Nutrition at retirement age

By M.  DIANE HOLDSWORTH  and LOUISE DAVIES, Gerontology Nutrition Unit,

Queen  Elizabeth  College,  University  of  London,  Campden Hill,  Kensington,

London W8  7AH

Why  is  it  important  to  investigate  the  nutrition  and  wellbeing  of  men  and

women at the age of  retirement from work? Surely by this time their meal patterns

and  food choice  are  already  set,  and  it would  seem more  important  to  influence

those at a younger age.

Nevertheless, retirement from work  is  a  time when food  habits may change:

former employees no longer eat at their place of work, food shopping patterns may

alter (possibly because of  a  fall in income), meal times and meal companions could

be different. If  steps towards healthier eating could be encouraged at this time of

change, habits which are conducive to positive health could be built  into this new

lifestyle. Moreover,  at  retirement from  work people  are  likely  to  be concerned

about their present and future health, therefore any programme for preventive care

implemented  at this  stage in  life could  reach  a target group  which  is  highly

motivated.

In a study of  the nutrition of  people approaching retirement, we found that  it is

not  easy  to conduct a survey among people who are still at work. The permission

of  the  employers for time  off  with  pay needs  to  be obtained.  The  employees

themselves  may reckon  that  they are too busy  to  co-operate. Another major

difficulty  was  that  most  firms  would, understandably, refuse permission for

outsiders to contact their  staff direct. The potential sample had to be approached

via  the personnel officers. Due  to this  it was not possible to find  the  reasons for

refusal to participate. Some may have been daunted by  the prospect of  losing time

off  work, or having a medical screening; on the other hand, others may have been

attracted by such requests.

The sample and plan of  research

Firms need  to be extremely large in  order  to have  sufficient numbers retiring

each year.  Eventually  employees  from two  firms cooperated;  they had  to  be

unconnected with  the  food  industry  in case  this  should affect  their nutritional

knowledge  or  food choice.  The  total  sample  of  213  men  and  women  closely

approximated  the  national average for  socio-economic groups and  the sex  ratios

within the general work force.

At  stage  I  of  our longitudinal survey,  base-line  data  were  collected

approximately  I  year before  the  subjects’ date  of  retirement. Approximately  6

months  later, nutritional  counselling  was  given  to  half  the  sample (Davies  &

Holdsworth,  1983).  In the current  longitudinal phase  of  the  survey, subjects are 304  SYMPOSIUM  PROCEEDINGS  I984

returning  for  repeat  investigations  not less  than  6 months  after  their retirement

date.

Within  the  research  programme  the  same individuals are being studied  at  two

vital periods  in  their  life,  and changes  are being monitored.  The present paper

presents findings from the completed base-line pre-retirement  data.

For  an  investigation  of  this age-group  it was considered  important  to examine

such  topics  as  physical  activity,  past and present smoking  habits,  alcohol

consumption,  stress factors and  attitudes  to retirement,  as well  as the nutritional

concerns such  as  shopping habits, meal  patterns  and  dietary  intake.  In  order  to

obtain such full details,  questionnaires  were designed  to  flow  as  a  guided

conversation  lasting approximately  I .  5  h.  In addition,  frequency of  consumption

lists were designed. The consumption of  specific foods was coded as daily, weekly,

fortnightly, monthly, occasionally, rarely or never. Thus some idea of  foods eaten

throughout  the  year could  be obtained.  This  information is  not necessarily

indicated in the normal dietary recalls or  I  week’s weighed dietary records.

It  was necessary  to  use  a standardized  technique for measuring stress.  We

therefore chose the Goldberg (1978) General Health Questionnaire which had been

extensively researched  to  identify somatic symptoms, anxiety and insomnia, social

dysfunction and severe depression.  It consists of  up  to sixty  questions with  four

scales of  response for each question.

All  the previously mentioned  information was obtained from the total sample of

213 men  and women.  In addition,  the  I  13  subjects from firm I  agreed to further

investigations which consisted of:

I.  A  health screening  at  the British  United  Provident Association (BUPA)

private medical  centre  in  London.  This  included  routine  medical  blood

analyses.

2.  Further biochemical analyses for assessment of nutritional status.

3.  A  7 d weighed and recorded dietary intake.

As with  all  nutritional  surveys, much  time  needs  to be  spent encouraging  the

sample  to meet  the demands  imposed. Public  relations, personal encouragement,

motivation and written instructions are important in surveys of  this type.

