Effect of micronutrient status on natural killer cell immune function in healthy free-living subjects aged 90 y

¹ From the Department of Internal Medicine, Cardioangiology, and Hepatology, the Department of Angiology and Blood Coagulation, and the Division of Geriatric Medicine, University Hospital Sant’Orsola-Malpighi, Bologna, Italy; the Laboratory of Immunology and Genetics and the Laboratory for Biocompatibility Research on Implant Materials, Rizzoli Orthopaedic Institute, Bologna, Italy; and the Department of Biochemistry G Moruzzi, University of Bologna, Italy.

² Supported in part by grants from the Ministero dell’Universita e della Ricerca Scientifica and from Ricerca Corrente, the Istituto Ortopedico Rizzoli.

³ Address reprint requests to G Ravaglia, Department of Internal Medicine, Cardioangiology, and Hepatology, University Hospital Sant’Orsola-Malpighi, Via Massarenti, 9, 40138 Bologna, Italy. E-mail: ravaglia{at}almadns.unibo.it.

ABSTRACT  
Background: Natural killer (NK) cells play a role in natural immunity against tumor and infected cells. Advanced aging is associated with functional impairment of NK cells and increased susceptibility to nutritional deficiencies.

Objective: Our objective was to test whether micronutrient status affects NK cell activity in an older population.

Design: The relations between NK cell variables (percentage of leukocytes and cytotoxicity) and blood concentrations of selected micronutrients were studied in 62 healthy, free-living northern Italian subjects (25 men, 37 women) aged 90–106 y. Anthropometric measurements were also made.

Results: All subjects were well nourished according to age-specific anthropometric norms but many of them had micronutrient deficiencies. The prevalence of micronutrient deficiency was highest for selenium (in 50% of both sexes), zinc (in 52% of men and 41% of women), and vitamin B-6 (in 40% of men and 59% of women), followed by vitamin A (in 16% of men and 27% of women) and vitamin E, vitamin B-12, and folate (each in <10% of both sexes). Ubiquinone-10 status was inadequate in 40% of women and 24% of men (P = 0.02). The percentage of NK cells was associated with serum zinc (men: r = 0.573, P = 0.007; women: r = 0.373, P = 0.031) and selenium (women: r = 0.409, P = 0.018) concentrations. In women only, NK cell cytotoxicity at different effector-target cell ratios was positively associated with plasma vitamin E and ubiquinone-10 concentrations (P < 0.05). No significant associations with NK cell variables were found for the other measured nutrients.

Conclusions: The results of this study strengthen the hypothesis that individual micronutrients may affect the number and function of NK cells in old age. The study also confirms the high prevalence of micronutrient deficiencies in healthy and apparently well-nourished persons aged 90 y.

Key Words: Natural killer cells • vitamins • trace elements • zinc • selenium • aging • nutrition • micronutrients • oldest-old age group • northern Italy

INTRODUCTION  
Natural killer (NK) cells are a distinct subpopulation of lymphocytes that show spontaneous cytolytic activity against various types of tumor or infected cells (1). NK cells therefore may play an important role in early natural surveillance against cancer and infectious disease, which are among the principal causes of mortality among elderly people in the Western world (2).

There are many descriptions of changes in the leukocyte subpopulations in aging, and these are not always comparable. A progressive age-related shift in the circulating lymphocyte population from conventional T cells to NK cells is well documented (3–7), although not a universal finding (8–10), in the literature. Conflicting data were also reported with regard to the functional activity of NK cells during aging [unchanged (4, 11, 12), decreased (3, 6, 7, 13), or increased (14)]. Possible explanations for these discrepancies are differences in donor selection criteria and sample size of studies (3, 4). However, abnormalities of NK cell function with age might also be related to microenvironmental changes and primarily to endocrine and nutritional factors (15).

Although nutrition is a critical determinant of immunocompetence, and the lack of specific micronutrients may be implicated in causing depressed cell-mediated immune responses in older age (16), only a few studies specifically investigated the relation between micronutrient status and NK cell function in elderly people (17–19). Moreover, none of these studies examined this relation in the growing age segment known as the oldest old (90 y), even though this age group represents a particularly vulnerable population with a precarious nutritional balance that can be easily disturbed by intercurrent illness, decreased functional capacity, or increased economic hardship (20).

