjects randomized to the Supplement Group ate the same depletion diet, with vitamin C added back in pill form. In the Vegetable or Fruit Groups, foods containing vitamin C were added to the depletion diet. The AA content of the repletion foods was analyzed twice weekly by HPLC, and the amount fed was adjusted to maintain a constant AA intake over the repletion periods and in all study groups.
Compliance.
Subjects were required to consume all food provided and were closely monitored by study staff during meals; any uneaten lunch or weekend food was returned and weighed. Calculations at the end of the study indicated that uneaten food was negligible and had no effect on average vitamin C intake. A daily checklist report of protocol violations and a final, post-reimbursement anonymous questionnaire revealed that reported dietary omissions or commissions were minor and had no effect on average energy, vitamin C or other nutrient intake. These monitoring procedures indicated excellent compliance with the dietary regimen. These compliance measures and conclusions are similar to those reported by the DASH study [12], which used a comparable feeding study design.
Data Collection.
Information was obtained at baseline about usual diet, physical activity, prior smoking history and current passive smoke exposure, stress and demographic characteristics. The relationship between plasma nutrient values at baseline and questionnaire nutrient estimates has been reported elsewhere [13,14]. During the controlled diet portion of the study, an event checklist was completed by the subjects daily, on which they indicated whether they had been sick, taken medications, engaged in physical exercise or experienced environmental or social stress. Subjects were weighed weekly. Body mass index (weight/height2) was calculated, and fat mass and lean mass were measured by bioelectric impedance.
The factors obtained from the daily checklist (sickness, medication use, operation or dental work, exposure to environmental stress, experience of social/psychological stress, vigorous physical activity that day) were tallied and a score obtained for each study period. For example, a separate score was calculated for number of days of vigorous physical activity during each depletion and each repletion period. These derived variables were then examined as covariates in multiple regression analyses.
Plasma Analyses.
Venous blood was obtained from fasting subjects, at approximately the same time of day, prior to breakfast. It was obtained at baseline, biweekly during the depletion periods and weekly during the repletion periods. Blood was collected in vacutainers containing EDTA, iced and centrifuged within 30 minutes. Plasma was stabilized with freshly prepared 10% meta-phosphoric acid, vortexed, stored at -70°C and analyzed within 10 days of collection. Plasma ascorbic acid concentration was determined spectrophotometrically using 2,4-dinitrophenylhydrazine [15], which has been shown to correlate highly with HPLC methods [16–19]. Ascorbic acid is presented here in mg/dL as well as in µmol/L, for comparison with earlier data.
Statistical Analyses.
Statistical analyses were conducted using SAS and included t tests and multiple regression analyses, with significance at p<0.05. For some analyses, a measure of integrated tissue exposure to AA over a time period was estimated for each individual as the sum of the plasma concentrations at each week of repletion. We refer to this as "area under the plasma repletion curve."
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Depletion on Nine mg/Day
Plasma AA fell from a mean of 39.8 µmol/L (0.70 mg/dL) at the beginning of depletion to a mean of 12.2 µmol/L (0.21 mg/dL) after one month on nine mg/day during the first depletion cycle and to 13.4 µmol/L (0.24 mg/dL) during the second cycle (Table 1). It was not our intention to reduce subjects to scorbutic levels, and no symptoms of scurvy were observed.
There was considerable variability in the AA level to which subjects fell on the depletion dose, with individual respondent values ranging from 13.6 to 29 µmol/L (0.24 to 0.51 mg/dL). The plasma AA value at the beginning of depletion was positively associated with the level reached on depletion (r=0.52). Excluding initial plasma AA level, total body weight was highly significant and a slightly stronger predictor of the extent of the fall in AA than was lean body mass. In multiple regression analysis, no other factors affected the extent of the fall in AA on depletion, including age, BMI, reported physical activity or use of vitamin supplements prior to the start of the study.
Repletion on 117 mg/Day
At the end of the Cycle 1 repletion, mean plasma AA reached 59.1 µmol/L (1.04 mg/dL); after the Cycle 2 repletion, mean plasma AA reached only 48.6 µmol/L (0.86 mg/dL). Both mean AA at the end of repletion and area under the repletion curve were significantly different between Cycle 1 and Cycle 2 as evaluated by paired t tests (p<0.0001) (data not shown). In Cycle 2 only 10% of subjects equaled or exceeded the plasma AA levels they reached on the first repletion.
