Original Article

Body composition, energy expenditure and physical activity

Feasibility of a SenseCam-assisted 24-h recall to reduce under-reporting of energy intake

  • European Journal of Clinical Nutrition volume 67, pages 10951099 (2013)
  • doi:10.1038/ejcn.2013.156
  • Download Citation
Published online:



The SenseCam is a camera worn on a lanyard around the neck that automatically captures point-of-view images in response to movement, heat and light (every 20–30 s). This device may enhance the accuracy of self-reported dietary intake by assisting participants’ recall of food and beverage consumption. It was the objective of this study to evaluate if the wearable camera, SenseCam, can enhance the 24-h dietary recall by providing visual prompts to improve recall of food and beverage consumption.


Thirteen volunteer adults in Oxford, United Kingdom, were recruited. Participants wore the SenseCam for 2 days while continuing their usual daily activities. On day 3, participants’ diets were assessed using an interviewer-administered 24-h recall. SenseCam images were then shown to the participants and any additional dietary information that participants provided after viewing the images was recorded. Energy and macronutrient intakes were compared between the 24-h recall and 24-h recall+SenseCam.


Data from 10 participants were included in the final analysis (8 males and 2 females), mean age 33±11 years, mean BMI 25.9±5.1 kg/m2. Viewing the SenseCam images increased self-reported energy intake by approximately 1432±1564 kJ or 12.5% compared with the 24-h recall alone (P=0.02). The increase was predominantly due to reporting of 41 additional foods (241 vs 282 total foods) across a range of food groups. Eight changes in portion size were made, which resulted in a negligible change to energy intake.


Wearable cameras are promising method to enhance the accuracy of self-reported dietary assessment methods.


The 24-h dietary recall is widely used in dietary assessment due to its low participant burden, ease of administration and suitability for a wide range of populations including participants with low literacy.1 However, as a retrospective method it relies on self-report without verification, thus lapses in participant’s memory, errors in portion size estimation or intentional misreporting remain unidentified, but contribute to measurement error.2, 3, 4

Analyses of data from national nutrition surveys in many countries including Australia, New Zealand and the United States of America, which used the 24-h recall have demonstrated substantial under-reporting of energy intake, particularly in females, overweight and obese participants.5, 6, 7 Considerable under-reporting has also been revealed in doubly labelled water validation studies using the 24-h recall administered in-person or via the telephone.4, 8, 9, 10 Under-reporting leads to attenuation and misclassification error, thus impeding the ability to identify associations between diet and disease. Consequently, the need for technological innovation is well recognised,11, 12 as technology has the potential to provide objective dietary intake data, reduced participant burden, standardised and automated methods of assessment, and new methods, previously not feasible, to assess dietary intake.

The development of structured interviewer-administered and self-administered computer-based 24-h recall systems can reduce interviewer bias and measurement error, but such systems still rely on self-report without the ability to verify dietary recall.13, 14, 15 The recent development of wearable cameras with a point-of-view lens may provide a complementary objective measure to assist self-report, as these devices can record food consumption objectively and passively; which was previously only possible using a trained observer. Such devices are common place in adventure sports, but have recently been integrated directly into smart glasses (for example, Google Glass) for everyday use, to complement and enhance smart phone features. Thus, the rapid development and use of wearable cameras in society may provide new opportunities to obtain objective dietary intake data.16

Arab et al.17 first demonstrated the feasibility of wearable cameras for dietary assessment in using a customised mobile phone worn around the neck that captured images automatically every 10 s. The images were sent to an internet server automatically and assisted participants self-report dietary intake using a web-based 24-h recall (Image-DietDay), which revealed promising results;18 however, the cameras narrow field of view and insufficient battery life were noted limitations.

The SenseCam is a wearable camera worn around the neck and captures wide-angle point-of-view images passively, every 20–20 s, in response movement (tri-axil accelerometer), heat (infrared heat sensor) and light (light intensity sensor) sensors.19 SenseCam and other similar devices can capture images automatically over entire days, weeks or years, and are used to create digital life-logs. Life-logging is the digital capture and storage of personal data with the aim to record complete and searchable personal digital archives to assist people with everyday tasks, and/or remember past events.20, 21, 22 Thus, wearable cameras, such as SenseCam, are used to assist people with memory impairment, as the captured images provide a powerful cue for memory recall,23, 24 but have also been used in health research,25 to enhance self-reported measures of physically active travel,26, 27 sedentary behaviour,28 physical activities29 and as a supplement to food records.30

Wearable cameras may also enhance the 24-h dietary recall by providing visual prompts to improve recall of food and beverage consumption. Therefore, we conducted a feasibility study to evaluate the potential of the SenseCam to assist an interviewer-administered 24-h dietary recall. The study was designed to assess the degree to which the images assisted and changed the participants’ self-reported dietary intake. Specific objectives were to determine the effect viewing the images had on self-reported energy and nutrient intakes.