We reported recently the existence of positive associations between NK cell activity and vitamin D status in a sample of healthy, free-living Italian persons 90 y old (15). In the present study we examined whether, in the same population, NK cell variables were associated with blood concentrations of other selected vitamins and trace elements.

SUBJECTS AND METHODS  
Subjects
Details of recruitment and enrollment of participants and assessment methods were published previously (15). Briefly, 62 noninstitutionalized subjects aged 90–106 y and living in the Bologna province (Emilia Romagna region of northern Italy) were enrolled. According to the suggestions of the SENIEUR protocol admission criteria for immunogerontologic studies in humans (21), all subjects were in apparently good health, had normal hematologic and biochemical values, and were not taking nutritional supplements or medications known to affect the immune system (steroidal and nonsteroidal antiinflammatory drugs, hormones, and analgesics). Given the advanced age of these subjects, it is not surprising that, among the clinical conditions not included in the SENIEUR exclusion criteria, the most frequently recorded in this population was arteriosclerosis, manifesting itself either as hypertension (n = 10) or as a previous (>6 mo old) episode of cardiac insufficiency (n = 4); myocardial infarction (n = 3); stroke (n = 3); or arterial recanalization (n = 2). Other minor complaints were cataract (n = 3), benign prostatic hyperplasia (n = 3), osteoarthritis (n = 3), and colonic diverticula (n = 1).

As in the SENIEUR protocol, however, none of our subjects was receiving a specific pharmacologic treatment for his or her medical condition at the time of this study. All subjects gave their informed consent to enrollment in the study, which was approved by the ethical committee of the Department of Internal Medicine, Cardioangiology, and Hepatology of the University of Bolgna.

Nutritional assessment
Anthropometric measurements
All subjects were weighed on the same scales while barefoot and in light clothing. Because shrinkage of the spine with aging can affect the validity of height measurements in the elderly, height was calculated from the knee-height measurement, according to the equations of Chumlea et al (22). Body mass index (in kg/m²) was calculated. Arm muscle area and arm fat area were calculated from midarm circumference and triceps skinfold thickness by using standard formulas (23). These measurements were performed according to standardized procedures (24), with the patient in a recumbent position to make measurement easier and prevent falsification of results by alterations in mobility.

Blood nutrients
Peripheral blood samples were collected from 0800 to 0900, after the subjects had fasted overnight. The samples were put on ice, transported to the laboratories within 1 h, and processed immediately. Plastic tubes containing tripotassium EDTA and metal-free evacuated tubes containing no additives (Becton Dickinson, Meylan, France) were used for collecting plasma and serum, respectively. Plasma was separated by centrifugation (1500 x g, 30 min, 4°C) and analyzed immediately. Serum was separated by centrifugation (3000 x g, 30 min, 4°C) and samples were appropriately stored at -70°C until analyzed.

Plasma retinol, -tocopherol, and ubiquinone-10 were assayed simultaneously by reversed-phase HPLC (Millennium 2010; Waters, Milford, MA) (25, 26). Because plasma lipids influence blood concentrations of lipid-soluble vitamins, plasma retinol, -tocopherol, and ubiquinone-10 concentrations were adjusted to plasma cholesterol and triacylglycerols (27), measured according to routine procedures.

Serum selenium concentrations were measured in triplicate with a graphite furnace atomic absorption spectrometer with Zeeman background correction (Solaar 939QZ Unicam; Cambridge, United Kingdom) by using a standard addition method. Bovine serum with a low selenium concentration was used to prepare the standard curve. Human sera [standard reference material 1598; National Institute of Standards and Technology (NIST), Gaithersburg, MD] were used to validate the accuracy and precision of the method. Samples were compared with the standard curve by using the standard curve linear least-squares fit analysis. The detection limit for selenium was 0.093 µmol/L (7.4 ng/mL).

Serum zinc concentrations were measured by using a flame atomic absorption spectrometer (PU 9400, NL; Pye Unicam, Philips, Eindhoven, Netherlands) equipped with an air-acetylene flame burner. A linear calibration curve was created by using certified standard NIST solutions at 3 concentrations (0.1, 0.2, and 0.3 mg/L). Specimens were diluted 1:5 with 5% glycerol and ultrapure bidistilled and deionized water and analyzed in duplicate. Seronorm trace elements (Nycomed Pharma, Oslo, Norway) and serum trace elements control toxicology (normal range; Utak Laboratories Inc, Valencia, CA) were used as controls for validating the accuracy and precision of the method. The detection limit for zinc was 0.153 µmol/L (0.01 µg/mL).