The vitamin C dose per kg of body weight was strongly associated with extent of repletion as measured by total area under the repletion curve (AUC) (Fig. 1). This association (r=0.66) was stronger than that between dose per kg of lean body mass and AUC (r=0.57) or that between dose per kg of body weight and AA reached at the end of the first repletion (r=0.42) (data not shown).
|
|
 |
 |
where Area1 and Area2 are the areas under the plasma AA repletion curve for Cycles 1 and 2, respectively; AA5 and AA13 are the plasma AA values at Weeks 5 and 13, the beginning of each repletion cycle; and dosekg is the vitamin C intake per kg of body weight. In both cycles, intake per kg of body weight was significant at p<0.0001. When body weight is substituted for intake per kg of body weight in the above models, a negative coefficient but an essentially identical R2 is achieved, as would be expected, since they are simply the same variable (body weight) multiplied by a constant (117 mg/day). In the context of a study such as this, in which vitamin C intake was the same among all subjects, body weight may be interpreted as an effect modifier of the relationship between intake and resulting plasma repletion.
Based on these regression models, estimated plasma levels that would be attained after various time intervals in persons with various body weights and depletion histories were calculated (Table 3). After two weeks of 117 mg vitamin C per day, the 59 kg (130-lb) subject will have achieved a plasma AA value of 60.8 µmol/L (1.07 mg/dL), while the 91 kg (200-lb) subject will have achieved only 36.4 µmol/L (0.64 mg/dL). Table 3 also indicates that the plasma AA value attainable at a given dose is not only a function of the dose and body weight (or dose per kg of body weight), but also a function of the pre-existing depletion experience and presumably, therefore, of tissue status. In the second cycle, subjects achieved a lower plasma AA and took approximately two weeks longer to reach the AA level that they had achieved in Cycle 1. For example, subjects in the 59 kg group reached 60.8 µmol/L after two weeks in the first repletion, and took four weeks to reach 59.1 µmol/L in the second repletion (Table 3).
|
|
However, a great deal of individual variability in response to a given dose remains, even within body-weight groups. Figure 2c shows the variability within a single body-weight quartile. There was approximately a two-fold range in the plasma AA value achieved on repletion, within a single-body weight quartile. For example, in Body Weight Quartile 4, the value at the end of the first repletion ranged from 34.0 to 70.7 µmol/L, and the area under the repletion curve ranged from 72.8 to 162.0 µmol/L.
To explore the effect of such variability on results of studies with small sample sizes, we drew nine random samples of seven subjects each, and reexamined our data. These "studies" of seven subjects produced widely differing conclusions. For example, in one sample, subjects achieved a plasma AA of at least 37.5 µmol/L after one month on 117 mg/day, while in a different seven-person sample, subjects achieved a plasma AA of at least 68.7 µmol/L. The standard deviations in these nine samples ranged from 2.8 µmol/L to 17.0 µmol/L (data not shown).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The level of plasma AA reached after one month of depletion is consistent with the results of other investigators [3,6,8,11,20]. The depletion results confirm those of Kallner et al. [21], that initial plasma level is the principal factor determining how far plasma levels will fall on depletion and that body weight is the only other influential factor.
The extreme variability observed in these data after repletion with an identical, fairly substantial dose (117 mg/day) can be seen in Figure 2, where levels ranged from 26.8 to 85.8 µmol/L after the first repletion and from 16.9 to 75.6 µmol/L after the second repletion. It is also seen in the large SDs after each repletion (Table 1), larger still after the second than after the first. This suggests that one month at 117 mg/day was insufficient to fill body pools and allow subjects to reach a plasma plateau. This variability was seen despite the fact that this was a relatively homogeneous study group: it did not include women, men in their 20s or over age 60, poor people, smokers, alcohol consumers or people with common chronic diseases such as high blood pressure, elevated cholesterol, diabetes, heart disease or the like.