Materials and methods

A convenience sample of 13 healthy adults between the age of 18 and 65 years were recruited through an advertisement posted on notice boards in Oxfordshire, United Kingdom. Participants were excluded if they followed strict dietary regimes (for example, vegan) to ensure a wide range of foods were captured, or were unable to complete usual activities of daily living. Participants were provided a brief training session and information sheet, explaining how to use and wear the SenseCam. The camera is very simple to operate and only requires the participant to turn the device on for continuous operation. A privacy button can be activated to temporally stop image capture when required (SenseCam automatically starts capturing images again after 7 min). Participants were instructed to wear the SenseCam for two full days while continuing their usual daily activities. The 2-day period allowed participants to become familiar with using the SenseCam and participants were informed their diet would be assessed using a standard method for nutrition research. On day 3, participants had their diet assessed over the previous 24 h (day 2) by a trained dietitian (LG) using an interviewer-administered multiple pass 24-h dietary recall (MP24), based on the United States Department of Agriculture multiple pass method.14 A portion size guide and standard household measures were used to assist participants’ self-reported dietary intake.31 After the MP24, participants were given the opportunity to privately screen the captured images using a freely available SenseCam browser32 and instructed to delete any photos they wished. This procedure followed ethical guidelines for SenseCam research to ensure the privacy of the participants was maintained (some images may contain private and/or sensitive content). The remaining SenseCam images were viewed by participants and dietitian together to identify all eating episodes and to confirm or modify details reported in the initial 24-h recall (MP24+SenseCam). Dietary details recorded in the MP24 were re-stated, but the information was not scrutinised and no changes were suggested by the dietitian to reduce the possibility of interviewer-bias. However, unreported foods visible in the images were queried and the participant confirmed or modified their intake accordingly. Unreported foods, changes to portion size and other misreporting errors (for example, exchanging or removing foods) were grouped by the following food categories: breads and cereals; beverages (including milk; excluding water due to no energy content); fruit and vegetables; meats and dairy (for example, cheese); biscuits, sweets and snacks; condiments (spreads, sauces, dips and dressings). After completion of the MP24+SenseCam (SC) participants completed a brief feedback survey to explore the user experience of wearing the device. The survey used seven-point Likert scales or categorical scales with open text response sections. The study was approved by the Central University Research Ethics Committee University of Oxford (Ref: SSD/CUREC1A/12–008) and the University of Auckland Human Participants Ethics Committee (Ref: 7942).


Foods and beverages were analysed using nutrient analysis software WISP (Tinuviel Software, V 3.0, Warrington, UK). Data analysis was performed using SPSS Statistics (V 20.0; IBM, Armonk, NY, USA). A paired t-test was used to compare the differences in self-reported energy and nutrient intakes between the MP24 and MP24+SC. Participant characteristics were described with summary statistics. Statistical significance was set at α0.05.



Thirteen participants were recruited but ten participated in the study (8 males and 2 females), aged 33±11 years with a mean BMI 25.9±5.1 kg/m2 were included in the analysis (Table 1). Two participants were excluded due to a technical fault and loss of images, and one was excluded for protocol non-adherence (SenseCam infrequently worn). The participants were predominantly healthy and physically active (excluding one participant with obesity), all had completed a bachelor’s degree or higher, and all were currently employed in professional positions. The excluded participants (1 male, 2 females) were slightly older 39±13 years but had a similar BMI and education.

Table 1: Characteristics for the 10 participants included in the analysis.

Dietary recall

Energy and macronutrient intakes are presented in Table 2. The MP24+SC resulted in a significantly greater reported mean energy intake than the MP24 alone (12 888±2977 kJ vs 11 455±2099 kJ, P=0.02) with a mean difference of 12.5%. Significantly higher reported intakes of protein, total fat, saturated fat and mono-unsaturated fat were also evident with the MP24+SC compared with the MP24.