Serum folate and vitamin B-12 concentrations were assayed by competitive immunoassay using direct, chemiluminescent technology (Chiron Diagnostics Co, East Walpole, MA). Plasma concentrations of pyridoxal-5'-phosphate (the active coenzyme form of vitamin B-6) were assayed by HPLC. Briefly, plasma proteins were precipitated by adding 1 mol perchloric acid/L (0.5 mL) to plasma (0.5 mL) and centrifuging at 13000 x g for 15 min at 4°C. Then, 0.1 mL of a 2.62% sodium bisulfate solution (pH 1.0) was added to 0.5 mL of the supernate and 50 µL of the obtained solution was injected onto a C₁₈ column (Bakerbond; MetaChem, Torrance, CA) protected by a saturation column packed with 40 µm C₈ (Isolute; International Sorbent Technology, Jones Chromatography Lmt, Mid Glamorgan, United Kingdom). The mobile phase was 0.1 mmol sodium phosphate/L buffer (pH 1.5). The eluted peaks were monitored by a fluorometric detector set at an excitation of 300 nm and an emission of 400 nm. Intra- and interassay CVs were <5%. The detection limit for plasma pyridoxal-5'-phosphate was 4 nmol/L.

Immunologic tests
Mononuclear cell preparation and flow cytometry analysis
A complete blood profile and count was obtained for each subject. Peripheral blood mononuclear cells were separated from heparin-containing blood by conventional gradient centrifugation. The following monoclonal antibodies (MoAbs), directly conjugated with fluorescein isothiocyanate (FITC) or phycoerithrin, were used to analyze the surface antigens of peripheral blood mononuclear cells: anti-CD3, recognizing all T cells; anti-CD16, recognizing the low-affinity receptor for FcIgG, reactive with a subset of cells with NK cell activity; anti-CD56, recognizing the nuclear cell adhesion molecule, reactive with resting and activated CD16⁺; anti-CD2, recognizing an intercellular adhesion molecule binding the leukocyte function-associated antigen 3 (LFA-3) and expressed by both T and NK cells; anti-CD11a and anti-CD11b, recognizing different chains of the ß₂ integrin family; anti-CD18, recognizing the common ß subunit of the ß₂ integrin family; and anti-CD 29, recognizing the common ß chain of the ß₁ integrin family.

Anti-CD3, -CD16, -CD56, -CD2, and -CD29 were purchased from Becton Dickinson (Mountain View, CA). Anti-CD11a (LFA-1, TS1/22 clone), -CD11b (L2/1 clone), and -CD18 (LFA-1ß, TS1/18 clone) were prepared in our laboratory from culture supernates of hybridomas.

Incubation of 2 x 10⁵ mononuclear cells with MoAbs was performed in V-bottom plates for 30 min at 4°C, followed by the secondary antibody (goat anti-mouse FITC-immunoglobulin, 1:20 dilution; Becton Dickinson) when the MoAbs were not directly conjugated with fluorochrome. Incubation with 1:10 mouse normal serum (Dako, Glostrup, Denmark) was performed to reduce nonspecific binding.

Negative control cells were incubated with only immunoglobulin isotype control, FITC-conjugated goat anti-mouse immunoglobulin, or both. Mononuclear cells were then washed with phosphate-buffered saline–fetal calf serum and resuspended in 1% paraformaldehyde. The analysis was performed by using cytometry with a FACStar Plus cell sorter (Becton Dickinson) (3).

NK cell lytic activity
K562 tumor target cells (2 x 10⁶) were incubated with 3.7 MBq radioactive sodium chromat (specific activity: 1480 x 10¹⁰/–/4440 x 10¹⁰ /Bq/g) (NEN, Köln, Germany) for 1 h at 37°C and shaken occasionally. Tumor cells were then washed 3 times with cold medium, centrifuged at 100 x g for 7 min at 4°C, and resuspended at a concentration of 10⁷/L. ⁵¹Cr labeled target cells [10⁷/L (5 x 10²/50 µL)] and a varying number of effector cells [from 5 x 10⁸ to 5 x 10⁶/L (5 x 10⁴ to 5 x 10²/100 µL)] were incubated in triplicate in V-bottom 96-well plates [effector-target (E-T) ratios from 100:1 to 1:1]. Then, 75 mL supernate was harvested and counted in a gamma counter, and ⁵¹Cr release was measured as described previously (28).

