The Night Vision Threshold Test (NVTT): a Simple Instrument for Testing Dark Adaptation in Young Children

The Night Vision Threshold Test (NVTT): A Simple Instrument for Testing Dark Adaptation in Young Children by Burris Duncana, Louise Canfieldb, Brent Barbera, John Greivenkampc, Francis O. Oriokotd, and Florence Naluyindad aDepartment of Pediatrics, College of Medicine and Children’s Research Center, College of Medicine, University of Arizona, USA bDepartment of Biochemistry, College of Medicine, and cThe College of Optical Sciences, University of Arizona, 1501 N Campbell Avenue, Tucson, AZ 85724, USA dDepartment of Pediatrics and Mwanamugimu Nutrition Unit, Makerere Medical School, Kampala, Uganda

Summary It is estimated that 41 per cent of the population aged under 5 in the developing world has an inadequate vitamin A dietary intake resulting in increased morbidity and mortality. Half a million children go blind each year as a result of vitamin A deficiency. Thirteen and a half million have night blindness, the first sign of vitamin A deficiency. Unfortunately, there is no simple, sensitive and inexpensive means to identify the child who has marginal levels of vitamin A and thus institute means to prevent their development of severe deficiency. A low cost, simple, easy-to-use instrument designed to detect a young child’s ability to adapt to darkness was tested in children admitted to the Mwanamugimu Nutrition Unit at Makerere Medical School in Kampala, Uganda. Despite the severe degree of malnutrition found in these children, Night Vision Threshold Test results and serum retinol levels were related (r ¼ 0:41, p < 0:05). Further efficacy trials for this instrument are planned at community sites in Nepal.

Introduction relatively inexpensive, their distribution is expensive and UNICEF estimates that 231 million children under the a means to target those communities where there is a age of 5 years or 41 per cent of the entire population documented significant prevalence of vitamin A defi- under 5 years old in the developing world have an ciency would be economically sound. However, all inadequate dietary intake of vitamin A. Thirteen and a currently available methods for the assessment of a half million children have night blindness, the first population for vitamin A deficiency have serious limitations: some require serum and expensive analysis clinical sign of vitamin A deficiency. Three million 3 children have xerophthalmia and 500 000 have severe of retinol levels; some lack sensitivity and specificity, 1 and others require considerable cooperation and are eye damage resulting in blindness. Vitamin A is the 4–10 most important cause of preventable childhood blindness inappropriate for young children. and represents one-third of all cases of childhood A history of night blindness remains the most blindness world-wide. More than one-half of these frequently used and the most reliable indicator for the earliest and mildest expression of clinical xerophthal- children die within a few months of becoming blind. In 11 addition, children with inadequate vitamin A status have mia. The history of night blindness, however, is based a 23 per cent greater risk of dying from common on the person’s perception; it does not determine the childhood illnesses, such as gastroenteritis, respiratory degree or severity of dark adaptation and its accuracy is diseases and measles.12 The efficacy of vitamin A highly questionable in the infant or very young child. supplementation for the young child with vitamin A A reliable, inexpensive, non-invasive, rapid test for deficiency in decreasing morbidity and mortality from assessing the prevalence of vitamin A deficiency in measles and measles-associated diarrhoeal, and respira- infants and young children is needed. Such a test would tory illness has been documented2 prompting vitamin A provide the means for the selection of at-risk individuals distribution programmes as a current strategy of the or communities to receive vitamin A supplementation WHO and UNICEF. Although vitamin A capsules are and nutrition educational programmes. It could be used for monitoring the effectiveness of vitamin A interven- tion programmes designed to restore vitamin A status in individuals and in their communities. In addition, most Correspondence: B. Duncan, Department of Pediatrics, Room 3335, University Medical Centre, 1501 N Campbell Avenue, previous techniques have targetted severely vitamin A Tucson, AZ 85724, USA. E-mail . developing clinical symptoms of vitamin A deficiency.

30 ᭧ Oxford University Press 2000 Journal of Tropical Pediatrics Vol. 46 February 2000 B. DUNCAN ET AL.