The dietary  weighing and recording was  taught during a picnic-style lunch at

the  medical centre.  Checks were made  on  the  validity  of  the  I  week’s weighed

records  by  using simultaneously four  methods  of  estimating  food  intakes

(Holdsworth et  al. 1984). These methods were: (a)  I week’s weighed and recorded

dietary intakes examined  in  conjunction with  (b) frequency of  consumption lists;

(c) recall  of  food  and  drink customarily  consumed  (this was  not  the  usual  24  h

recall but  rather an example of  eating and drinking style throughout a normal day);

(d) carefully-designed questions within the main questionnaire on the consumption

of, for example, alcohol, butter, margarine, breads. and the use of  the salt cellar.

Generalfindings compared with recommendations

The present paper compares some of our general findings with recommendations

made  in  recent  reports  such  as  the  National  Advisory  Committee  on  Nutrition Vol. 43  Nutrition and food problems of  dzywent age groups  305

Table  I.  The choice of  butter or margarine  with bread:  men aged  63-64 years

and women aged 58-59 years (n 2 I 3)

Percentage response to question:

Do you mostly have butter or

margarine with bread?

- Men Women

Butter  54  51

Margarine  39  38

Both  6  9

Neither  I  2

Education  (1983) discussion  document;  the  ‘Obesity Report’ (Royal College  of

Physicians, 1983) and  the report on ‘Alcohol and Alcoholism’ (Special Committee

of  the Royal College of  Psychiatrists,  1979). Our data included general information

about the shopping and eating styles of  the total sample of 213, i.e.  136 men and 77

women.  This  is  the  generation who  were  not brought up in  the  age  of

supermarkets but it was obvious that new shopping styles, at least in London and

its  outskirts,  had  been  adopted:  93%  bought  their  food  from supermarkets

although 4170 also regularly used a small grocer.

Food choice, however, may tend  to cling to traditions  laid down in youth, and

this may prove of  importance in implementing future nutritional guidelines. It may

have been the habit  of  a lifetime that kept up the popularity of  butter as shown in

the response to the question in Table  I.

However,  the tradition  of  the Great  British Breakfast does  seem  to  have

changed. Concern  is  sometimes expressed at  the  number  of  schoolchildren who

skip breakfast (Bender et  al.  1972). The men  and women  in our  study were old

enough  to be grandparents.  They were  asked whether they  ate breakfast,  either

cooked or uncooked. Table 2  indicates that  as many as 25%  did not eat breakfast

every day.

Many  of  the  subjects were weight-conscious:  45%  reckoned  that  they had  a

tendency  to overweight. To the question, ‘in the past year have you been altering

what  you  eat  in  order  to control your weight?’ 35%  responded yes, representing

29%  of  the men and 4670 of  the women. This may have influenced the fact that so

Table 2.  Breakfast  habits of men aged 63-64 years and women aged 58-59 years

(n 213)

Percentage eating

breakfast

&

Men Women

Every day  74  75

Weekdays only  2  0

Weekends only  10  5

Sometimes  4  8

Never  I0  12 306  SYMPOSIUM  PROCEEDINGS  '984

Table 3.  Sugar taken in tea and coffee by men aged 63-64 years and women aged

58-59years  (70 oftotal) (n 213)

Tea  Coffee

0.5 teaspoon  5  5

I  teaspoon  '7  24

z teaspoons  I4  18

3  teaspoons  2  3

Sone  62  50

many  of  them drank tea  and  coffee without  sugar.  The number  of  teaspoons  of

sugar taken  in  tea and coffee is shown  in Table 3.  For coffee 4070 of  the men and

6570 of  the women did not add sugar.

Given  this  type  of  response  perhaps  one should  query  the  advisability  of

encouraging  manufacturers  to  switch  from sucrose  to  an artificial sweetener.  It

might be more acceptable to take advantage of  the capacity of people to live with a

less-sweet  palate and  to make products less sweet by  reducing  the  sugar content.

Certainly artificial sweetners were not  popular  in drinks; only  16% of  the  sample

used them regularly, even though so many were trying to slim.

It may  be  relatively easy  at  the  age of  retirement  to implement  the guidelines

recommending increased consumption of  fruits and vegetables, and the importance

of maintaining physical activity. Even before retiring from work, as many as 4670

of  the total sample were growing fruit or vegetables in their gardens or allotments.

Findings fromjrm I  including I  week's weighed dietary records

From  the medical screening  and  I  week's weighed  dietary records  of  the  113

subjects  from  firm  I,  it  was  possible  to  study the  prevalence  of,  for  instance,

hypertension, obesity,  hypercholesterolaemia. We also wished  to  answer  the

following  questions. Did their diet quantitatively  show evidence  of  high  fat,

including animal fats, sugar, salt and alcohol intakes? Were they already having the

low-potassium, low-fibre diets common amongst some elderly?