Statistical methods
Data are reported as means ± SDs, medians with 5th and 95th percentiles, or numbers of subjects with percentages, as appropriate. Because of the relatively small number of subjects and the nonnormal distribution of many nutritional variables, data were analyzed by using nonparametric statistics. Differences between groups were compared by using a Mann-Whitney U test or a chi-square test, as appropriate. The strength of the associations between NK cell immune function variables and the variables of interest was assessed by using the Spearman correlation coefficient (r_s). P values <0.05 were considered significant. Statistical calculations were performed by using SIGMASTAT 2.0 (Jandel Scientific, Erkrath, Germany).

RESULTS  
Nutritional characteristics
No significant differences in age or anthropometric characteristics were found between men and women (Table 1), except for arm muscle area, which was higher in men than in women. Consistent with the advanced age of the study population, anthropometric measurements were on average very low, but values for all subjects fell within the normal range (5th and 95th percentiles) for the population aged 90 y and living in the Emilia Romagna region (24). The same was not true for blood micronutrient concentrations, which are shown in Table 2 along with the 5th and 95th percentiles for blood nutrient concentrations obtained from a sample of 50 healthy men and women aged 20–64 y and living in the Bologna province. Subjects in the current study whose blood nutrient concentrations fell below the limits of these reference standards were defined as having deficiencies. The prevalence of micronutrient deficiency was highest for selenium (in 50% of both sexes), zinc (in 52% of men and 41% of women), and vitamin B-6 (in 40% of men and 59% of women), followed by vitamin A (in 16% of men and 27% of women) and vitamin E, vitamin B-12, and folate (each in <10% of both sexes). No sex-related differences in the prevalence of micronutrient deficiencies were found except for ubiquinone-10, the deficiency of which was more frequent in women (40%) than in men (24%).

View this table:
TABLE 1.. Anthropometric data¹  

View this table:
TABLE 2.. Blood micronutrient data  
Immunologic characteristics
The immunologic characteristics of this population were extensively described elsewhere (15). Briefly, there was no significant difference between men and women in the distribution of T (CD3⁺) and NK (CD16⁺ and CD56⁺) cells nor in the distribution of some adhesion molecules involved in the interaction with target cells [CD2, CD29, and the ß₂ integrin family (CD11b and CD18)] both on the total mononuclear cells (Figure 1A) and on electronically gated CD16⁺ cells (Figure 1B). Of note, this subject population as a whole had lower percentages of CD3⁺ (P < 0.001), CD56⁺ (P < 0.05), CD11a⁺ (P < 0.01), CD18 (P < 0.01), and CD16⁺/CD11a⁺ cells (P < 0.05) than did a younger control group (10 men and 7 women aged 21–38 y; ± SD: 27 ± 6 y). The percentage of CD16⁺ cells in the peripheral blood of subjects 90 y old was directly related to the percentage of CD56⁺ cells (P < 0.001) and inversely correlated with the percentage of CD3⁺ T cells (P = 0.002).

View larger version (33K):
FIGURE 1. . Surface marker expression of fresh peripheral blood lymphocytes and electronically gated CD16⁺ lymphocytes from men (n = 25; lymphocyte count: 1.5 ± 0.6 x 10⁹/L) and women (n = 37; lymphocyte count: 1.8 ± 0.5 x 10⁹/L) aged 90 y. The young control subjects (n = 17; lymphocyte count: 1.9 ± 0.7 x 10⁹/L) were aged 21–38 y. Results are expressed as mean (±SD) percentages of positive cells.

 
NK cell cytolytic activity tended to be slightly but not significantly higher in women than in men (Figure 2). As a whole, subjects aged 90 y had significantly lower cytotoxic activity (P < 0.001 for E-T ratios from 100:1 to 12.5:1) than did younger control subjects (29 men and 10 women aged 21–38 y; ± SD: 29 ± 5 y). NK cell cytolytic activity of subjects aged 90 y was also directly correlated with their CD16⁺ cell number for all the E-T ratios examined (P < 0.05).

View larger version (13K):
FIGURE 2. . Natural killer (NK) cell cytotoxicity for different effector-target (E-T) ratios in healthy, free-living men (n = 25) and women (n = 37) aged 90 y. The young control subjects (n = 39) were aged 21–38 y. Results are expressed as mean (±SD) percentages of ⁵¹Cr release.