Table 2: Energy and macronutrient intakes from one interviewer-administered MP24 and the MP24 with the assistance of SenseCam images (MP24+SC)

Information on unreported foods, changes to portion size and misreported foods detected by SenseCam are presented in Table 3. A total of 41 unreported foods in the MP24 was revealed by viewing the SenseCam images (MP24+SC), 241 vs 282 total foods, respectively (additional 17% foods items). The additional foods were from all food categories and had energy contents from 0 kJ to 1820 kJ. The beverage category had greatest number of unreported items (n=9), but provided less energy than other groups excluding fruits and vegetables. Unreported breads and cereals, meats and dairy, and biscuits sweets and snacks accounted for the greatest additional energy. The condiments category had the fewest unreported foods (n=4) but on average these additions each provided substantial additional energy (773±737 kJ).

Table 3: Energy provided from unreported foods, change to portion size or misreported foods revealed from one SenseCam-assisted interviewer-administered MP24 (MP24+SC).

There were only eight changes to portion size (five increases in portion size, three reductions in portion size) that provided little effect on energy intake overall (−259 to 189 kJ). Misreporting errors, that is, foods incorrectly reported in the initial MP24 and modified after image review ( for example, salmon exchanged for chicken) had a greater effect on energy intake (−787 to 165 kJ) compared with changes to portion sizes and the majority of misreporting errors were in the fruits and vegetables category (n=8).

Feedback survey

Results from the participant feedback survey are presented in Table 4. The survey revealed participants found the images helped them remember foods they had forgotten about (median score 7) and were able to provide more accurate information (median score 7) but had less impact on helping them to remember extra details of how foods were prepared or purchased (median score 4). Participants indicated wearing the SensCam and using it to assess their diet was a low burden but three sometimes felt uncomfortable in public situations, such as riding the bus or purchasing items (open text feedback). Wearing the SenseCam never (50%), seldom (10%) or sometimes (40%) affected dietary behaviours. Qualitative feedback revealed five participants were conscious of the device during food consumption and some considered making healthier food choices and may have eaten fewer snacks, but the degree of behaviour change was not clear. Participants indicated that about 1 week would be the maximum time they would wear a SenseCam.

Table 4: Participant evaluation of the SenseCam-assisted 24-h recall (MP24+SC).


This study demonstrated that a wearable camera can be used to enhance the 24-h dietary recall by revealing unreported foods and misreporting errors. Overall, reviewing the images revealed 17% additional food items and increased reported energy intake by 12.5%.

As a novel study, no direct comparison was possible, however, the 24-h recall method typically provides energy intake data that is under-reported by approximately 8–24% compared with doubly labelled water.2 Thus, the increase in self-reported energy intake, revealed from viewing passively captured images, indicates that wearable cameras may help to reduce levels of under-reporting. Although, additional and more rigorous testing of wearable cameras is necessary.

O’Loughlin et al.30 conducted a similar study (n=34) in Ireland that used SenseCam to enhance a 1-day food kept record by trainee Jockeys, Gaelic football players and university students, which revealed a 10–18% increase in energy intake after image review. No data regarding what accounted for the increase was reported but unreported foods and changes to portion size were indicated by the investigators. Arab et al.18 also reported promising results using images captured automatically on mobile phones to assist a self-reported 24-h recall (Image-DietDay) but the study did not classify how the images assisted recall.

Other image-assisted methods to assess dietary intake are also in development. Sun et al.33 have described a customised wearable camera system for objective dietary assessment. Similar to the SenseCam the device is worn at chest height but aims to use automated image analysis to objectively assess intake; however, at the time of writing no dietary intake data has been presented. Other research investigating image-assisted electronic food records on smart phones (or cameras) is encouraging.34, 35, 36 However, this approach is fundamentally different as participants are still required to initiate data collection, thus a similar burden to traditional self-report methods is placed on the participant to actively record food consumption. Although, it is likely future developments of technology will allow these two image-assisted approaches to complement each other in an effort to objectively assess dietary intake.