Raised  blood pressure. Raised blood pressure was prevalent among our sample,

especially the men: 4170 had  resting  diastolic  blood  pressures  above 90 mm Hg,

3070 were above 95 mm Hg; for women the corresponding values were 19  and 7%

(Fig.  I). Of  the men  in  the  sample, 257'  could be classified as hypertensive, with

both a resting systolic blood pressure of  >160 mm Hg and a resting diastolic blood

pressure of >90 mm Hg.

Obesity.  Using  the  quetelet  index for body  mass  (weight/height2), Garrow

(1981) defines those with an index of  25-29.9  as having grade I obesity, and 3-40

as having grade  I1 obesity, which  is clinically serious.  In our sample, 5070 of  the

men had  an  index  of  between  25  and  29.9 and  1170 an index  of  over  30.  For

women  in our  sample, corresponding values were 40 and 670  respectively. Fig. 2

demonstrates the wide ranges from the very thin to the obese. Vol. 43  Nutrition and food problems of dryerent age groups  307

.

Fig.  I.  Resting diastolic and  systolic blood pressure in a sample of men and women approaching

retirement from work. (W) Men aged 63-64 years, n 46; (0) women aged 58-59  years, n 67.

Fig. 2.  Quetelet index (weight/height2) of men and women approaching retirement from work. (a)

Men aged 63-64 years, n 46; (v) women aged 58-59  years, n 67. 308  SYMPOSIUM  PROCEEDINGS  I984

Raised  serum  cholesterol.  Serum  cholesterol values  are  shown in  Fig.  3.  The

mean value for  the men was  6.72 mmol/l,  and  for  the women  7.16 mmol/l.  An

average value  below  5.2 mmol/l  (2  g/l)  is  considered  as  low  as  is  likely  to be

achieved in the UK (World Health Organization, 1982).

A  serum  cholesterol concentration  of  4-4.5  mmol/l  (I '6-1.8  g/l)  is associated

with  a  very low  risk  of  coronary  heart  disease (American  Health Foundation,

1979). Only one of our subjects fell into this category.

Dietary fats. Our  pre-retirement  sample was  from  Southeast England, where

the  incidence of  coronary  heart  disease  is high.  There  is a strong  consensus of

opinion  that  the  incidence  of  coronary  heart  disease could  be  lowered  by  the

reduction  of  total fat  in  the  diet  to  3 ~ 3 5 %   of  total  energy (Royal  College of

Physicians and British Cardiac Society, 1976). A recent, relatively small survey has

suggested  that  in  the diets of  men within a wide age group  living  in  Southeast

England, approximately  3870 of  their total energy was derived from fats (Bingham

et  al.  1980). For  the men in our pre-retirement sample,  the  energy derived from

fats  was  4170, for  the  women  43%.  However, energy derived  from  fats was

recalculated at 4370 for men and 44% for women when alcohol was excluded from

the total energy intake.  The percentage  of energy from carbohydrate and protein

was  39.1  and  15.8 respectively.  These  values  correspond with the  mean daily

intakes of energy and nutrients shown in Table 4.

O  Men  L- Women

Fig. 3.  Serum cholesterol values (mmoV1) in a sample of men and women approaching retirement

from work. (m) Men aged 63-64 years, n 45; (0) women aged 58-59  years, n 64. Vol. 43  Nutrition and food problems of dizferent age groups  309

Protein  Fat  Carbohydrate Alcohol

RDA, recommended daily amounts (Department of Health and Social Security, 1979).

Approximately 7570 of  the  ingested  fats came from animal sources: among the

contributions  to  total  fat were meat  and meat dishes 24.2Oj0, milks 8.970, other

dairy products (including butter) 22,970.

There  were several  blatant  examples  of  a  high  fat  intake  resulting from

misinterpretations  of  low-carbohydrate slimming diets. In  the diet  record  of  one

woman  was written  ‘I  Energen  roll  spread  thickly with butter’;  another woman

recorded ‘Nimble bread 5/6 oz and then as much as 2/3 oz butter and  I+ oz jam’.

Sugar and salt intakes.  In the  sample, sugar contributed  12.270 of  total energy

intake. Those who routinely sweetened  their beverages could be advised simply to

reduce the amount.