 
Relation between NK cell variables and blood nutrients
Percentages of CD16⁺ and CD56⁺ cells of subjects aged 90 y were positively associated with serum zinc concentrations in both sexes (Figure 3). In women, percentages of total CD16⁺ and CD16⁺ cells expressing adhesion molecules of the ß₂ integrin family (CD11b and CD18) were positively associated with serum selenium concentrations (Figure 4). In men, only the association of the percentage of total CD16⁺ cells with the serum selenium concentration was nearly significant (r_s = 0.437, P = 0.053).

View larger version (21K):
FIGURE 3. . Relations among percentages of CD16⁺ and CD56⁺ lymphocytes and serum zinc concentrations in healthy, free-living men and women aged 90 y.

 

View larger version (18K):
FIGURE 4. . Relations among percentages of CD16⁺, CD16⁺/11b⁺, and CD16⁺/18⁺ lymphocytes and serum selenium concentrations in healthy, free-living women aged 90 y.

 
No significant associations were found among the selected nutrients and the percentages of CD3⁺, CD2⁺, CD11a⁺, CD11b⁺, CD18⁺, and CD29⁺ peripheral blood cells. No significant associations were found among the selected nutrients and the percentages of CD16⁺ cells expressing CD2, CD11a, or CD29. Significant positive correlations of NK cell cytolytic activity with plasma lipid-adjusted vitamin E and ubiquinone-10 concentrations were found in women for all but one of the examined E-T ratios (Figure 5). No associations were found between NK cell cytolytic activity and blood nutrient concentrations in men.

View larger version (35K):
FIGURE 5. . Relations among natural killer (NK) cell cytotoxicity (⁵¹Cr release) and plasma vitamin E and ubiquinone-10 in healthy, free-living women aged 90 y.

 

DISCUSSION  
This study showed that NK cell immune function of healthy, free-living subjects aged 90 y was associated with blood concentrations of certain individual micronutrients of which deficiencies are very common in advanced age. Our results agree with the data of Chandra (19), which showed an improvement in NK cell variables of healthy subjects aged >65 y after 12 mo of supplementation with physiologic amounts of micronutrients. In contrast, Dowd et al (17) reported no significant correlations between NK cell function and individual blood nutrients, except for vitamin C. Their study population, however, included hospitalized patients, and very old subjects were scarcely represented.

Our results also differed from those of Payette et al (18), who found no correlations between NK cell function and micronutrient status in a group of elderly Canadian subjects. Comparisons with this study, however, are difficult because the Canadian study included subjects taking micronutrient supplements (29), and more than half of the study population had unusually low concentrations of interleukin 2.

On the basis of age- and region-specific anthropometric norms, no one in the population of the current study had clinically overt protein-energy undernutrition. On the basis of blood nutrient reference values for young subjects living in the same area and with similar dietary habits, however, about half of the study subjects had zinc and selenium deficiencies. Dietary intake was not estimated, but several conditions common in advanced age (reduction in total energy intake, lack of variety and characteristic self-selection of food items, poor socioeconomic conditions, malabsorption, and drug-nutrient interactions) were likely responsible for the reduced concentrations of these trace elements in this and other studies (20). Our data also agree with previous reports of possible micronutrient deficiencies even in independently living and apparently healthy elderly populations (30).

A remarkable finding in the current study was the strong association between the relative number of peripheral lymphocytes expressing markers of NK cell activity and zinc and selenium concentrations. Several studies in animals and humans showed decreased NK cell activity during zinc-deficient states (31). Zinc influences the activity of multiple enzymes involved in activation, replication, and programmed death of lymphocytes and also acts as an antioxidant (32). However, exposure of NK cells to high concentrations of zinc in vitro appears to inhibit NK cell cytotoxicity, perhaps inducing a down-regulation of CD16 (33). Selenium is also involved in NK cell–mediated immunity because it appears to up-regulate the interleukin 2 receptor on the surface of NK cells, resulting in enhanced proliferation and clonal expansion of cytotoxic precursor cells (34).

In a previous study (3) we found that 35 SENIEUR study subjects aged 71–95 y had a progressively higher number of NK cells than did young control subjects, whereas in the current study we had a consistent population of SENIEUR study subjects aged 90 y who did not show such a difference. In our previous study, however, as well as in many other studies of leukocyte subsets in very old people (4–11), zinc and selenium were not taken into account. Indeed, the evaluation of zinc and selenium is not required by the SENIEUR protocol and marginal deficiencies of trace elements are hardly ever recognized during clinical examination. On the basis of the strong association between the number of NK cells and blood concentrations of both zinc and selenium found in the current study, it is possible that differences in trace element status might explain the controversial results in the literature about the effect of aging on the number of NK cells.