The use of images captured from a wearable camera to assist dietary recall in this study revealed a diverse range of unreported foods throughout the day (only one food item revealed at supper). Some foods, including tea, black coffee, mandarins and peas provided little energy but other foods including avocado on toast, potato crisps, cheese and lemon drizzle cake provided considerable additional energy. The wide variety of foods and the fact that both healthy and unhealthy foods were under-reported indicated under-reporting was likely random. The degree of under-reporting revealed may reflect the physically active sample, as increased levels of physical and varied lifestyle have been shown to reduce the accuracy of self-reported intake.30, 37 Conversely, this finding may simply reflect widespread under-reporting that plagues traditional dietary assessment methods.2

The misreported foods, as identified by participants (that is, removed or corrected following image review), further exposed and reduced random error associated with existing dietary assessment methods. The errors were predominately single food items (for example, orange corrected to banana) but some involved multiple foods within the same eating episode (small salmon fillet and salad corrected to chicken breast, jacket potato, with salad and caeser dressing). Fewer changes to portion size were probably attributed to poor image quality (foods in low-light, food partially obscured or undesirable angles) and the decision not to question the participant’s self-report. This reduced the potential for interviewer bias and any change to portion size may lead to a correlated error (as the true portion size remained unknown).38, 39, 40

Participant feedback indicated viewing the images assisted with portion size estimation but also revealed wearing the camera may have affected dietary behaviours; thus may not be a true representation of usual intake. Similar participant feedback was reported by Arab et al.24 with mobile phones worn around the neck to capture dietary intake. Nearly all participants found the images helpful during the Image-DietDay 24-h recall, but some were self-conscious in public, which may have affected normal behaviour.

In the present study, the magnitude of behaviour change was not clear but social desirability and other psychosocial factors including social approval and fear of negative evaluation affect dietary behaviours41, 42, 43, 44, 45 and self-reported dietary intake.9, 46, 47 Wearable cameras may exacerbate behaviour change as devices are visible to others and objectively record food consumption in real-time. Moreover, the ability to omit foods during subsequent self-reported dietary intake is minimised as all food are recorded (unless the device was switched off), therefore, may further affect behaviours, especially in groups prone to misreporting including women, overweight and obese.48, 49, 50

It was observed that often participants’ memory was prompted to recall additional foods before the specific image revealing the unreported food was displayed. The mechanism that acts to enhance memory recall is unknown, but it is suggested images are possibly captured at the same moment memory encoding is taking place,22 and functional magnetic resonance imaging has revealed regions of the brain where memories are processed and stored are activated when viewing SenseCam images, but not activated reviewing a written diary of the same events.51 Thus, viewing the images provides both a powerful cue for memory recall and an objective record of the day’s events if memory is not prompted. In combination, this provides participants a better opportunity to correctly self-report dietary intake and reduce under-reporting.

The strength and novel feature of this study was the direct comparison between traditional self-report and self-report with the assistance of SenseCam. Often studies identify under-reporting but can only suggest why or how under-reporting has occurred. The SenseCam and other wearable cameras allow under-reporting to be explored in detail providing insight into the cause of under-reporting, previously not feasible in a free living situation.

However, the SenseCam has some limitations. The imaging frequency of 20–30 s was not always sufficient to capture all foods consumed, posture and body shape can affect lens angle resulting in non-useful images (more commonly in women), and poor image quality in low-light environments were limitations of the SenseCam device with respect to assessing dietary intakes. Further, the device malfunctioned for two participants that resulted in a loss of images and one participant was non-compliant (camera worn infrequently). Similar reliability and compliance issues were noted by O’Loughlin et al. when the SenseCam was to enhance written food records.30 These issues would need to be resolved before wide-scale use of wearable cameras in dietary assessment.

This study also had limitations. A sole dietitian (LG) conducted the dietary recalls without a gold-standard reference method, thus interviewer error cannot be ruled out, and the sample size was small and primarily consisted of well-educated active adult men. Therefore, the findings of this pilot may not apply to women or other groups, and should to be repeated in an adequately sized sample to confirm these findings.

In summary, the use of wearable cameras appears to be a promising method to enhance self-reported dietary assessment. Viewing the images revealed a number of unreported foods, and other misreporting errors, which increased self-reported energy intake. Overall, this provided a clear indication to how wearable cameras may reduce levels under-reporting. Further research using gold-standard methods, such as doubly labelled water, is required to validate the use of wearable cameras in dietary assessment, and the degree to which wearable cameras affect dietary intake and behaviours should be explored.


  1. 1.

    , . Dietary assessment methodology. In: Coulston A, Boushe C (eds). Nutrition in the Prevention and Treatment of Disease. Academic Press: San Diego, 2001, pp 3–30.

  2. 2.

    , . Markers of the validity of reported energy intake. J Nutr 2003; 133 (Suppl 3), 895S–920S.

  3. 3.

    , , , , , et al. Accuracy of energy intake data estimated by a multiplepass, 24-hour dietary recall technique. J Am Diet Assoc 2000; 100: 303–311.

  4. 4.

    , , , , , et al. 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.

  5. 5.