This group represents about 2 per cent of the population 30 cm. A green LED is used to match the spectral of most developing countries. Unfortunately, the effects response of the eye. Since the light output of an LED is a of severe vitamin A deficiency are not reversible and function of the electrical current through the LED, the children at this stage often die despite supplementation. illuminance of the projected spot can be easily varied by There are, however, many more children with marginal connecting the LED to a battery through a series of vitamin A stores (0.35–0.7 mmol/l). In this group, resistors. A 12-position rotary switch was used allowing vitamin A deficiency is completely reversible. More the projected illuminance to vary from about 400 mlux importantly, vitamin A supplementation significantly down to about 0.1 mlux. Each switch position reduced the protects this group of children against respiratory and illuminance by a factor of about a half. A digital ammeter gastrointestinal infections which together are responsible with numeric display on the outside of the case is provided for a very large number of deaths of children world- to measure the electrical current through the LED and to wide.1 Thus, this study was designed to evaluate a monitor the light output while the system is in use. technique for identifying children with marginal vitamin A deficiency so that interventions can be Testing compliance implemented prior to the development of severe vitamin A game was designed and adapted after the method used A deficiency. Specifically, this study was designed to by audiologists for testing the hearing of young children. test the efficacy of an inexpensive, non-invasive, easy- In the case of hearing, a sound stimulus is activated in to-use instrument, the Night Vision Threshold Tester one corner of a sound-proof room and the hearing-abled (NVTT), to evaluate infants and young children for their child will turn to the place where the sound originates. In ability to adapt to darkness or for the presence of night the game we designed to test for dark adaptation; after blindness; the first clinical sign of vitamin A deficiency. sitting in a dark room for 10 min, a light is directed to one corner of the room and scanned from one side to the other. If the child can see the light, he/she should follow Materials and Methods it and light perception can be determined by observing Instrument design the child’s head turn in response to the movement of the The design of the instrument is shown in Fig. 1. Its use is light. based on the premise that the function of the rods in the Twenty-four children between the ages of 7 and 45 retina, the eye’s ability to adapt to darkness, is a reflection months attending the Pediatric Clinic at the University of of the body’s retinol stores. The instrument is contained in Arizona were evaluated to determine if young children a plastic box approximately 12 × 18 × 5 cm. A light- would play this ‘game’. Four (17 per cent) could not be emitting diode (LED) is imaged and magnified by a tested because of inattention or being afraid of the dark. simple lens to project a round uniform circle of light on The remaining children (83 per cent), 13 (65 per cent) of the wall of a darkened room. The system was designed for whom were less than 24 months of age, co-operated with operation at 3 m that produces an illuminated spot of about the test procedure.

Fig. 1. The Night Vision Threshold Tester (NVTT).

Journal of Tropical Pediatrics Vol. 46 February 2000 31 B. DUNCAN ET AL.

Fig. 2. Light perception of subjects with retinitis pigmentosa. Results of SSTDA as compared with light intensity from the NVTT instrument.

Standardization of the NVTT Dietary information Nine subjects between the ages of 8 and 61 months with Both a 24-h and a 7-day food frequency dietary history, retinitis pigmentosa (RP) were tested for night blindness including an estimate of serving size, were obtained with the NVTT and the results compared with those from the child’s mother by the field research team. A obtained using the more sophisticated SST Dark qualitative analysis of the dietary data was performed to Adaptometer (SSTDA). RP is a condition that affects describe the dietary pattern for the study population. the rods of the retina and in its early phases causes night blindness. Each eye of nine subjects was tested Determination of the night vision threshold individually. Twelve of the 18 eyes tested were not The child sits on the mother’s lap in a completely dark able to detect light from the highest intensity emitted room facing a wide strip of white paper attached to the from the SSTDA but could detect the light beam when wall 3 m in front of the subject. The observer stands shown through the brightest light of our instrument, as it behind the mother and child and, while waiting for dark was far brighter than the brightest light emitted from adaptation to occur, discusses the importance of vitamin the SSTDA. The varying levels of light intensity detected A and how to enrich the child’s diet with it. After 10 min by each eye of the other twelve subjects (six eyes) had a in absolute darkness, the tester then turns on the light linear relationship with the amount of light emitted from source from the NVTT and directs it at the white strip of our instrument and that from the SSTDA (Fig. 2). paper to the child’s right or left and scans the light from one side to the other. Observing movement of the child’s Malnourished children head is facilitated by a small cap decorated with ‘glow- Forty-eight children between the ages of 6.7 and 67 in-the-dark’ stars that the child wears. The intensity of months were enrolled in the study in Uganda. Forty- the light is decreased until the child no longer follows the two had been admitted to the Mwanamugimu Nutrition light and that level of intensity is noted as the level of Unit, Makerere Medical School and Mulago National dark adaptation for the child. Referral Hospital in Kampala and the other six were recruited from an ophthalmology clinic in a neighbour- ing community. Demographic information and clinical Vitamin A supplementation and dietary histories were obtained from the mothers. Each child was administered the contents of a capsule The children were examined, anthropometric measures containing vitamin A as retinyl palmitate in oil supplied taken and the dark adaptation test was done as described by the Helen Keller Foundation. The dosage was 100 000 below. A blood sample was obtained and the child was International Units (IU) for infants between 6 and 12 given a vitamin A supplement. months and 200 000 IU for the older children.