It has also been speculated that the increase in the number of NK cells with age could be a compensatory mechanism for coping with the reduced cytolytic activity of single NK cells (3, 4). If so, zinc and selenium might be of central importance in maintaining an effective natural immunity in the oldest-old age group. However, there is no definite evidence that differences in the number of NK cells among elderly subjects actually influence the ability of the elderly to fight infections and cancer. Ligthart et al (11) reported that NK cell function was directly proportional to the number of CD16⁺ cells in the peripheral blood. Ogata et al (7), however, reported that the activity, but not the number of, NK cells was related to the subsequent development of severe infections in elderly Japanese subjects. Our study also showed positive associations between NK cell cytotoxicity and plasma vitamin E and ubiquinone-10 concentrations in women. The lack of similar associations in men must be considered with caution because of the relatively small number of subjects in this subgroup and their better ubiquinone-10 status.

The low frequency of vitamin E deficiency in our subjects agrees with previous reports that aging per se has little effect on plasma vitamin E concentrations (30, 35), although other authors found low age-related vitamin E platelet values (36). Ubiquinone-10, a vitamin-like substance that is found in small amounts in a wide variety of foods but that can also be synthesized in all human tissues, is well known as a redox component in the mitochondrial respiratory chain and is an effective membrane antioxidant (37). The basal plasma concentrations of ubiquinone-10, which reflect metabolic demand (37), have been described as both increasing (38) and decreasing (39) with aging. Chronic malnutrition is presumed to affect ubiquinone-10 status both directly (by reducing ubiquinone-10 dietary intake) and indirectly (by reducing the dietary intake of other vitamins, such as vitamin B-6, folic acid, and vitamin B-12, that are involved in its biosynthesis) (37).

Several animal and human studies showed that adequate vitamin E intake is essential for immune function and that vitamin E supplementation can successfully improve some aspects of the age-related decline in immunity (40). Although the mechanism behind the immunostimulatory effect of vitamin E is still unknown, there is compelling evidence that vitamin E may exert its immunoenhancing effect by regulating prostaglandin synthesis, decreasing free radical formation, or both (41).

By contrast, little is known about the role of ubiquinone-10 as an immunomodulating agent. Folkers et al (42) reported an increase in IgG and T4 lymphocyte blood concentrations after ubiquinone-10 administration in human subjects, but no data are available about the possible influence of ubiquinone-10 on NK cell activity. In our study, plasma ubiquinone-10 and vitamin E concentrations were associated with NK cell activity. A possible explanation for our findings could be the sparing effect of ubiquinone-10 on vitamin E as a result of its effective antioxidant action and, alternatively (or additionally), its ability to recycle oxidized vitamin E (43).

In contrast with a few reports suggesting that vitamin A (44), vitamin B-6 (45), folate (46), and vitamin B-12 (47) may affect NK cell activity in overtly deficient subjects with various diseases, we did not find associations between NK cell variables and blood concentrations of these vitamins in our study population. Differences in age and health status of subjects in these studies may have contributed to this inconsistency in findings. Alternatively, many of these vitamins might indirectly affect NK cell function through their involvement in the multistage process of endogenous ubiquinone-10 biosynthesis.

In conclusion, the results of this study strengthen the hypothesis that individual micronutrients are related to NK cell function in old age. They also confirm the high prevalence of micronutrient deficiencies in apparently healthy and well-nourished subjects aged 90 y. The cross-sectional design of the study, however, does not permit one to infer that individual micronutrient deficiencies actually caused impairment of NK cell function. Further intervention studies are required to determine whether the provision of nutritional supplements effectively enhances NK cell immune function in elderly subjects. Meanwhile, a thorough nutritional assessment, paying particular attention to the dietary intake of essential nutrients, is strongly recommended in geriatric practice.