    , , , , . Dietary methods research in the third National Health and Nutrition Examination Survey: underreporting of energy intake. Am J Clin Nutr 1997; 65: 1203S–1209S.

  6. 6.

    , , . An evaluation of energy intakes in the 1983 Australian National Dietary Survey of Adults. Eur J Clin Nutr 1993; 47: 604–606.

  7. 7.

    , , . Under-reporting of energy intake in the 1997 National Nutrition Survey. N Z Med J 2004; 117: 1–11.

  8. 8.

    , , , . In-person vs telephone-administered multiple-pass 24-hour recalls in women: validation with doubly labeled water. J Am Diet Assoc 2000; 100: 777–783.

  9. 9.

    , , , , , et al. Psychological measures of eating behavior and the accuracy of 3 common dietary assessment methods in healthy postmenopausal women. Am J Clin Nutr 2000; 71: 739–745.

  10. 10.

    , , , , , et al. Validation of a self-administered food-frequency questionnaire administered in the European Prospective Investigation into Cancer and Nutrition (EPIC) Study: comparison of energy, protein, and macronutrient intakes estimated with the doubly labeled water, urinary nitrogen, and repeated 24-h dietary recall methods. Am J Clin Nutr 1999; 70: 439–447.

  11. 11.

    , , , , . Need for technological innovation in dietary assessment. J Am Diet Assoc 2010; 110: 48–51.

  12. 12.

    , , , , , . Review and evaluation of innovative technologies for measuring diet in nutritional epidemiology. Int J Epidemiol 2012; 41: 1187–1203.

  13. 13.

    , , , . The USDA Automated Multiple-Pass Method accurately estimates group total energy and nutrient intake. J Nutr 2006; 136: 2594–2599.

  14. 14.

    , , , , , et al. The US Department of Agriculture Automated Multiple-Pass Method reduces bias in the collection of energy intakes. Am J Clin Nutr 2008; 88: 324–332.

  15. 15.

    , , , , , et al. The Automated Self-Administered 24-Hour Dietary Recall (ASA24): a resource for researchers, clinicians, and educators from the National Cancer Institute. J Acad Nutr Diet 2012; 112: 1134–1137.

  16. 16.

    Google 2013, (cited 16 May 2013). Available from .

  17. 17.

    , , , , . Eight self-administered 24-hour dietary recalls using the Internet are feasible in African Americans and Whites: the energetics study. J Am Diet Assoc 2010; 110: 857–864.

  18. 18.

    , , , , . Feasibility testing of an automated image-capture method to aid dietary recall. Eur J Clin Nutr 2011; 65: 1156–1162.

  19. 19.

    , , , , , et al. SenseCam: a retrospective memory aid. In: Dourish P, Friday A (eds). UbiComp 2006: Ubiquitous Computing; 2006 Sep 17-21; Springer-Verlag: Orange County, California, USA, 2006, pp 177–193.

  20. 20.

    , , , , (eds). Passive capture and ensuing issues for a personal lifetime store. Proceedings of the 1st ACM workshop on continuous archival and retrieval of personal experiences. 15th October, Microsoft, New York, USA, 2004.

  21. 21.

    . Total recall [life recording software]. Spectrum, IEEE 2005; 42: 24–30.

  22. 22.

    , , . SenseCam: a wearable camera which stimulates and rehabilitates autobiographical memory. Memory 2011; 19: 685–696.

  23. 23.

    , , , , , et al. SenseCam improves memory for recent events and quality of life in a patient with memory retrieval difficulties. Memory 2011; 19: 713–722.

  24. 24.

    , , , , , et al. The use of a wearable camera, SenseCam, as a pictorial diary to improve autobiographical memory in a patient with limbic encephalitis: a preliminary report. Neuropsychol Rehabi 2007; 17: 582–601.

  25. 25.

    , , , , , et al. Wearable Cameras in Health: The State of the Art and Future Possibilities. Am J Prev Med 2012; 44: 320–323.

  26. 26.

    , , , , , et al. Evaluating the feasibility of measuring travel to school using a wearable camera. Am J Prev Med 2012; 2012: 5.

  27. 27.

    , , , , , . Can we use digital life-log images to investigate active and sedentary travel behaviour? Results from a pilot study. Int J Behav Nutr Phys Act 2011; 8: 44.

  28. 28.

    , , , , , et al. Using the SenseCam to improve classifications of sedentary behavior in free-living settings. Am J Prev Med 2013; 44: 290–296.

  29. 29.

