Obesity in a Primary Health Care Centre: A Retrospective Study

Introduction

Obesity is not an uncommon finding, particularly in affluent societies^[1,6], and it is the associated medical and social risks that especially make it a problem. Garrow^[1] devised a chart for classifying obesity based on the body mass index (BMI), a concept that was first suggested over a hundred years ago by Quetelet, a Belgian mathematician^[2]. Billewicz et al^[3] compared various indices of adiposity and concluded that, while none was perfect, Quetelet's obesity index was reasonably satisfactory over a wide range of heights. Information on the prevalence of obesity in various countries^[4] has been different mainly because there is no universally accepted index of obesity^[5]. However, Garrow^[1] estimated that, among adults in the U.K. and U.S. (using Quetelet's index), 25 to 33% have grade I obesity (BMI, 25 to < 30 kg/m²), 4% have grade II obesity (BMI, 30 to 40 kg/m²), and 0.05% have grade III obesity (BMI > 40 kg/m²). Other studies^[6], using 20% above desirable weight as an index of obesity, have estimated the prevalence rate for the U.S. population to be 24% for women and 14% for men, aged 18 to 74 years.

There are no known official data on the prevalence rate of obesity in the population at large in Saudi Arabia, though it is suspected to be high. The objective of this study was to estimate the percentage of women who are obese and to compare them with their male counterparts. Only Saudi men and women were studied. No attempt was made to examine socioeconomic factors, marital status, or parity.

SUBJECTS AND METHODS

Patients seen at the primary health care centre of King Fahd Hospital of the University routinely have their heights and weights measured at the first visit. Only weight is recorded at subsequent visits. Weight was measured (with patients wearing light indoor clothing) in kilograms to the nearest 100 grams using a lever balance made by Seca (Germany) placed on a hard, level, uncarpeted floor. The instrument is recalibrated every six months. Height is measured with the patient standing upright with the entire foot planted on the platform. Measurements are made against a flat vertical surface using the fitted sliding headpiece of the lever balance and recorded to the nearest 0.5 cm.

The files of patients were retrospectively reviewed and only heights and weights recorded at the first registration were used. Only Saudi men and nonpregnant Saudi women between the ages of 18 and 74 years were included. Patients with either type of diabetes mellitus or diseases associated with fluid retention were excluded. The BMI of each patient was then calculated by dividing the weight (in kilograms) by the height (in meters squared).

Data were entered into the computer and analyzed using SPSS/PC+. The BMI for each subject was computed and prevalence rates of the various grades of obesity were estimated. Tests of hypothesis to determine the significance of the difference between the two sexes were carried out.

Obesity was subdivided into three grades in accordance with the Garrow classification^[1]; mild (grade I), BMI 25 to < 30 kg/m²; moderate (grade II), BMI 30 to 40 kg/m²; and severe (grade III), BMI > 40 kg/m².

RESULTS

The records of 1072 Saudis (477 men and 595 women) between the ages of 18 and 74 years were analyzed.

Table 1 shows the distribution of females and males according to their BMI. Of the total females, 206 (34.6%) were nonobese. Of those who were obese, 203 (34.1%) had grade I obesity, 166 (27.9%) had grade II obesity, and 20 (3.4%) had grade III obesity. Of the 477 men, 231 (48.4%) were nonobese, 167 (35%) had grade I obesity, 76 (15.9%) had grade II obesity, and only three (0.63%) had grade III obesity.

Table 1. Distribution of 595 women and 477 men aged 18–74 years, according to BMI.

The between-sex differences in obesity rates are shown in Table 2. Severe obesity was significantly more prevalent in females as compared with males (P = 0.001). Although there was no significant difference in the prevalence rate of mild obesity (P = 0.4), the overall difference in the rates of all grades of obesity was significantly higher in women (P <0.0001).

Table 2. Test of significance in obesity rates between men and women aged 18–74 years.

Detailed statistical studies of simple regression analysis of weight on age showed poor correlation for both sexes, (r² = 0.03154 for females, 0.04324 for males). Similarly, poor correlation was demonstrated between BMI and age for both sexes (r² = 0.05386 for females, 0.081892 for male).

Table 3 and Figure 1 show a comparison of distribution of males and females according to BMI and subdivided by the arbitrary age groups 18–35 yr, 36–55 yr, and 56–74 yr. In all age groups, more women were moderately or severely obese. However, an equal proportion of men and women were mildly obese in age groups 18 to 35 and there was a slight male preponderance in the 36 to 55 year age group. Subjects in age group 36 to 55 years were significantly more obese than those in age group 18 to 35 years (Table 4), and women predominated in both age groups (P <0.0006).

Table 3. Comparison of distribution of men and women according to BMI and age groups.

Table 4. Test of significance in obesity rates by age groups and gender.

DISCUSSION

Although mild obesity (grade I) is not considered to pose an appreciable risk^[1], marked obesity (grades II and III) is associated with increased morbidity and mortality^[7]. Therefore it is important to determine its prevalence in a society so that proper preventive measures can be instituted, if findings indicate a serious problem.

In the present study, marked obesity was found in 16.5% of the men and 31.26% of the women. Overall, marked obesity was found in 24.72% of this adult population. This is considerably higher than the percentage cited by Garrow^[1], who found marked obesity in only 4.05% of adults of both sexes in the U.K. and U.S. The reasons for these differences are uncertain, but may be related to differences in dietary and sociocultural habits and the environment. The hot climate, especially in the summer, tends to discourage even the more ambitious from engaging in more physically demanding recreation.

Using a BMI > 30 kg/m² as the criterion for obesity in 467 nonpregnant Saudi women aged 15 to 49 years, Khwaja et al^[8] observed an overall prevalence rate of 27% which was not statistically different (P = 0.2) from the 29.7% (95% confidence interval, 21.4–38.0) observed in our study which used similar criteria. However, they found that obesity was significantly related to age, and this was not the case in our study. The significant difference in prevalence for adult women and men cannot be accounted for by a difference in age, as both subpopulations had a similar mean age. This sex-related difference has also been observed in the Ten State Nutrition Survey^[9]. The reasons are uncertain, but the difference may reflect different eating patterns or hormonal secretions. The obesity rate in a population increases considerably at about 40 years of age^[10]. In our study population, obesity was uncommon in the relatively younger age groups. In our environment, and particularly in urban areas, women tend to lead a predominantly sedentary life, thus increasing the likelihood of obesity. Women do not have as much active outdoor life as men.

Attitudes toward obesity vary with culture. A community-based study conducted by one of the authors (TB) revealed that women who were visibly obese did not consider themselves so. This sociocultural attitude was also suggested by Dawson^[11], who analyzed the ethnic differences in overweight females using data from the 1985 National Health Interview Survey and who remarked that “women's perceptions of whether or not they are overweight are more strongly influenced by their weight relative to their peers than by their weight relative to an arbitrary, health-based standard.”

In the present population, mild and moderate obesity were common. Our results are not necessarily representative of other areas in the Kingdom and thus further studies are needed to delineate the extent of obesity on a national level. Our results, however, do suggest that obesity is a major problem for the family doctor and the primary health care team who must be actively involved in its prevention. In view of the polygenic origin^[12] of obesity, it is important in constructing treatment programs for the obese that factors such as sociocultural background be considered. The primary health care teams are ideally suited for detecting and preventing obesity.