32 Journal of Tropical Pediatrics Vol. 46 February 2000 B. DUNCAN ET AL.

Sample collection and preparation Results and Discussion Two to three ml of blood was obtained by venipuncture Demographics and transferred to glass tubes. Blood was also drawn Twenty-six of the 48 children were males and 22 were from one of us (BD) on each sample day and served as a females. The average age was 21.9 months (range 6.7– control. The blood was centrifuged and the serum 67.3 months). Fourteen (29 per cent) were less than 12 removed and stored in a ¹40ЊC freezer with a back-up months; 18 (37.5 per cent) were between 13 and 24 power source until it could be transported to the laboratory months; nine (18.7 per cent) were between 25 and 36 in Tucson, Arizona for analysis. Serum samples were months; and seven (14.6 per cent) were more than 36 precipitated with ethanol to remove protein and the lipids months of age. Ten (21.8 per cent) of the 48 were the were extracted from the supernatant with hexane as 10 only children in the family, although three of the 10 were previously described. only children as a result of sibling deaths. There were three sets of twins but in only two were both twins in the HPLC analysis malnutrition unit. There were two other families each of Hexane extracts were evaporated to dryness under N2 whom had two siblings in the unit and were included in and re-suspended in 150 ml of the mobile phase, the study. Thus four of the 44 families had two children [methanol:tetrahydrofuran (THF),(90:10 v/v) containing each in the malnutrition unit. Fourteen (29 per cent) were 0.25 g/l butylated hydroxytoluene (BHT)]. The extract either twins or had younger siblings in the family. (50 ml of the 150 ml sample) was injected onto a YMC (Morris Plains, NJ) reversed phase C18 column using an Anthropometrics IBM auto sampler (Model LC/9050 SE). Samples were Only two of the 48 children were above the National eluted isocratically at a flow rate of 1.7 ml/min using a Center for Health Statistics’ 50th centile weight for age Waters model 510 pump. The HPLC effluent was detected while 77 per cent (37/48) were below the 5th centile and at 325 nm using a Milton Roy programmable detector, 64 per cent (31/48) were at or below the first centile. model SM 4000 controlled by a Maxima 820 Chromato- One-third (16/48) of the children had oedema which graphy Workstation (Warters Associates, Milford, MA). inflated the weight data. Only four of the 48 were above Automated integration of vitamin A on all chromato- the 50th length for age centile while 71 per cent (34/48) grams was verified manually by the technician. Where were below the 5th centile which reflects long-standing baseline resolution could not be achieved, tangent skim- undernutrition or stunting. Forty-two per cent (20/48) of ming analyses were performed to remove bias. Using an the children were below the 5th centile weight for length. injection volume of 50 ml, limits of detection (defined as a Twelve (25 per cent) of the 48 children were still peak with a signal:noise ratio Ն3) for retinol was receiving at least a portion of their nutrition from 30 nmol. breastmilk. Of the 36 who were no longer being breastfed, 14 per cent had been weaned before the 6th Quantitation month and 39 per cent were no longer breastfeeding by The HPLC was calibrated at the onset of the study their first birthday. Seventy-two per cent (26/36) had using standard curves established from authentic retinol been weaned before 18 months and 94 per cent were no [National Institutes of Standards and Technology (NIST)] longer receiving breastmilk at 24 months of age. and the same curve was used for standardization throughout the study. Dietary analysis (Table 1) Nutritional intake for this group of children reflects Data analysis their severe undernourished condition. The 7-day food Descriptive statistics and correlations were performed frequency information reveals a monotonous diet con- using Microsoft Excel 5.0 (Microsoft Corp.) sisting primarily of milk, corn and bananas. Although,

Table 1 Categories of children’s 7-day food consumption

Type of food categories eaten by the children

Servings per weeka Vegetables Fruits Seeds Breastmilk Milk Meats, eggs, fish

< 1 4 9 11 37 6 5 1–5 14 19 23 3 14 26 6–10 12 11 6 7 25 10 11–15 10 5 6 0 3 5 16–20 3 1 0 0 0 1 >20 5 3 0 0 0 1

aReported by mothers using a 7-day food frequency instrument.