REFERENCES  

1. O'Shea J, Ortaldo JR. The biology of natural killer cells: insights into the molecular basis of function. In: Lewis CE, McGee JO, eds. The natural immune system. The natural killer cell. Oxford, United Kingdom: IRL Press, 1992:2–40.
2. Goodwin JS. Decreased immunity and increased morbidity in the elderly. Nutr Rev 1995;53:80S–85S.
3. Facchini A, Mariani E, Mariani AR, Papa S, Vitale M, Manzoli FA. Increased number of circulating Leu 11+ (CD16) large granular lymphocytes and decreased NK activity during human ageing. Clin Exp Immunol 1987;68:340–7.
4. Sansoni P, Cossarizza A, Brianti V, et al. Lymphocyte subsets and natural killer cell activity in healthy old people and centenarians. Blood 1993;82:2767–73.
5. Xu X, Beckman I, Ahern M, Bradley J. A comprehensive analysis of peripheral blood lymphocytes in healthy aged humans by flow cytometry. Immunol Cell Biol 1993;71:549–57.
6. Miyaji C, Watanabe H, Minagawa M, et al. Numerical and functional characteristics of lymphocyte subsets in centenarians. J Clin Immunol 1997;17:420–9.
7. Ogata K, Yokose N, Tamura H. Natural killer cells in the late decades of human life. Clin Immunol Immunopathol 1997;84:269–75.
8. Bender BS, Chrest FJ, Adler WH. Phenotypic expression of natural killer cell associated membrane antigens and cytolytic function of peripheral blood cells from different aged humans. J Clin Lab Immunol 1986;21:31–6.
9. Canonica GW, Caria M, Venuti D, Cipro G, Ciprandi G, Bagnasco M. T cell activation through different membrane structures (T3/Ti, T11, T44) and frequency analysis of proliferating and interleukin-2 producer T lymphocyte precursors in aged individuals. Mech Ageing Dev 1988; 42:27–35.
10. Beregi E, Regius O, Rajczy K. Comparative study of the morphological changes in lymphocytes of elderly individuals and centenarians. Age Ageing 1991;20:55–9.
11. Ligthart GJ, Schuit HR, Hijmans W. Natural killer cell function is not diminished in the healthy aged and is proportional to the number of NK cells in the peripheral blood. Immunology 1989;68:396–402.
12. Krishnaraj R. Immunosenescence of human NK cells: effects on tumor target recognition, lethal hit and interferon sensitivity. Immunol Lett 1992;34:79–84.
13. Mariani E, Monaco MCG, Cattini L, Sinoppi M, Facchini A. Distribution and lytic activity of NK cell subsets in the elderly. Mech Ageing Dev 1994;76:177–87.
14. Kutza J, Murasko DM. Effects of aging on natural killer cell activity and activation by interleukin-2 and IFN-alpha. Cell Immunol 1994;155:195–204.
15. Mariani E, Ravaglia G, Meneghetti A, et al. Natural immunity and bone and muscle remodelling hormones in the elderly. Mech Ageing Dev 1998;102:279–92.
16. Chandra RK. Nutritional regulation of immunity and risk of infection in old age. Immunology 1989;67:141–7.
17. Dowd PS, Kelleher J, Walker BE, Guillou PJ. Nutrition and cellular immunity in hospital patients. Br J Nutr 1986;55:515–27.
18. Payette H, Rola-Pleszczynski M, Ghadirian P. Nutrition factors in relation to cellular and regulatory immune variables in a free-living elderly population. Am J Clin Nutr 1990;52:927–32.
19. Chandra RK. Effect of vitamin and trace-element supplementation on immune responses and infection in elderly subjects. Lancet 1992;340:1124–7.
20. Morley JE. Nutritional status of the elderly. Am J Med 1986; 81:679–95.
21. Ligthart GJ, Corberand JX, Fournier C, et al. Admission criteria for immunogerontological studies in man: the SENIEUR protocol. Mech Ageing Dev 1984;28:47–55.
22. Chumlea WC, Roche AF, Steinbaugh ML. Estimating stature from knee height for persons 60 to 90 years of age. J Am Geriatr Soc 1985;33:116–21.
23. Frisancho AR. New norms of upper limb fat and muscle area for assessment of nutritional status. Am J Clin Nutr 1981;34:2540–5.
24. Ravaglia G, Morini P, Forti P, et al. Anthropometric characteristics of healthy Italian nonagenarians and centenarians. Br J Nutr 1997;77:9–17.