It  is  recognized  that  food  manufacturers  have  a  responsibility  in  helping  to

diminish sodium intakes. However,  our  results  also  indicate  the  need  for health

education  for  people  at  retirement  age  on  the  use  of  table salt.  The majority

customarily added salt to the food once it was on the plate, as shown in Table 5.

Potassium  and  dietary $fibre.  The  preliminary findings indicated, for men,  a

weak  inverse  relationship between  dietary  K  intake  and diastolic blood  pressure

(Fig.  4). This  finding  was not statistically  significant (correlation coefficient

-0.21). However,  it  is noteworthy that the men who were hypertensive (indicated

in  Fig.  2) were among those  with  the  lowest  dietary K  intakes. The mean daily

dietary K  intake was:  hypertensive  men  64 mmol (2.5 g);  those  classified as not

hypertensive  87 mmol  (3.4 g).  For  the  men in  our  sample,  fruit  and vegetable

Table 5.  Salt  added to food on  the plate by men  aged 63-64 years and women

aged 58-59 years (7’ of  total) (n I I 3)

Always  50

Often  I 1

Occasionally  ‘5

Seldom  I 1

Never  ’3 1507

130  llyl

SYMPOSIUM  PROCEEDINGS

 

Dietary K  intake (S/d)

Fig 4.  Dietary potassium intakes (g/d) and diastolic blood pressure (mm Hg) of men approaching

retnement from work (aged 63-64 years, n 46). (M), Men with  systolic blood pressure >160 mm Hg.

consumption  was negatively correlated  with  systolic  blood  pressure (correlation

coefficient -0.46, P<o.oor) (Fig. 5).

Also of  concern in the UK are low fibre intakes; the average British diet contains

20 g dietary  fibre/d  (Southgate  et  al.  1978) and  it  has  been suggested  that  for

adults  this  should be increased  to  30 g/d  (National  Advisory  Committee  on

Nutrition Education, 1983). From our results the mean daily intake was only 18 g:

men  19 g (range 8-36 g); women  16 g (range 7-35 g). Of this sample, 9470 were

below the recommendation of  30 g dietary fibre/d.

In a survey on people about to retire, and thus alter their lifestyle, it is important

to examine factors such as alcohol, smoking and exercise.

Alcohol.  The Special Committee  of  the  Royal College of  Psychiatrists (1979)

suggests  that  an  intake  of  4 pints  of  beer daily,  or  its equivalent,  constitutes a

reasonable guideline for the upper safe limit of drinking. Wilson (I 980), conducting

research on drinking in England and Wales, indicates heavy drinkers as those who

consume over  50  units per week  for men  and over 35 units per week  for women.

This represents  for men  3.5 pints of  beer (60 g alcohol) and for women 2.5 pints

(40 g alcohol) daily, or its equivalent. It should be emphasized  that these levels are

not  intended as suggestions for routine consumption.

The  I  week's  weighed  records  of  our  subjects  showed widely varying  alcohol

intakes, ranging from the expected zero to the high intake for one gentleman of  IOO

g daily. Vol. 43  Nutrition and food problems of dz.erent age groups  311

2

Fruit and vegetable consumption (g/d)

mean approaching retirement from work (aged 63-64 years, n 46).

Fig.  5.  Fruit  and vegetable consumption (g/d) and resting systolic blood  pressure  (mm Hg) of

It  was important  for  the  interviewer  to  record on  the  questionnaire relevant

comments  from  the  subjects; these comments illuminated  attitudes,  e.g.  the

heaviest drinker (subject no.  103) stated  that  ‘gin  is a woman’s drink; its alcohol

effect  on  men  is  negligible’.  When questioned  further  he gave  the  following

answers: ‘For lunch at work  I often have no  food, instead  I have  a gin. . . about

three pub measures, with a  splash of martini  to colour  it. Then, maybe, I would

have  another  such drink and sometimes a third.  If  I  am  thirsty  I have 2 pints of

beer afterwards.  At  dinner  I share a bottle  of  wine with my  wife  (she  has one

glass). In the evening I often drink a bottle of wine also’. Interviewer: ‘Do you ever

have whisky?’ Subject  no.  103: ‘No, most certainly not.  I  never  touch  the  hard

stuff’.

As could be expected his  serum glutamic oxalacetic transaminase (EC 2.6. I. I ;

SGOT) levels were markedly raised  above a normal upper limit of 50  id1 at 225

i d ,  and  his y-glutamyl  transpeptidase  (EC 2.3.2.2; GGT) activity was ninefold

greater than the  upper  limits of  normal. Of  our sample,  3%  had raised  SGOT

levels. 32%  of our total sample of 213 said they would take a drink ‘most days’. In

contrast, over half the sample were infrequent drinkers (Fig. 6).