Journal of Tropical Pediatrics Vol. 46 February 2000 33 B. DUNCAN ET AL.

Fig. 3. Correlation between dark adaptation and serum retinol levels in children whose retinol level was < 0.7 mmol/l. most of the children received at least one serving of milk with vitamin A deficiency and management recommen- each day, almost two-thirds (31/48) received less than dations include the administration of large doses of one serving of either meat, eggs or fish each day and vitamin A on admission to the clinic or hospital. about the same percentage received no more than one In summary, in preliminary trials, an inexpensive, serving of vegetables (31/48) nor more than one serving simple-to-use instrument designed to measure a young of fruits (29/48) each day. child’s ability to adapt to darkness, or night vision, correlates with serum retinol concentrations. This Relationship of NVTT and serum retinol concentration instrument shows potential for identifying children (Fig. 3) with marginal vitamin A stores and thus offers the For the group, serum retinol concentrations and NVTT opportunity to institute the means to prevent their values were related (r ¼ 0:3) but the correlation did not developing severe vitamin A deficiency. reach significance (p < 0:12). However, when NVTT values with retinol levels in subjects whose serum retinol concentrations were less than 0.35 mmol/l, the correlation References was stronger (r ¼ 0:41, p < 0:05). These data indicate that the NVTT may be a valuable field method for 1. The State of the World’s Children. Oxford University Press, Oxford, 1995. detecting children with marginal levels of vitamin A 2. Semba RD. Vitamin A, immunity and infection. Clin Infect deficiency. Because serum retinol is not a reliable Dis 1994; 19: 489–99. predictor of vitamin A status, the present data do not 3. Emran N, Sommer A. Lissamine green staining in the allow us to evaluate directly the efficacy of the NVTT for clinical diagnosis of xerophthalmia. Arch Opthalmol 1979; predicting the vitamin A status of an individual. For this 97: 2333–35. work, more sensitive indicators such as the mRDR13,14 4. Natadisastra G, Wittpenn JR, Muhilal, West KP, Mele L, are needed. Studies to evaluate this relationship are Sommer A. Impression cytology: A practical index of currently in progress. vitamin A status. Am J Clin Nutr 1988; 48: 695–701. The present data indicate that the NVTT is not suitable 5. Thornton SP. A rapid test for dark adaptation. Ann Ophthal- mol 1977; 731–34. for evaluating the severely malnourished child. These 6. Solomons NW, Russell RM, Vinton E, Guerrero A-M, data are consistent with the poor response of these Mejia L. Application of a rapid dark adaptation test in children to external stimuli due to the malaise accom- children. J Pediatr Gastroenterol Nutr 1982; 4: 571–74. panying malnutrition. In addition, severely malnourished 7. Olson JA. A guide to current methods of assessing vitamin children often present with clinical symptoms consistent A status. Rapid dark adaptation time. Part 6. International

34 Journal of Tropical Pediatrics Vol. 46 February 2000 B. DUNCAN ET AL.

Vitamin A Consultative Group (IVACG) Secretariat, xerophthalmia screening. Am J Clin Nutr 1980; 33: 887–91. Nutrition Foundation,1993; 29–30. 12. Reddy V, Bhaskaram P, Raghuramulu N, et al. Relation- 8. Udomkesmalee E. A guide to current methods of assessing ship between measles, malnutrition and blindness: A vitamin A status. Vision restoration time. Part 6. Interna- prospective study in Indian children. Am J Clin Nutr tional Vitamin A Consultative Group (IVACG) Secretariat, 1987; 44: 924–30. Nutrition Foundation, 1993; 27–28. 13. Stephensen CB, Alvarez JO, Kohatsu J, Hardmeier R, 9. Zabriskie NA, Kardon RH. The pupil photostress test. Kennedy JI, Gammon RB. Vitamin A is excreted in the Ophthalmology 1994; 101: 1122–30. urine during acute infection. Am J Clin Nutr 1994; 60: 10. Congdon N, Sommer A, Severns M, et al. Pupillary and 388–92. visual threshold in young children as an index of population 14. Tanumihardjo SA, et al. Vitamin A status in preschool-age vitamin A status. Am J Clin Nutr 1995; 61: 1076–82. Indonesian children as assessed by the modified relative- 11. Sommer A, Hussaini G, Muhilal, Tarwotjo I, Susanto D, dose-response assay. Am J Clin Nutr 1990; 52: 1068–72. Saroso JS. History of night blindness: a simple tool for

Journal of Tropical Pediatrics Vol. 46 February 2000 35