25. Arnaud J, Fortis I, Blachier S, Kia D, Favier A. Simultaneous determination of retinol, alpha-tocopherol and beta-carotene in serum by isocratic high-performance liquid chromatography. J Chromatogr 1991;572:103–16.
26. Aberg F, Appelkvist EL, Dallner G, Ernster L. Distribution and redox state of ubiquinones in rat and human tissues. Arch Biochem Biophys 1992;295:230–4.
27. Gey K, Puska P. Plasma vitamin E and A are inversely correlated to mortality from ischemic heart disease in cross cultural epidemiology. Ann N Y Acad Sci 1989;570:268–82.
28. Mariani E, Monaco MCG, Sgobbi S, De Zwart JF, Mariani AR, Facchini A. Standardization of a microtoxicity assay for human killer cell lytic activity. J Immunol Methods 1994;172:173–8.
29. Bendich A. Nutritional immunology in elderly people. Am J Clin Nutr 1991;53:983–4 (letter).
30. Hallfrisch J, Muller DC, Singh VN. Vitamin A and E intakes and plasma concentrations of retinol, beta-carotene, and alpha-tocopherol in men and women of the Baltimore Longitudinal Study of Aging. Am J Clin Nutr 1994;60:176–82.
31. Shankar AH, Prasad AS. Zinc and immune function: the biological basis of altered resistance to infection. Am J Clin Nutr 1998; 68(suppl):447S–463S.
32. Fabris N. Neuroendocrine-immune aging: an integrative view on the role of zinc. Ann N Y Acad Sci 1994;719:353–68.
33. Borrego F, Lopez-Beltran A, Pena J, Solana R. Downregulation of Fc gamma receptor IIIA alpha (CD16-II) on natural killer cells induced by anti-CD16 mAb is independent of protein tyrosine kinases and protein kinase C. Cell Immunol 1994;158:208–17.
34. Kiremidjian-Schumacher L, Roy M. Selenium and immune function. Z Ernahrungswiss 1998;37(suppl 1):50–6.
35. Campbell D, Bunker W, Thomas AJ, Clayton BE. Selenium and vitamin E status of healthy and institutionalized elderly subjects: analysis of plasma, erythrocytes and platelets. Br J Nutr 1989;62:221–7.
36. Vatassery GT, Johnson GJ, Krezowski AM. Changes in vitamin E concentrations in human plasma and platelets with age. J Am Coll Nutr 1983;2:369–75.
37. Littarru GP, Lippa S, Oradei A, Fiorini RM, Mazzanti L. Metabolic and diagnostic implications of blood CoQ10 levels. In: Folkers K, Yamagami T, Littarru GP, eds. Biomedical and clinical aspects of coenzyme Q. Vol 6. Amsterdam: Elsevier, 1991:167–78.
38. Di Palma A, Masini M, Cimatti G, Di Palma P, Adamanti S. Ubiquinone-10: a probable longevity marker. In: Folkers K, Yamamura Y, eds. Biomedical and clinical aspects of coenzyme Q. Vol 5. Amsterdam: Elsevier, 1986:489–92.
39. Ravaglia G, Forti P, Maioli F, et al. Ubiquinone-10 plasma levels and body composition in elderly males. Arch Gerontol Geriatr 1996;1(suppl):539–43.
40. Meydani M. Vitamin E. Lancet 1995;345:170–5.
41. Meydani SN, Beharka AA. Recent developments in vitamin E and immune response. Nutr Rev 1996;56:49S–58S.
42. Folkers K, Morita M, McRee J Jr. The activities of ubiquinone-10 and vitamin B6 for immune responses. Biochem Biophys Res Commun 1993;193:88–92.
43. Frei B, Kim MC, Ames BN. Ubiquinol-10 is an effective lipid-soluble antioxidant at physiological concentrations. Proc Natl Acad Sci U S A 1990;87:4879–83.
44. Semba RD. The role of vitamin A and related retinoids in immune function. Nutr Rev 1998;56:38S–48S.
45. Baum MK, Mantero-Atienza E, Shor-Posner G, et al. Association of vitamin B6 status with parameters of immune function in early HIV-1 infection. J Acquir Immune Defic Syndr 1991;4:1122–32.
46. Rosenthal GJ, Germolec DR, Lamm KR, Ackermann MF, Luster MI. Comparative effects on the immune system of methotrexate and trimetrexate. Int J Immunopharmacol 1987;9:793–801.
47. Kubota K, Kurabayashi H, Kawada E, Okamoto K, Shirakura T. Restoration of abnormally high CD4/CD8 ratio and low natural killer cell activity by vitamin B12 therapy in a patient with post-gastrectomy megaloblastic anemia. Intern Med 1992;31:125–6.