The variability in plasma level attained on a given dose has also been observed by other investigators [4,10,21]. Levine et al. [10] found that the ultimate plasma plateau ranged from 14.9 to 58.8 µmol/L on an intake of 60 mg of vitamin C per day, and from 57.1 to 75.1 µmol/L on 200 mg/day, in seven young men. Unlike the present study, Levine et al. [10] continued each dose until a plasma plateau was achieved. They found that the time required to achieve a plateau also varied among the seven subjects.
The failure of 25% of our subjects to reach plasma levels of 40 µmol/L (the lower level of normal) in the second repletion of one month on 117 mg/day was unexpected. It suggests that repeated depletions may lead to an increasing impoverishment of body pool sizes and a widening gap in plasma AA levels between persons who have and have not experienced such repeated depletions. Previous research has not explored the effect of repeated depletions and the time and dose required to fully replete subjects.
This study quantifies the effect of body weight on the plasma level attained on a given dose. It indicates that on dietary intakes in this range, a heavy subject may attain a plasma AA level as much as 10 µmol/L lower than that of a light subject after a single depletion and as much as 18 µmol/L lower after a second depletion (Table 3). Others have observed the association with body weight, most often in the context of differences between men and women in plasma level [8,11,13,21–29]. Some authors have suggested a role for ovarian hormones or sex differences in renal absorption to explain observed gender differences in plasma concentration [30]. Others have suggested that gender differences, as well as the variability in plasma level in a single gender, may be due to differences in lean body mass or fat free mass [4,11,31,32]. The present data support the latter explanation: the gender effect, and variability in attained dose within a gender, may be substantially related to differences in body weight (a given dose in a smaller volume producing a higher concentration). In the present study we examined both total body weight and lean body mass and found similar results for both, perhaps because this was a single-gender study with relatively fewer differences in percent of body fat. Weight is used here because it is more commonly understood and measured. In addition, the multivariate analyses suggest that apparent age effects may be attributable to the weight gain that often accompanies aging, rather than to effects of aging per se (at least in the age range 30 to 59).
While it is clear from the present study that body weight is an important determinant of plasma level, and one that influences both the AA level attained on any given dose and the time required to reach that level, substantial unexplained sources of variability remained. Additional sources of variability in response may include the presence of several compartments in addition to plasma, varying degrees of tissue binding in those compartments [21] and genetic differences. A role for individual variations in emotional stress level [4] is not supported in the present data, nor are factors such as physical stress, minor illnesses or physical activity level. However, a role for such factors also cannot be ruled out; there may have been insufficient variability in those factors among persons in this sample, and a larger study and one designed specifically to answer such questions would be required. In addition, unmeasured environmental exposures could have contributed to differences in plasma level; however, we did obtain more information on a daily basis than had any previous research, including daily reports on passive smoke exposure and environmental stress, but were unable to identify factors that influenced the resulting blood levels.
In summary, these analyses have shown that the plasma AA that will be attained depends on several factors: the daily intake (compare the 9 mg, 60 mg and 117 mg levels in Table 1), body weight or dose per kg of body weight (Table 2 and Table 3), tissue or body pool status, reflected here in the range of levels reached on depletion and in the effect of previous depletion experience (compare AA levels after repletions one and two, in Table 1), duration of repletion (note incomplete repletion and absence of a plateau after one month, in Cycle 2 [Fig. 2b]) and unknown individual factors (Fig. 2a and 2c).
Several of these results may have implications for health policy and nutrition recommendations, clinical practice and research design. The role of body weight in partially explaining attained plasma level might be considered in making public health recommendations. In national survey data, approximately one-third of U.S. adults are overweight, and approximately one-third of men in this age range in the United States have plasma AA values below 40 µmol/L [22], sometimes considered the lower level of "normal" AA [33,34]. These data suggest that any recommended intake expressed in mg/day will result in lower blood AA levels in men than in women and lower blood AA levels in overweight than in normal-weight persons. The same would be true for recommendations regarding numbers of servings of fruits and vegetables. The extreme variability in response to a given dose of AA in repletion (Fig. 2a), over and above that explained by body weight, also supports the idea that it may be useful to express health-related recommendations in terms of desirable plasma concentrations, rather than in terms of milligrams of intake per day. The great advances in prevention of heart disease are due in part to campaigns to "know your (serum cholesterol) number" and guidelines that medical care practitioners can follow to identify blood levels in the desirable range.