The average weekly  intake  of  the  68Y0 of  the  sample who consumed  alcohol

during the week of  their weighed dietary records was, men I 16 g alcohol (14 units,

equivalent to fourteen  single measures of  spirits weekly) and women  78 g alcohol

(10 units, equivalent to ten single gins weekly).

Smoking  and exercise. Of  the total  sample 25%  had never smoked, and by the

age of  retirement from work  4770 had  given up  smoking. Of  these, one-fifth had 3 12

 

Alcohol consumption (g/7 d)

Fig. 6.  Alcohol consumption  during 7 d weighed and recorded dietary intakes of men and women

approaching retirement  from work. (El) Men aged 63-64  years, n 46; (0) women aged 58-59  years,

n 66.

spontaneously mentioned  an  increase  in  weight  on  giving  up  cigarettes. Of  the

28Y0 smoking at the time of  the survey, 8170 came from social classes 111, IV and

V.

In  the  longitudinal  study  the Gerontology Nutrition Unit  at Queen Elizabeth

College will be  examining  the results on  alcohol  intake and cigarette  smoking in

conjunction  with  signs  of  stress.  Over  25%  of  our  sample were  in  a  high-risk

coding according  to  their answers  to  questions  in  the  General Health

Questionnaire (Goldberg, I 978), indicating anxiety and insomnia states.

The  subjects  were asked  if  they did regular brisk physical activity such  as

walking, jogging, cycling, swimming  or  other  sports,  including more  strenuous

types of  gardening.  In a  separate question,  an  estimation was  given of  the  time

spent on such  activities each week.  For  our  total  sample of  213, 80%  spent  less

than 10 h/week on brisk activities.

This could  indicate  the  importance  of  the  National  Advisory Committee  for

Nutrition Education (1983) recommendations to keep up a greater physical activity

right through the age-groups.

These preliminary findings indicate that preventive measures need to be taken at

the earlier  stage of  retirement  from work  if our increasing elderly population  is to

enjoy a healthy retirement. Vol. 43  Nutrition and food problems of  different age groups  3'3

This research is  part  of  the Community programme under  Regulation  (EEC)

nos. 2935/79 and 271/82. The authors wish  to thank British Petroleum and North

Thames Gas for their essential co-operation, and their employees for participating

so willingly in  this detailed study.  The authors  are grateful to Dave Macfarlane,

Computer Programer, Gerontology Nutrition Unit, for his valuable assistance.

REFERENCES

American Health Foundation  (1979). Preventiwe Medicine 8,612-678.

Bender, A. E., Magee, P. & Nash, A. H.  (1972). British MedicalJournal ii,  383.

Bingham, S., McNeil, N. L. & Cummings, J. H.  (1980). British Journal ofNutrition 45,23-35..

Davies,  L.  &  Holdsworth,  M.  D.  New  Directions  in  Pre-Retirement  Education.

Stoke-on-Trent : Beth Johnson Foundation.

Department  of  Health and Social Security  (1979). Recommended Daib Amounts of Food Energy

and Nutrients for Groups of People  in the UK. Report  on Health and Social Subjects no.  15

London: H.M. Stationery Office.

Garrow,  J.  S.  (1981).  Treat  Obesity Seriously. A Clinical  Manual.  London: Churchill-

Livingstone.

Goldberg, D.  (1978). Manual of the General Health Questionnaire. Windsor: NFERPublishing

co.

Holdsworth, M. D., Davies, L. & Wilson, A.  (1984). Human Nutrition: Applied Nutrition 38A,

National  Advisory  Committee  on  Nutrition  Education  (1983).  Profosals for  Nutritional

Guidelines for Health Education  in Britain. London: Health Education Council.

Royal College of Physicians (1983). Journal ofthe Royal College ofPhysicians 17.3-58.

Royal College of Physicians and British Cardiac Society  (1976). Journal of the Royal College of

Southgate, D. A. T., Bingham, S. & Robertson, J.  (1978). Nature 274. 51-52.

Special Committee of  the Royal College of  Psychiatrists  (1979). Alcohol anddlcoholism. London:

Tavistock Publications.

Wilson, P.  (1980). Drinking  in England and Wales. London: H.M. Stationery Office.

World Health  Organization  (1982). Prevention  of  Coronary Heart Disease.  Technical Report

(1983).

132-137.

Physicians 10, 213-275.

Series no. 678. Geneva: WHO.