    , , , , , et al. Use of wearable cameras to assess population physical activity behaviours: an observational study. Lancet Public Health Science Conference 2012; 380: S35.

  30. 30.

    , , , , . Using a wearable camera to increase the accuracy of dietary analysis. Am J Prev Med 2013; 44: 297–301.

  31. 31.

    . Food portion images for dietary assessement: graduated food portion photos of Australian foods. The University of Wollongong: Wollongong, 2012.

  32. 32.

    , , . Automatically assisting human memory: a sensecam browser. Memory 2010; 19: 785–795.

  33. 33.

    , , , , , et al. A wearable electronic system for objective dietary assessment. J Am Diet Assoc 2010; 110: 45–47.

  34. 34.

    , , , , , et al. Validation of photographic food records in children: are pictures really worth a thousand words&quest. Eur J Clin Nutr 2009; 63: 1025–1033.

  35. 35.

    , , , , , et al. Novel technologies for assessing dietary intake: evaluating the usability of a mobile telephone food record among adults and adolescents. J Med Internet Res 2012; 14: e58.

  36. 36.

    , , , . Trial of a mobile phone method for recording dietary intake in adults with type 2 diabetes: evaluation and implications for future applications. J Telemed Telecare 2011; 17: 318–323.

  37. 37.

    , , , , . Relationship of high energy expenditure and variation in dietary intake with reporting accuracy on 7 day food records and diet histories in a group of healthy adult volunteers. Eur J Clin Nutr 2002; 56: 358–367.

  38. 38.

    , , , , , et al. Portion size estimation and expectation of accuracy. J Food Compost Anal 2006; 19: S14–S21.

  39. 39.

    , , . Food photography. I: The perception of food portion size from photographs. Br J Nutr 1994; 72: 649–663.

  40. 40.

    , , , , , et al. Accuracy in the estimation of food servings against the portions in food photographs. Eur J Clin Nutr 2007; 62: 674–681.

  41. 41.

    , , , , , . Psychosocial predictors of energy underreporting in a large doubly labeled water study. Am J Clin Nutr 2004; 79: 795–804.

  42. 42.

    , , . Behavioral and body size correlates of energy intake underreporting by obese and normal-weight women. J Am Diet Assoc 1999; 99: 300–306.

  43. 43.

    , , , , , et al. Discrepancy between self-reported and actual caloric intake and exercise in obese subjects. N Engl J Med 1992; 327: 1893–1898.

  44. 44.

    , , , , . Social desirability bias in dietary self-report may compromise the validity of dietary intake measures. Int J Epidemiol 1995; 24: 389–398.

  45. 45.

    , , , , , et al. Gender differences in social desirability and social approval bias in dietary self-report. Am J Epidemiol 1997; 146: 1046–1055.

  46. 46.

    , , , , , . 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.

  47. 47.

    , , , . Assessment of selective under-reporting of food intake by both obese and non-obese women in a metabolic facility. Int J Obes 1998; 22: 303.

  48. 48.

    , , , . Validity of reported energy intake in obese and nonobese adolescents. Am J Clin Nutr 1990; 52: 421–425.

  49. 49.

    , , , , , et al. High levels of energy expenditure in obese women. BMJ 1986; 292: 983.

  50. 50.

    , , . Literacy and body fatness are associated with underreporting of energy intake in US low-income women using the multiple-pass 24-hour recall: a doubly labeled water study. J Am Diet Assoc 1998; 98: 1136–1140.

  51. 51.

    , , , , , et al. The neural basis of effective memory therapy in a patient with limbic encephalitis. J Neurol Neurosurg Psychiatry 2009; 80: 1202–1205.

Download references


We thank the participants and the British Heart Foundation Heart Promotion Research Group, University of Oxford, for their expertise, guidance and technical support for SenseCam research.

Author information


  1. National Institute for Health Innovation, University of Auckland, Auckland, New Zealand

    • L Gemming
    •  & C Ni Mhurchu
  2. British Heart Foundation Health Promotion Research Group, University of Oxford, Oxford, UK

    • A Doherty
    •  & P Kelly
  3. Epidemiology and Biostatistics, University of Auckland, Auckland, New Zealand

    • J Utter


  1. Search for L Gemming in:

  2. Search for A Doherty in:

  3. Search for P Kelly in:

  4. Search for J Utter in:

  5. Search for C Ni Mhurchu in:

Competing interests

The authors declare no conflict of interest.

Corresponding author

Correspondence to L Gemming.