At present there is no consensus on the importance of AA level in health promotion (although suggestive data exist [2,35–39]) and insufficient information on which to base health recommendations expressed in terms of plasma concentrations (although again, suggestive data exist [2]). However, such information can be amassed, if epidemiologic researchers, health surveys and clinicians routinely collect plasma AA data. In the face of variability in plasma response to intake, a reliance on dietary intake measures to evaluate nutrient-disease relationships can only lead to inadequate and perhaps incorrect interpretations.
The slower rate of repletion and lower plasma AA levels attained on the second repletion, after an apparently adequate first repletion, may also have public health implications. Although both repletions began with subjects at similar plasma levels, both provided an intake of 117 mg/day and both lasted for one month, the mean plasma AA was significantly lower on the second repletion than on the first (p<0.0001), and most subjects had not yet reached a plateau after one month. There were large differences between the first vs. the second repletion in the percent of subjects in who failed to reach 40 µmol/L (7% vs. 25%, respectively) or failed to reach 58 µmol/L (1.0 mg/dL) (37% vs. 75%, respectively). It may be that the first cycle depleted major body pools, which were not fully repleted after the first one-month repletion on 117 mg/day. In that case, repeated depletions which are not fully compensated may result in increasingly low circulating AA levels. Such repeated depletions may not be uncommon among the poor, among those with chronic illnesses or among the elderly who are subject to increasingly frequent illness or hospitalization. Numerous studies have found that hospitalized persons or those with chronic diseases have depleted plasma AA values [40–44].
These data also have implications for research design. First, it would appear that one month is insufficient for plasma levels to reach a plateau in all subjects, if the dose is in the dietary ranges, c. 100 to 150 mg/day. Second, studies should match treatment groups on body weight, stratify randomization within body weight groups or control for this factor. Finally, the great variation in both average levels and standard deviations seen with the random subsets of n=7 in this data set, and by others [21], emphasizes the difficulty in determining an accurate 2SD "margin of error" in recommendations. In his estimate of ascorbate requirements, Kallner [21] noted that, "Because of the low number of data points, the interindividual variation can hardly be calculated by correct statistical means and only rough estimations are possible;" his later study on smokers found that "a larger scattering was observed than in nonsmokers" [45]. Although this might have been due to differences in smoking habits, as theorized by Kallner, it may also have been due to variations resulting from small samples. The success of studies like this one, Blanchard et al. [11], and the DASH study [12] demonstrate the possibility of conducting large-scale controlled-diet studies on free-living subjects. While exquisitely-controlled metabolic-unit studies on very small numbers will always have an important role in nutrition research, it would seem that larger samples may be needed to fully appreciate the complexities of human responses to this and other nutrients.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Gladys Block, PhD, is currently at the University of California, Berkeley.
Received June 1, 1999. Accepted September 1, 1999.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. Dietrich, G. Block, E. P Norkus, M. Hudes, M. G Traber, C. E Cross, and L. Packer Smoking and exposure to environmental tobacco smoke decrease some plasma antioxidants and increase {gamma}-tocopherol in vivo after adjustment for dietary antioxidant intakes Am. J. Clinical Nutrition, January 1, 2003; 77(1): 160 - 166. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. M Preston, C. Rodriguez, C. E Rivera, and H. Sahai Influence of environmental tobacco smoke on vitamin C status in children Am. J. Clinical Nutrition, January 1, 2003; 77(1): 167 - 172. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. BLOCK Ascorbic Acid, Blood Pressure, and the American Diet Ann. N.Y. Acad. Sci., April 1, 2002; 959(1): 180 - 187. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Block, A. R. Mangels, E. P. Norkus, B. H. Patterson, O. A. Levander, and P. R. Taylor Ascorbic Acid Status and Subsequent Diastolic and Systolic Blood Pressure Hypertension, February 1, 2001; 37(2): 261 - 267. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
